Macro Roundup

This paper highlights that the way that housing inflation was measured in the CPI made previous inflationary cycles look more volatile and responsive to Fed policy. We draw two sets of conclusions. First, our observations imply that the current inflation regime is closer to that of the late 1970s than it may at first appear. In particular, the rate of CPI disinflation engineered in the Volcker-era is significantly less when measured using today’s treatment of housing. In order to return to 2 percent core CPI today, we need nearly the same 5 percentage points of disinflation that Volcker achieved. In its pre-1983 form, shelter inflation would decline mechanically as interest rate increases subsided, adding fuel to rapid disinflations. Not only is the current housing component of CPI unlikely to decline swiftly, but recent research indicates that housing inflation is likely to continue growing in the coming monthsWith private sector rent growth currently still at 16 percent, residential inflation is likely to move towards 7 percent by the end of 2022, contributing almost 3 percentage points to core CPI inflation. Compared to previous inflationary cycles, housing will thus serve as a significant hindrance to rapid disinflation whereas it used to move the official series substantially lower.Second, our estimates suggest that past inflation cycles were more volatile than today’s due to the greater weight of transitory goods components in past indices. With current procedures, the inflation taken out of the system in 1950s and 1970s looks significantly more modest. Overall, some optimistic interpretations which underplay the magnitude of the required disinflation to return to trend inflation today are called into question by the analysis in this paper. Our estimates suggest that the current inflation rate is closer to the peak of other cycles than the official CPI data suggest. Figure 3 shows that the peak of the Volcker-era inflation (March 1980), currently understood to have been at 14.8 percent, is only 11.4 percent when adjusted for the switch from homeownership costs to OER. The growth in core CPI at its peak in June 1980 falls from 13.6 percent to 9.1 percent when measured using the OER method as seen in Figure 4. The large differences between the official and adjusted series reflect both the substantial weight of OER in the index, especially in core CPI, and the lower peaks of estimated OER relative to homeownership costs. From a low of 14.5 percent weight in 1983, as Americans have shifted more of their consumption towards housing, OER has risen to represent 24.3 percent of overall CPI and 30.6 percent of core CPI in 2022. While past inflation peaks are lower using the consistent methodology, the average inflation rate pre-1983 would also have been lower. Our estimates show that the mean headline inflation rate between 1949 and 1983 is 0.4 percentage points lower when accounting for the shift to OEROne thing that is not called into question is the great work performed by the people of the Bureau of Labor Statistics; our corrections do not at all denigrate current processes, they are just a tool for better intertemporal contextualization. Most importantly, prior to 1983 measurement of shelter inflation was mechanically responsive to Federal Reserve interest rate policy via mortgage rates. This method made pre-1983 peak CPI inflation measures, especially during the Volcker-era, artificially high at the beginning of the tightening cycle, and declines look artificially fast. Using publicly available Bureau of Labor Statistics (BLS) data for the post-war period, we develop new estimates of CPI headline and core inflation that can be better compared across time. The full series is available on our website athttp://larrysummers.com/category/inflation/. Our analysis reveals that current inflation, especially core inflation, is considerably closer to previous peaks than in the official series. Official core CPI inflation peaked at 13.6 percent in June 1980, whereas we estimate that core inflation was 9.1 percent in that same month when adjusting for the treatment of shelter inflation. Our estimates also suggest that the local trough of core CPI inflation in 1983 was considerably higher than originally reported. Overall, these estimates imply that the rate of core CPI disinflation caused by Volcker-era policies is significantly lower when measured using the current treatment of housing: only 5 percentage points of decline instead of 11 percentage points in the official CPI statistics. To return to 2 percent core CPI today, we thus need disinflation of a similar magnitude as Chairman Volcker achieved. We show that due to the greater weight of transitory goods components —especially food and apparel— in the index of the 1940s and 1950s, past inflation spikes were higher and more short-lived than today’s. When using current weights, we estimate that the peak of core CPI inflation in June 1951 falls from 7.2 to 5 percent, and the peak of headline CPI inflation falls from 9.4 to just 3.3 percent. These two points serve as a caution against overly optimistic forecasts of an inexpensive disinflation in the current cycle —the disinflation that needs to be achieved now is large by historical standards.

Methodology

“…To better contextualize the current run-up in inflation, this paper constructs new historical series for CPI headline and core inflation that are more consistent with current practices and expenditure shares for the entirety of the post-war period. Using publicly available Bureau of Labor Statistics (BLS) data for the post-war period, we develop new estimates of CPI headline and core inflation that can be better compared across time. The full series is available on our website athttp://larrysummers.com/category/inflation/. Our analysis reveals that current inflation, especially core inflation, is considerably closer to previous peaks than in the official series. Official core CPI inflation peaked at 13.6 percent in June 1980, whereas we estimate that core inflation was 9.1 percent in that same month when adjusting for the treatment of shelter inflation….”

Implication

“…Our estimates also suggest that the local trough of core CPI inflation in 1983 was considerably higher than originally reported. Overall, these estimates imply that the rate of core CPI disinflation caused by Volcker-era policies is significantly lower when measured using the current treatment of housing: only 5 percentage points of decline instead of 11 percentage points in the official CPI statistics. To return to 2 percent core CPI today, we thus need disinflation of a similar magnitude as Chairman Volcker achieved….”

Other Comparisons

“…Similar issues affect conclusions drawn from comparisons of current inflation with other periods of elevated inflation. Recent work suggests that the years following World War II have strong similarities to the current inflation environment (e.g., Rouse et al., 2021; DeLong, 2022). We show that due to the greater weight of transitory goods components —especially food and apparel— in the index of the 1940s and 1950s, past inflation spikes were higher and more short-lived than today’s. When using current weights, we estimate that the peak of core CPI inflation in June 1951 falls from 7.2 to 5 percent, and the peak of headline CPI inflation falls from 9.4 to just 3.3 percent. These two points serve as a caution against overly optimistic forecasts of an inexpensive disinflation in the current cycle —the disinflation that needs to be achieved now is large by historical standards…”

The Evidence

“…Our estimates suggest that the current inflation rate is closer to the peak of other cycles than the official CPI data suggest. Figure 3 shows that the peak of the Volcker-era inflation (March 1980), currently understood to have been at 14.8 percent, is only 11.4 percent when adjusted for the switch from homeownership costs to OER. The growth in core CPI at its peak in June 1980 falls from 13.6 percent to 9.1 percent when measured using the OER method as seen in Figure 4. The large differences between the official and adjusted series reflect both the substantial weight of OER in the index, especially in core CPI, and the lower peaks of estimated OER relative to homeownership costs. From a low of 14.5 percent weight in 1983, as Americans have shifted more of their consumption towards housing, OER has risen to represent 24.3 percent of overall CPI and 30.6 percent of core CPI in 2022. While past inflation peaks are lower using the consistent methodology, the average inflation rate pre-1983 would also have been lower. Our estimates show that the mean headline inflation rate between 1949 and 1983 is 0.4 percentage points lower when accounting for the shift to OER. More broadly speaking, past inflationary cycles would have been less volatile using the consistent methodology that uses OER. The pace of reflation during the cycle upswings, and the pace of disinflation during the cycle downswings are lower under today’s methodology, as summarized in Table 1.These differences imply that the responsiveness of the CPI to monetary policy was considerably lower during the 1960s and 1970s than the official CPI statistics suggest. We stress that any consequences of the difference in measurement are larger for core CPI than for headline CPI, due to the considerably larger weight of OER in core CPI..”

Marijn Bolhuis, Judd Cramer and Lawrence Summers, "Comparing Past and Present Inflation," National Bureau Of Economic Research, June 2022, https://www.nber.org/papers/w30116

David Autor, "The Labor Market Impacts of Technological Change: From Unbridled Enthusiasm to Qualified Optimism to Vast Uncertainty,"National Bureau Of Economic Research, June 2022, https://www.nber.org/papers/w30074

Autor, “…I began by asking what the role of technology—digital or otherwise—is in determining wages and shaping wage inequality. I presented four answers corresponding to four strands of thinking on this topic: the education race, the task-polarization model, the automation-reinstatement race, and the era of AI uncertainty.. Fundamentally, technological change expands the frontier of human possibilities, but one should expect it to create many winners and many losers, and to pose vast societal challenges and opportunities along the way. What are the policy implications of these observations? The question is so broad that almost any answer is bound to appear vague and inadequate. One can reliably predict that technological innovations will foster new ways of accomplishing existing work, new business models, and entirely new industries, and these in turn will generate new jobs and spur some productivity gains. But absent complementary institutional investments, technological innovation alone will not generate broadly shared gains. sketch a long-form policy vision of what form these investments may take, focusing on three domains: education and training; labor market institutions; and innovation policy itself. I provide a capsule summary of those conclusions here, stopping short of policy recommendations. As a starting point, educating and training the workforce to meet the demands of the moment can improve access to good jobs for workers who would otherwise face barriers and boost the quality of existing jobs by creating opportunities for career ladders. While boosting the supply of skills is never controversial, pure supply-side policies are inadequate to meet the gravity of the labor market challenge. In 1979, 60% of U.S. males at the median of the wage distribution possessed high school or lower education, whereas 21% had some college, and 20% held a bachelor’s degree or above. By 2019, 31% of males at the median of the earnings distribution had some college, 36% had attained at least a four-year college degree — a 75% increase — and only one-third had high school or less education. Despite this dramatic skill upgrading, the real earnings of the median male rose by only 10% in these forty years. More broadly, over more than four decades, the link between rising productivity and commensurate improvements in job opportunities and earnings has decoupled for the majority of U.S. workers. The poor quality of jobs available to workers lacking four-year college degrees or specialized credentials provides one of the starkest examples of this failure. Low-wage U.S. workers earn substantially less than low-wage workers in almost all other wealthy industrialized countries. The divergence between the upward path of productivity growth and the near plateauing of median wage growth among U.S. workers was not an inevitable consequence of technology, globalization, or market forcesRather, a set of U.S.-specific institutional and policy choices failed to blunt—and in some cases magnified—the consequences of technological and globalization pressures on the U.S. labor market. To contend effectively with these challenges would require institutional and policy reforms that realign labor market opportunities with the rising productivity and societal wealth that the U.S. has reaped from decades of innovation and investments in human and physical capital. Thus, a second locus for investment would be to revitalize the governmental, nongovernmental, and private sector institutions that translate—or fail to translate—rising productivity into shared prosperity. Some such steps might include updating and more vigorously enforcing labor standards, recalibrating federal minimum wage policies, extending the scope and flexibility of the unemployment insurance system, and transforming the U.S. employer-based health insurance provision into a system with portable benefits. further call for reevaluating the U.S.’ uncritical embrace of pure shareholder capitalism. While shareholder capitalism can plausibly be credited with some of the productive dynamism of the U.S. economy, it has also arguably helped fuel the drive to curtail wages and benefits for low-wage workers and, more broadly A third potential domain for policy is to directly shape innovation itself to speed productivity growth and complement the skills of the labor force. It is well known that the U.S. has a strong national innovation system, fueled by federal R&D investments, to develop fundamental science and new technologies, that has led to scientific leadership and new industries Less recognized, however, is the crucial link between those new industries as complements to the inevitable loss of jobs that results from productivity-enhancing technologies. New industries grew out of a flourishing innovation ecosystem that created new companies and new applications, alongside older industries that increased mechanization and automation as they matured. Arguably, the U.S. has let those important R&D investments wither Through increased and targeted R&D investments supported by a reinvigorated federal R&D program, as well as a tax policy that keeps workers and social challenges at the forefront, the country’s innovation system could be put to work for a broader number of people and regions than it has in recent decades Although I would prefer to end this essay with optimistic assurances,I will instead end with one uncertainty, one certainty, and one admonition. The uncertainty is that we have less clarity about our technological future than we did two decades ago. AI has extended the frontier of technological possibility towards boundaries that are barely visible at present. The tasks that machines will be able to accomplish, the rate at which new innovations may emerge, and the speed with which socially impactful technological innovations may diffuse is unknown. But the range of possibilities has surely gotten broader, and our certainty about the boundaries has accordingly diminished. Conversely, the certainty is that these technological advances will expand the set of desirable possibilities that are within the reach of humanity. Artificial Intelligence has the potential to help humanity tackle some of its most pressing challenges: climate change, disease, poverty, malnutrition, and inadequate education. But whether societies will successfully realize this potential, or instead squander it or, worse, disastrously misuse it, is highly uncertain and, I would argue, fundamentally indeterminate. The admonition is this: given the potential applicability of AI to a vast set of purposes, we collectively (meaning individuals, organizations, and governments) should not simply be asking what AI will accomplish but what we want it to accomplish. How can we use AI most productively to complement workers, raise productivity, and more broadly, tackle humanity’s most pressing challenges? Simultaneously, how can we blunt or reshape the commercial incentives to use AI for socially counterproductive objectives such as displacing workers, preying upon people’s cognitive and emotional frailties, or consolidating the power of governments or corporations to exercise social control As we ponder our uncertain AI future, our goal should not merely be to predict that future but to create it….”

“…How do pretax incomes vary in Europe and the United States today? Table III provides a first answer to this question by displaying the average incomes and income shares of key income groups in Western Europe, Northern Europe, Eastern Europe, and the US in 2017. The average national income per adult stood at e52,700 in the US at purchasing power parity, compared to e44,900 in Northern Europe, e35,300 in Western Europe, and e21,700 in Eastern Europe. In Europe, only Norway (e55,000) and Luxembourg (e102,000) have higher average national incomes than the US.23Things look very different at the bottom of the pretax income distribution. Thebottom 50% earned only about e12,300 in the US in 2017, compared to e21,600 inNorthern Europe and e14,600 in Western Europe. Of the twenty-seven countriesconsidered in this paper, the US thus ranks third in terms of average national incomeper adult but nineteenth when it comes to the average income of the poorest 50%.24On average, pretax income inequality at the bottom is lowest in Northern Europe(with a bottom 50% share of 24%), followed by Western Europe (21%) and EasternEurope (20%). With a bottom 50% pretax income share of only 11.7%, the US is byfar the most unequal of all countries, followed by a distant Serbia (16%) and veryfar from the values observed in the Czech Republic, Iceland, Norway, and Sweden(all above 25%).25These differences appeared even more pronounced at the verybottom of the distribution: the average income of the poorest 20% was e11,600 inNorthern Europe in 2017, more than three times larger than its counterpart in theUnited States (e3,800).The same differences are visible at the top end of the distribution: the top 1% captured 21% of total pretax income in the US in 2017, compared to 12% in Eastern Europe, 11% in Western Europe, and less than 9% in Northern Europe. In 2017, the top 0.001% average pretax income exceeded e92 million in the US, nearly ten times the value observed in Northern Europe. The European countries with lowest top 1% income shares are the Netherlands, Slovenia, Iceland, Belgium, and Finland (less than 9%), while those with highest top income concentration are Germany, the United Kingdom, Greece, and Poland (13-15%).26In summary, while the US stands out as being richer than most European countriestoday, differences in average national incomes mask substantial heterogeneity. With inequality levels surpassing by far those observed in any European country, theUS displays bottom pretax average incomes that barely exceed those observed inpoorer Eastern European countries. In contrast, the lower inequality levels and higheraverage incomes observed in Northern Europe imply significantly better standardsof living for the majority of the population than in the United States….”

Here is Blanchet, “…theUS displays bottom pretax average incomes that barely exceed those observed inpoorer Eastern European countries…”

David Autor, David Mindell and Elizabeth Reynolds, “The Work Of the Future,” MIT, 2020, https://workofthefuture.mit.edu/wp-content/uploads/2021/01/2020-Final-Report4.pdf

Autor, “..By almost every measure of job quality — pay, working environment, prior notice of job termination, and access to paid vacation, sick time, and family leave — less-educated and low-paid U.S. workers fare worse than comparable workers in other wealthy industrialized nations. Figure 7 provides one such benchmark, comparing purchasing-power adjusted gross (pre-tax) hourly pay of low-skill workers in 22 OECD countries in 2015.32 Low-skill U.S. workers earn only 79% as much as low-skill U.K. workers, only 74% as much as low-skill Canadian workers, and only 57% as much as low-skill Germans. While no single metric makes for a complete comparison, numerous analyses support the qualitative picture painted by Figure 7…”

“…Yet, as shown in the table below, the financing turned out to be off by at least $282 billion. Instead of a reduction, the ACA has caused an estimated $158 billion increase in the deficit during its first ten years (or about 40 percent of the health care budget increase). Moreover, because many of the now missing taxes and savings were back-loaded in the ACA, the increase in the deficit from the ACA in its second decade is even larger. This clearly confirms that the ACA was never fully paid for….”

Mark Warshawsky, "The ACA Was Never Fully Paid For,"Real Clear Policy, May 27, 2022, https://www.realclearpolicy.com/articles/2022/05/27/the_aca_was_never_fully_paid_for_834472.html

The ACA Was Never Fully Paid For

The White House celebrations for the twelfth anniversary of the passage of the Affordable Care Act (ACA) remind us that twelve years is a sufficient passage of time to compare the initial claims of the ACA designers and promoters to the actual outcomes and consequences of this massive, partisan, legislation. In the following analysis, I focus on the ACA financing.

The Congressional Budget Office (CBO) estimated on March 20, 2010 that the ACA’s expansion of health insurance coverage would cost the federal government $938 billion from 2010 through 2019. This gross expenditure would consist mainly of subsidies for purchasing insurance on the exchanges, expanding Medicaid eligibility, and enhancing Medicaid cost-sharing with the states. The CBO further estimated that — taking into account the reductions in payments to providers and certain tax features — the ACA would increase the federal budgetary commitment to health care by $390 billion. The health care and revenue provisions actually were estimated to reduce the deficit by $124 billion from 2010 to 2019, and by up to one-half of a percentage point of GDP in the subsequent decade. The financing would come from many sources and new programs, including several new taxes, mainly payroll ones, and spending cuts to Medicare providers. While promoting the ACA in 2009 and 2010, President Obama announced that it would be fully paid for.

Yet, as shown in the table below, the financing turned out to be off by at least $282 billion. Instead of a reduction, the ACA has caused an estimated $158 billion increase in the deficit during its first ten years (or about 40 percent of the health care budget increase). Moreover, because many of the now missing taxes and savings were back-loaded in the ACA, the increase in the deficit from the ACA in its second decade is even larger. This clearly confirms that the ACA was never fully paid for.

Another miss was the so-called Cadillac tax, which was an excise tax on expensive employer health insurance. Aimed at employees, it was a back-door and inexact limitation on the deductibility of health insurance premiums from income. This was intended to reduce the pace of health care price increases by discouraging complete insurance coverage. The provision was unpopular especially with unions and government workers. It was twice postponed and finally eliminated in 2019, by a large majority in Congress.

Other taxes removed in 2019 on a bipartisan basis were the excise tax on medical devices and fees on health insurers. The penalty on individuals without insurance was reduced to zero by Congress in 2017. A similar penalty on large employers who do not offer their workers adequate health insurance is still in the law, but according to the Treasury Inspector General, using current assessment rates, less than five percent of the revenues projected by the Joint Committee on Taxation will be collected, so I count it as a complete loss. The ACA reduced Medicaid payments to hospitals with a disproportionate share (DSH) of Medicaid and uncovered patients. Because this reduction has been repeatedly postponed and suggested for elimination, I regard it too as a financing loss. Finally, the ACA created a new government agency, the Center for Medicare and Medicaid Innovation (CMMI), whose purpose was to develop successful projects able to reduce costs. In the CBO’s estimate, this would result in significant net budget savings over the long-run. The reality, however, has been quite disappointing and the CMMI is now emphasizing health equity in its agenda.

Some argue that I should also add the spending misses on the plus side for the ACA. These arose because enrollment in the exchanges was lower than expected, some states did not expand Medicaid, and overall health care prices and use have come in below projections owing to recent economic trends. But then I should also count the significant increases in the exchange subsidies, and Medicaid enrollment and spending arising from the passage of the American Rescue Plan in 2021. These countervailing considerations may cancel out, but are quite difficult to calculate. It thus makes more sense to focus on discrete items from the original CBO score.

Almost all Americans care about health care insurance. A more stable and complete financing for legislation expanding health care coverage would have surely occurred had it passed originally on a bipartisan basis.

Excess savings is the rise in cumulative total assets that is attributable to an abnormal rise in personal saving. As an example, in April 2020 the personal saving rate hit a record 33% (compared to about 7% pre-pandemic), boosting excess savings. Today, the aggregate personal saving rate is 4.4%, drawing down excess savings. What’s now pulling the stock of excess savings down? It’s chiefly a rise in nominal spending, not a drop in personal income. By mid-2021 personal outlays had bounced back to their pre-pandemic trend and by early 2022 they jumped far above trend. As the personal saving rate shrank the cumulative stock of excess savings began to decline. So far, Americans have spent down approximately $150B of the $2.5T in extra aggregate savings that they accumulated in 2020 and 2021. Excess savings have decreased, driven by high-income households spending more (and being taxed on their income/capital gains).

Joseph Politano, "Americans are Spending Their Excess Savings,"Apricitas Economics, June 4, 2022, https://apricitas.substack.com/p/americans-are-spending-their-excess

Americans are Spending Their Excess Savings

Or, well, not exactly. The jump in savings are a mechanical result of how deficit-financed government transfers show up in the official National Income and Product Accounts. For the public sector to run a deficit the private sector must run a surplus; for the public sector to go into debt the private sector must be building up savings. A massive shock to the size of the public sector deficit will create a corresponding large shock to the size of private sector savings—hence why deficit-financed stimulus resulted in excess savings.

Neither of those points means that excess savings are fake—just that they’re more complicated than they initially appear. It also means that we have to be careful in analyzing how Americans are spending down those savings as the economy continues to reopen.

An Excess Penny Saved…

First, let’s clarify some points. Personal saving—in the National Income and Product Accounts—represents the sum total of aggregate personal income that is not spent on aggregate personal outlays. In other words, if you summed together all the wages, dividends, government transfers, etc of US households and subtracted out their consumption spending, taxes, interest expenses, etc you would arrive at personal saving. That leaves a lot of ways for personal saving to manifest—increased bank account balances, additional retirement contributions, cryptocurrency purchases, paying down the principal of mortgages or other household debt, and even home remodeling (technically categorized as “residential investment”). In other words, it is not like there is precisely an additional $2.5 trillion sitting in household bank accounts because the personal saving rate has increased by $2.5 trillion.

It’s also important to distinguish between the flow of saving and the stock of savings. The personal saving rate represents the share of income saved in a given time period (say, one month) while personal savings represent the total value of accumulated assets. Excess savings is the rise in cumulative total assets that is attributable to an abnormal rise in personal saving. As an example, in April 2020 the personal saving rate hit a record 33% (compared to about 7% pre-pandemic), boosting excess savings. Today, the aggregate personal saving rate is 4.4%, drawing down excess savings.

The biggest positive contributors to excess savings are obvious; increased government transfers and decreased personal consumption expenditures. Those categories more than offset the drop in labor and capital income that occurred in 2020 and 2021. But again remember that for the public sector to go into debt the private sector must be building up savings; a positive contribution from government transfers only indicates that deficit financed spending was directed at households. Also, it would be hard for the contribution from government transfers to decrease significantly as definitionally public spending is classified as transfers and public revenue is separately classified as taxes.

Still, it is important to note that the big downward shifts in the stock of excess savings come from rising tax payments and personal consumption expenditures. The rise in consumption spending comes from a steady increase in services spending as the economy reopens, but the rise in taxes is a bit more complicated. Comprehensive data is unfortunately unavailable, but it looks as though rising tax receipts come from a combination of higher income taxes thanks to rising wages and employment levels alongside a jump in capital gains taxes given appreciating asset prices.

Now, it is important to mention here that it remains wrong to reason from accounting identities regarding excess savings. While it is true, definitionally, that for domestic savings to decrease then the savings of the government, foreign, or corporate sectors must increase it is not true that an increase in non-household savings will cause a decrease in domestic savings. As an example, an explicit attempt to increase the trade deficit could hypothetically result in lower domestic employment and lower government income tax revenues, thereby keeping aggregate domestic private savings constant even as the trade deficit increased.

It is also important to note the shifts in the composition of financial assets even as aggregate savings remain unchanged. The Federal Reserve’s Quantitative Easing can result in the transformation of household/corporate Treasury/MBS assets into bank deposits, and Quantitative Tightening can do the opposite. Indeed, that is exactly consistent with the real life bank deposit data; aggregate deposits increased tremendously at the beginning of the pandemic as QE went into effect and are decreasing now as QT goes into effect.

The Micro View

At some level, this should not be surprising. The top 1% hold about 30% of all assets in America, and the top 10% earn about 30% of all income. We know that these households have the highest discretionary spending budgets, lowest marginal propensity to consume, and were least likely to lose their jobs during the pandemic. It makes sense that they would end up holding the bulk of excess savings, stimulus checks or not. Critically, it’s also worth putting the $2.5 Trillion in excess savings in context against the $35 Trillion increase in aggregate household net worth—which also disproportionately benefitted higher-income and wealthier Americans.

Conclusions

Back in January, I laid out my thoughts on the future path of excess savings as such:

While most of the excess savings is held by high-income Americans, who do not have a particularly high marginal propensity to consume, I do think that high income earners will consume more of their excess savings than they would under a traditional increase in wealth. For one, high-income households have been stockpiling extremely liquid financial assets, an indication that they may only be waiting for a good opportunity to spend their money. For two, historical episodes of excess savings (like the post-WWII era) have seen a slow but steady drawdown of excess savings over the following 5-10 years. For three, polling from earlier this year showed that consumers intended to spend down some of their excess savings, and personal consumption expenditures have already exceeded their pre-COVID trend.
Generally, that trend has borne out over the last 6 months. Excess savings have decreased, driven by high-income households spending more (and being taxed on their income/capital gains). However, I underappreciated the downside risks to lower and middle-income household’s balance sheets—risks that have only intensified since then.
The good news on that front is, despite an appreciable slowdown in the labor market, employment levels and wage growth remain strong. Most Americans’ most valuable asset is their job, and that is especially true of lower and middle-income Americans. Without a job, however, they would not be able to rely on excess savings.

When it comes to China, official U.S. trade statistics miss a big part of the story. They don’t include the value of roughly 800 million shipments a year that arrive in the U.S. under a set of rules known as de minimis, a Latin phrase meaning items too small to be bothered with, that applies to individual imports entering the U.S. valued under $800. Correcting for that suggests the deficit with China might have barely budged, and possibly grown, since the trade war. The Journal obtained data from a Customs database of known de minimis shipments and their declared value. The data showed nearly 800 million such packages enter the U.S. annually—more than two million a day—with as much as half coming directly from China. The data show that the known reported value of shipments surged from $40.5 million in 2012 to a peak of $67 billion in fiscal year 2020, with $46 billion directly from China. (In 2021, as the pandemic faded, the known value of shipments declined to about $40 billion.)There are flaws in this data. Many packages enter via the postal service. In some cases, Customs only knows their weight, not the value. Data show the country from which a shipment entered the U.S., rather than the country of origin. This is part of why it hadn’t been publicly released;Customs is working to produce official statistics.

Josh Zumbrun, "The Tiny Loophole That Understates the Trade Deficit With China,"Wall Street Journal, June 17, 2022, https://www.wsj.com/articles/the-tiny-loophole-that-understates-the-trade-deficit-with-china-11655458201

The Tiny Loophole That Understates the Trade Deficit With China

American politicians and economists have obsessed over the U.S. trade deficit with China for years.

Lowering the deficit was one of former President Donald Trump’s goals in imposing tariffs on the majority of Chinese imports. President Biden’s administration is debating whether to peel them off to alleviate inflation or keep them on to pressure China. After all, they did appear to achieve one of Mr. Trump’s purposes: reducing the bilateral deficit between the U.S. and China, which according to the official U.S. trade statistics topped $400 billion a year in 2018.

But here’s the catch: When it comes to China, official U.S. trade statistics miss a big part of the story. They don’t include the value of roughly 800 million shipments a year that arrive in the U.S. under a set of rules known as de minimis, a Latin phrase meaning items too small to be bothered with, that applies to individual imports entering the U.S. valued under $800. Correcting for that suggests the deficit with China might have barely budged, and possibly grown, since the trade war.

“The whole idea of ‘de minimis’ is you send in packages, and Customs isn’t checking the values,” said Jeff Ferry, an economist at the Coalition for a Prosperous America, a trade group that supports tariffs as a tool to help U.S. firms compete with China.

“This has implications for whether we can trust the figures on imports from China” where such shipments have become especially common, said Mr. Ferry, who in a study earlier this year determined U.S. trade statistics omit the value of such shipments.

The rise of e-commerce made it easy to search online for low-cost goods shipped directly to your door. The U.S.-China trade war created a tremendous additional incentive. One reason you find such cheap goods online is because it is increasingly common to ship them into the country de minimis, avoiding tariffs. A U.S. retailer, by contrast, would have to pay the tariff on a large shipment of the same item.

For most of the decades since its creation by Congress in 1938, the use of de minimis was small—as trivial as intended. Congress stipulated the purpose was “to avoid expense and inconvenience to the government disproportionate to the amount of revenue that would otherwise be collected.”

The Journal obtained data from a Customs database of known de minimis shipments and their declared value. The data showed nearly 800 million such packages enter the U.S. annually—more than two million a day—with as much as half coming directly from China.

The data show that the known reported value of shipments surged from $40.5 million in 2012 to a peak of $67 billion in fiscal year 2020, with $46 billion directly from China. (In 2021, as the pandemic faded, the known value of shipments declined to about $40 billion.)

There are flaws in this data. Many packages enter via the postal service. In some cases, Customs only knows their weight, not the value. Data show the country from which a shipment entered the U.S., rather than the country of origin. This is part of why it hadn’t been publicly released; Customs is working to produce official statistics.

As de minimis went from trivial to substantial, the Census Bureau’s official trade statistics didn’t change to reflect this. (Other countries have their own de minimis rules, but U.S. exports that qualify for that treatment aren’t thought to be significant. China’s corresponding threshold is only 50 yuan, or about $7.45.)

The Census Bureau said in a written statement that shipments meeting the de minimis criteria aren’t received by Census, and thus not included in its estimation process. The Census declined to comment on estimates about the scale of de minimis, but said that if Customs published official de minimis statistics “we would intend to research the quality of those statistics to evaluate whether they could be incorporated into the official merchandise trade statistics.”

Mr. Ferry, the economist, estimates the true value could be nearly twice as high as what Customs data captures. He looked at Securities and Exchange Commission filings of the largest U.S. internet retailers, and calculated they are bringing in $112 billion of goods a year that likely qualify for de minimis. Even this figure doesn’t capture China-based companies whose entire business model relies on duty-free shipments, but don’t file with the SEC.

For example, officially, the U.S.-China goods deficit reached $402 billion in fiscal year 2018 but narrowed to $302 billion in 2020, as tariffs and then the pandemic damped trade.

But in fact, de minimis shipments surged during this period, precisely to dodge tariffs and because of online shopping during the pandemic. The different measures and estimates of de minimis are crude; they have none of the precision of our data on imports of, say, cars. But they imply the fiscal 2020 deficit was probably at least $350 billion and perhaps larger than ever at $430 billion.

During those early days of the pandemic, people were ordering everything online. It was taking a long time to arrive. How much came from China? Certainly a lot. More than the statistics capture. How much exactly? No one knows.

When President Trump entered the Oval Office, CBO projected the cumulative 2017-2027 budget deficits would be $10.0 trillion. When he left office four years later, CBO’s projected deficits for the same period were $13.9 trillion. The president signed or enacted $7.8 trillion in new initiatives, the costs of which were partially offset by $3.9 trillion saved from economic growth revenues and technical re-estimates of taxes and spending levels. Economic and technical factors produced a substantial $3.9 trillion in actual and projected savings over this period. Of this amount, $2.7 trillion comes from falling interest rate projections, which reduced the projected cost of net interest on the national debt. Another $1.3 trillion comes from higher tax revenues produced by faster economic growth projections.Technical re-estimates have reduced mandatory spending projections but also tax revenues. Most of these savings are projected to occur later in the 2017-2027 period and thus may not materialize if economic growth slows or interest rates rise. President Trump signed legislation and approved executive actions costing $7.8 trillion over the decade—compared to $5.0 trillion for President Obama and $6.9 trillion for President Bush, and he enacted these costs in just a single four-year presidential term, compared to his predecessors’ eight years in the Oval Office. The largest drivers were pandemic relief legislation ($3.9 trillion), the 2017 tax cuts ($2.0 trillion), and legislation raising the discretionary spending caps ($1.6 trillion). presidents are also limited to signing what Congress will pass. And during the Trump presidency, much of the $8 trillion in enacted legislation was passed by overwhelmingly bipartisan Congressional majorities.Consider the eight bills that Congress passed and President Trump signed costing at least $100 billion over the baseline, plus a separate pandemic relief bill (P.L. 166-159) that accelerated nearly $100 billion in previously enacted expenses. The most partisan of these nine bills was obviously the 2017 tax cuts (which also repealed the Obamacare individual mandate penalty). That law was passed exclusively by Republicans without a single Democratic vote. (See Figure 7 in the appendix for the largest bills signed during the Trump presidency.) The other eight bills—which consisted of five pandemic-relief laws, the 2018 and 2019 Bipartisan Budget Acts that raised the discretionary spending caps, and the 2019 law repealing key Obamacare taxes—were generally bipartisan. In fact, each of the eight bills passed the Senate with the majority support of both Republicans and Democrats. Senate Democrats cumulatively voted 94% in favor of these bills, and Senate Republicans cumulatively voted 76% in favor—not counting the CARES Act that passed the Senate by voice vote. A similar phenomenon occurred during the Obama presidency, when $5.0 trillion in enacted legislation was dominated by $4.1 trillion in bipartisan extensions of various current tax policies such as the 2001 and 2003 tax cuts. In fact, the $1 trillion stimulus legislation in 2009 was the only major deficit-hiking bill passed by an almost-exclusively Democratic majority during those eight years. Over the previous two presidencies, the vast majority of new deficit-expanding bills has been enthusiastically bipartisan.

Brian Riedl, "Trump’s Fiscal Legacy: A Comprehensive Overview of Spending, Taxes, and Deficits,"Manhattan Institute, May 12, 2022, https://www.manhattan-institute.org/trumps-fiscal-legacy

The surviving descendants of the old elite have prospered. By 2010 they were again richer and more educated than the Chinese average, according to data gathered by an international group of academics. Adding new evidence from cities to previous work, which looked at rural areas alone, the authors now find that the elite’s grandchildren have even out-earned Communist Party members.To measure the initial impact of China’s reforms on inequality, the authors compiled data on land ownership in 1950 from archived records. Unsurprisingly, inequality of land holdings, measured by a Gini coefficient, fell sharply after land reform, from 0.5 to 0.1—close to perfect equality. To measure what has happened since then, the authors used a survey of 36,000 Chinese residents from 2010. It recorded earnings and education, as well as social class. The party created hereditary class labels in 1950, in part to punish the old guard. These allowed the researchers to distinguish between old elites and everyone else. The authors found that elites born before 1940 were 7% likelier than their contemporaries to have finished secondary school. Their stigmatised children were 3% less likely to have done so than others their age. By 2010 the children of old elites earned 5% less than other Chinese. But things flipped back. Descendants of the old elite born between 1966 and 1990 were 6% more likely to finish high school than their contemporaries. In 2010 they earned 12% more than other Chinese. They even earned 2% more than party members. The researchers found that the old elite’s grandchildren are more enterprising and work longer hours than the descendants of those who had lower social standing. Although the elite’s capital was destroyed 70 years ago, their social capital has endured.

The authors found that elites born before 1940 were 7% likelier than their contemporaries to have finished secondary school. Their stigmatised children were 3% less likely to have done so than others their age. By 2010 the children of old elites earned 5% less than other Chinese.

But things flipped back. Descendants of the old elite born between 1966 and 1990 were 6% more likely to finish high school than their contemporaries. In 2010 they earned 12% more than other Chinese. They even earned 2% more than party members.

The researchers found that the old elite’s grandchildren are more enterprising and work longer hours than the descendants of those who had lower social standing. Although the elite’s capital was destroyed 70 years ago, their social capital has endured.

Economist Staff, “The grandchildren of China’s pre-revolutionary elite are unusually rich,”The Economist, June 9, 2022, https://www.economist.com/graphic-detail/2022/06/09/the-grandchildren-of-chinas-pre-revolutionary-elite-are-unusually-rich

The grandchildren of China’s pre-revolutionary elite are unusually rich

The land ownership system of feudal exploitation by the landlord class shall be abolished.” So read China’s agrarian reform law of 1950. Land was seized from the better-off and given to poor farmers, whose share of farmland rose from 14% in 1947 to 47% in 1954. Liu Shaoqi, who was president during the Mao era, called it the “most thorough reform in thousands of years of Chinese history”.

Hundreds of thousands of landlords were murdered. Tens of millions of people died in a famine when farms were collectivised. Yet the surviving descendants of the old elite have prospered. By 2010 they were again richer and more educated than the Chinese average, according to data gathered by an international group of academics. Adding new evidence from cities to previous work, which looked at rural areas alone, the authors now find that the elite’s grandchildren have even out-earned Communist Party members.

To measure the initial impact of China’s reforms on inequality, the authors compiled data on land ownership in 1950 from archived records. Unsurprisingly, inequality of land holdings, measured by a Gini coefficient, fell sharply after land reform, from 0.5 to 0.1—close to perfect equality.

To measure what has happened since then, the authors used a survey of 36,000 Chinese residents from 2010. It recorded earnings and education, as well as social class. The party created hereditary class labels in 1950, in part to punish the old guard. These allowed the researchers to distinguish between old elites and everyone else.

We are all familiar with comparisons showing that Black people earn 50% less than white peers and women earn 70 cents for every dollar that a man earns. However, the most popular tools used to combat disparities in the workplace have produced almost no measurable results. The average impact of corporate diversity, equity, and inclusion (DEI) training is zero and some evidence suggests that the impact can become negative if the training is mandated.One of the most important developments in the study of racial inequality has been the quantification of the importance of pre-market skills in explaining differences in labor market outcomes between Black and white workers. In 2010, using nationally representative data on thousands of individuals in their 40s, I estimated that Black men earn 39.4% less than white men and Black women earn 13.1% less than white women. Yet, accounting for one variable-educational achievement in their teenage years--reduced that difference to 10.9% (a 72% reduction) for men and revealed that Black women earn 12.7 percent more than white women, on average. Derek Neal, an economist at the University of Chicago, and William Johnson were among the first to make this point in 1996: “While our results do provide some evidence for current labor market discrimination, skills gaps play such a large role that we believe future research should focus on the obstacles Black children face in acquiring productive skill.”

Roland Fryer, "It’s time for data-first diversity, equity, and inclusion,"Fortune, June 20, 2022, https://fortune.com/2022/06/20/data-first-diversity-equity-inclusion-careers-black-workers-gender-race-bias-dei-roland-fryer/

It’s time for data-first diversity, equity, and inclusion

I have spent hundreds of hours in diversity training over the past two decades-from descriptions of federal anti-discrimination laws to academic-style seminars on the perils of implicit bias, microaggressions, or misgendering.

Advocates of this kind of training have their hearts in the right place. We are all familiar with comparisons showing that Black people earn 50% less than white peers and women earn 70 cents for every dollar that a man earns.

However, the most popular tools used to combat disparities in the workplace have produced almost no measurable results.

The average impact of corporate diversity, equity, and inclusion (DEI) training is zero and some evidence suggests that the impact can become negative if the training is mandated.

“Statistical Snapshots,” which describe how employee outcomes differ by demographic group, are another popular tool. These numbers cannot provide proof of bias. Simple averages often mislead and, importantly, crafting strategies based on misleading data often does more harm than good.

Some business leaders, in their determination to increase diversity, leap directly from observing raw disparities to removing some information from application forms, another common practice meant to make workplaces more equitable. However, hiding information on applications often leads to worse outcomes for those it was intended to help-likely because hiring managers use race itself as a proxy for the information they’re no longer allowed to see.

Our intuition for how to decrease race and gender disparities in the workplace has failed us for decades. It’s time to stop guessing and start using the scientific method. Remember when we thought that the Bubonic Plague was caused by a triple conjunction of Saturn, Jupiter, and Mars in the 40th degree of Aquarius?

Here is a three-step approach that can turn earnest intentions into good science.

Understand disparities

For decades, social scientists have shown that raw gaps in employment outcomes like hiring or wages-the type of data typically provided to C-suite executives-misstate the amount of actual bias in an organization. This data omits many factors that are key to personnel decisions, factors that often vary by group, owing to disparities in society at large. Business leaders can and should work to address inequality in their communities, but they should not mistake society-wide gaps for bias by their own employees.

One of the most important developments in the study of racial inequality has been the quantification of the importance of pre-market skills in explaining differences in labor market outcomes between Black and white workers. In 2010, using nationally representative data on thousands of individuals in their 40s, I estimated that Black men earn 39.4% less than white men and Black women earn 13.1% less than white women. Yet, accounting for one variable-educational achievement in their teenage years--reduced that difference to 10.9% (a 72% reduction) for men and revealed that Black women earn 12.7 percent more than white women, on average. Derek Neal, an economist at the University of Chicago, and William Johnson were among the first to make this point in 1996: “While our results do provide some evidence for current labor market discrimination, skills gaps play such a large role that we believe future research should focus on the obstacles Black children face in acquiring productive skill.”

Recently, I worked with a network of hospitals determined to rid their organization of gender bias. The basic facts were startling: Women earned 33% less than men when they were hired and their wages increased less than men once on the job. Yet, accounting for basic demographic variables known about individuals prior to hiring, these differences decreased by 74%. A problem remained, but it was an order of magnitude smaller than the unadjusted numbers implied.

Find the root causes of bias

Social scientists tend to categorize bias into one of three flavors: preference, information, and structural. Preference bias is good old-fashioned bigotry. If company A prefers group W over group B then they will hire and promote them more even if they are similarly qualified.

Information bias arises when employers have imperfect information about workers’ potential productivity and use observable proxies, like gender or race, to make inferences (gender stereotypes are a classic example).

Structural bias occurs when companies institute practices, formally or informally, that have a disparate impact on particular groups, even when the underlying practices are themselves group blind. Employee referral programs can fall into this category.

Over the past fifty years, economists and other social scientists have developed brilliant ways of statistically distinguishing between different types of bias. Gary Becker, in his 1993 Nobel Prize acceptance speech, outlined one such statistical procedure known as the “outcomes test.” It operates by comparing the success rate of decisions across groups and then inferring whether different decision rules were used for different groups. For example, if women CEOs statistically outperform male CEOs, all else equal, that would suggest that a higher standard was applied to women in the selection process. This type of statistical test can be used for hiring, promotions, and attrition across an organization.

In 2013, collaborators and I developed a similar test to detect information-based bias. Our approach uses the insight that if employers have information-based biases at the time of hiring, but then learn more about an employee’s productivity once they are on the job, one would expect to see the returns to tenure within the company to be higher for the group that faced the initial bias. Using a nationally representative dataset of thousands of individuals, we found that there was a significant gap at the time of hiring for Black candidates relative to white peers but that, as predicted, Black candidates experienced a 1.1 percentage point higher return to tenure.

With the aforementioned hospital network, the data pointed to a structural bias in scheduling. Women and men who worked the same number of hours earned exactly the same wage, but men worked more hours due to how the company assigned schedules, not women’s desire to work less.

This is the key step that is missing in every DEI initiative I have seen in the past 25 years: a rigorous, data-driven assessment of root causes that drives the search for effective solutions. In other aspects of life, we would not fathom prescribing a treatment without knowing the underlying cause. Hiding information on resumes when information bias is present is as effective as using alcohol baths to treat fever.

Evaluate

We must rigorously evaluate what works and what doesn't. The old cardiac test-where you “feel it in your heart”-is not good enough. Once we know where potential biases exist, determine what caused them, and curate a set of solutions to test, we must meticulously evaluate what’s working and what’s not.

Solutions that yield measurable results can be substantiated into company policy, while those that don't should be discarded. In the case of the hospital network, once a small change was made to the structure of their scheduling, gender differences were reduced. Despite countless hours spent in training and seminars, their results were unchanged for years. The solution was hidden in plain data.

This will seem heretical to some-but it barely scratches the surface of what's possible with a data-first approach to diversity, equity, and inclusion. More corporate leaders should be trying to solve diversity challenges in the same way they solve problems in every other aspect of their business: through intelligent use of data, rigorous hypothesis testing, and honest inference about what is working.

We examine theinfluence of physical proximity on between-startup knowledge spillovers at one of the largest technology co-working hubs in the United States. Relying on the random assignment of office space to the hub's 251 startups, we find that proximity positively influences knowledge spillovers as proxied by the likelihood of adopting an upstream web technology already used by a peer startup. This effect is largest for startups within close proximity of each other and quickly decays: startups more than 20 meters apart on the same floor are indistinguishable from startups on different floors. The main driver of the effect appears to be social interactions. While startups in close proximity are most likely to participate in social co-working space events together,knowledge spillovers are greatest between startups that socialize but are dissimilar.Ultimately, startups that are embedded in environments that have neither too much nor too little diversity perform better, but only if they socialize.

Maria P. Roche, Alexander Oettl and Christian Catalini, "(Co-)Working in Close Proximity: Knowledge Spillovers and Social Interactions,"National Bureau Of Economic Research, June 2022, https://www.nber.org/papers/w30120

Core Findings

“…The setting for our study is one of the largest technology co-working spaces in the United States. The building consists of five floors, covering 9,300 m2 (100,000 sq.ft.). One challenge in examining the relationship between location and knowledge spillovers is that rms and individuals may choose to locate in areas where knowledge exchange is already likely to be high. To deal with this potential endogenous location choice, we rely on the random assignment of office space to the hub's 251 startups. We measure knowledge spillovers as the instance of adopting a component of a peer rm's technology stack. Using floor plans to measure geographic distance, we find that close physical proximity greatly influences the likelihood of these knowledge spillovers. This effect, however, quickly decays with distance where startup rms that are more than 20 meters (66 feet) away are no longer influenced by each other. Strikingly, being located more than 20 meters apart, but on the same floor does not appear to differ from being located on a different floor altogether. In addition, we find that when firms overlap with common areas at the hub (e.g., kitchens), the distance of influence increases, revealing the important role that these spatial features play in extending geographic reach and in promoting knowledge spillovers..”

The Evidence

“. .Table 3 presents the results from assessing the effect of distance on the amount of peer technology adoption (ln(AdoptCountij + 1)) using a standard OLS model and using a linear probability model to estimate the likelihood of adopting a technology from a peer rm 1(AdoptTechij ). In the full model (Columns 2 and 4), using rm-x-room fixed effects and controlling for industry, business model, gender, age and pre-period technology overlap, we find that the doubling of distance between two dyads reduces both the amount of peer technology adoption by 3.5% and the likelihood of any peer technology adoption by 1.7%, with both point estimates significant at the 1% level. As seen, the magnitude and statistical significance of the effect remains largely unchanged with the inclusion of additional controls….”

New York — which is where I happen to live — isn’t just safe relative to other big cities. The data in the above chart are from the Federal Bureau of Investigation and local police departments, but the Centers for Disease Control and Prevention also keep track of homicides as reported on death certificates, and one can sort them by, among other things, whether they befell residents of a county that contains all or part of the biggest city in a large (population of one million or more) metropolitan area, a fringe (suburban) county in such a metropolitan area, or elsewhere. Suburban counties in large metro areas have tended to have the lowest homicide rates, but New York City’s was even lower from 2017 to 2019 and was only slightly higher in 2020, the most recent year for which the CDC has full data. Some causes of death other than homicide seem pretty relevant to assessing safety. Your risk of getting killed on the subway or in a subway station is much higher in New York City than in a place with no subway, for example,but if you add up all risks related to getting to work or school or wherever else it is you need to go, it’s a different story. New Yorkers are only about as third as likely to die in transportation accidents of any kind as Americans are overall. Put homicide and transportation risks together, and New York starts looking like a refuge from the American carnage.All these accountings leave out the September 2001 terrorist attacks. The most dangerous areas are a more hardscrabble lot, although led by an affluent coastal metro (Baltimore-Columbia) where the main city has an awful violent crime problem. The city of Baltimore’s homicide rate was a staggering 58.5 per 100,000 residents in 2021, twice that of Chicago and more than 10 times New York City’s.

Justin Fox, "New York City Is a Lot Safer Than Small-Town America,"Bloomberg, June 7, 2022, https://www.bloomberg.com/opinion/articles/2022-06-07/is-new-york-city-more-dangerous-than-rural-america

New York City Is a Lot Safer Than Small-Town America

Since the summer of 2020 the United States has been experiencing a deadly wave of violence, and New York City has not been immune. The number of homicides in the city shot up 47% in 2020, and rose another 4% last year. While the tally is down 9% so far this year there have been some shocking murders, not to mention a mass shooting on the subway that miraculously didn’t take any lives.

Regular consumers of the news would thus be forgiven for believing that the Big Apple has become a terribly dangerous place. And yes, it is markedly more dangerous than it was before the pandemic. I’m not here to dismiss concerns about rising crime in the city. But a little context is in order. And by “a little context,” I mean a lot of charts.

The historical context, which a fair number of people are aware of, is that New York’s pandemic murder wave followed a long decline, with the city’s homicide rate in 2021 still less than a fifth what it was in 1990. The geographical context, which seems to be less widely understood, is that while homicide rates fell sharply all over the the US in the 1990s, that decline slowed in the 2000s and reversed starting in 2015. In Philadelphia the homicide rate is now worse than it was in the early 1990s, and in Chicago it’s close. Which means there’s a growing gap between New York City and most of the rest of urban America.

Some causes of death other than homicide seem pretty relevant to assessing safety. Your risk of getting killed on the subway or in a subway station is much higher in New York City than in a place with no subway, for example, but if you add up all risks related to getting to work or school or wherever else it is you need to go, it’s a different story. New Yorkers are only about as third as likely to die in transportation accidents of any kind as Americans are overall. Put homicide and transportation risks together, and New York starts looking like a refuge from the American carnage.

There are other accidental threats listed in ICD-10 that seem less salient in the city than in the country, and in some cases, the suburbs. To name a few: “contact with powered lawnmower,” “drowning and submersion while in swimming pool,” “contact with venomous snakes and lizards,” “victim of volcanic eruption.” So I ran the numbers again for mortality from all external causes.

The explosion of drug-overdose deaths in the US over the past decade is one of the most depressing trends one encounters in CDC Wonder, but it does take us somewhat afield from what people usually have in mind when they consider whether a place is safe or dangerous. Same goes for suicides, which have also risen but not as sharply. Then there are external causes such as falls and surgical complications that are usually fatal only for the very old. If included in the calculations, they tend to make retirement communities look terribly dangerous, which doesn’t seem right.

Here’s what mortality from external causes looks like if one excludes (1) accidental poisoning (how drug overdoses are classified), (2) intentional self-harm, (3) falls, (4) complications of medical and surgical care and (5) sequelae of external causes of morbidity and mortality (after-effects, that is, which are also fatal mostly for the elderly).

This is so opposite from conventional wisdom that it’s worth teasing out why. One obvious explanation is that murder scares people a lot more than accidents do, and homicide risk is in fact highest in urban counties — just not in New York City. Another is that, as a colleague put it to me, you may be less likely to be victim of a crime in New York City, but you’re more likely to witness one. In dense urban environments you simply see a lot more of what’s going on, for good and ill. It’s also possible that the issue is not so much urban versus rural as it is who lives where. An affluent rural or exurban county where lots of people can work from home might be lower risk than a poorer urban one where most can’t. Rural counties in Appalachia or the Deep South might have very different profiles from those in New England or the Plains states.

A look at 2020 selected-external-causes mortality by state offers some support for the latter interpretation. The Northeast is the safest part of the country, but some heavily rural states (Mississippi, Arkansas) are a lot more dangerous than others (Vermont, Iowa).

But let’s not forget those data limitations. While CDC Wonder knows everything about mortality in the US, for privacy reasons it only shares what it knows if there are 10 or more deaths to report, and because of reliability concerns it only calculates mortality rates if there are 20 or more deaths. For the collection of external causes I selected, Wonder churned out 2020 death totals for 1,711 of the nation’s 3,143 counties and county equivalents (parishes in Louisiana, census areas in Alaska, independent cities in Virginia and some other states) and mortality rates for 997. It provided enough data to calculate the mortality rates for all 1,711, but I chose not to for the charts here. If I had, three more Northern Virginia cities and counties would have entered the top 10 along with Story County, Iowa, and La Crosse County, Wisconsin.

If I selected a wider array of external causes, or combined data from several years, more counties with smaller populations would enter the mix. Still, pretty much any version of the most-dangerous-counties list would be at least as heavy on rural counties as this.

Also, because complete 2020 Census results have yet to be released, the CDC is still using outdated 2020 population estimates based on the 2010 census. For New York City, this estimate is 550,977 lower than the number from the 2020 census, and the CDC estimates for most of the 2010s are probably too low as well. To give a sense of the impact, the city’s CDC-calculated homicide rate of 5.4 per 100,000 in 2020 would drop to 5.1 using the census population. I decided to stick with the CDC’s numbers nonetheless because (1) the city likely lost several hundred thousand residents in the months following the April 1 census record date, and (2) redoing all the 2020 mortality-rate calculations using census data would have introduced a lot of opportunities for error. But keep in mind that all the mortality rates for New York City in recent years in the CDC-derived charts are probably slight overestimates.

Finally, it would be nice to see this data by metropolitan area, wouldn’t it? CDC Wonder’s most up-to-date deaths databases does not provide statistics by metro area. It is possible to download them by counties, then sort those into metropolitan areas, but with 2020 deaths data from my list of external causes available for only a little over half of US counties, that doesn’t really work. By combining deaths from the five most recent available years I was able to get a more complete county list, and fill in the blanks with a little extra effort. Here, then, are the safest metropolitan areas in the US from 2016 through 2020, by my admittedly somewhat idiosyncratic definition.

The most dangerous areas are a more hardscrabble lot, although led by an affluent coastal metro (Baltimore-Columbia) where the main city has an awful violent crime problem. The city of Baltimore’s homicide rate was a staggering 58.5 per 100,000 residents in 2021, twice that of Chicago and more than 10 times New York City’s.

Data sources for charts

Safest of the Six Largest U.S. Cities: The main data sources here are the FBI’s Crime Data Explorer and Census Bureau annual population estimates that I got from the Data Commons Place Explorer (for pre-2010 numbers) and the Census Bureau (for 2010-2021). I could not find city population estimates for 1985-1989 so I calculated my own by just assuming the cities grew (or shrank) at a uniform pace between the 1980 and 1990 censuses. The FBI hasn’t released 2021 crime data yet so I got that year’s numbers either directly from police departments or from news reports based on police department data. Also, for some reason the FBI’s post-2014 homicide numbers for Philadelphia are different from (and lower than) those on the Philadelphia Police Department website, so I went with the latter.

Murder in the Cities, Suburbs and Country: This is downloaded straight from CDC Wonder’s Underlying Causes of Death database, but there are two different ways to get homicide totals from that. One is to look under “Injury Intent and Mechanism” and select “homicide” as the injury intent, the other is to look through the ICD-10 codes used to classify all causes of death and select “assault” under “External causes of morbidity and mortality.” The former includes deaths from terrorist attacks and the latter does not, because they are classified under a separate ICD-10 category of “Codes for special purposes.” When this piece was first published, the chart showed homicide deaths with the September 2001 terrorist attacks subtracted out (for consistency with the FBI data and because otherwise New York’s high 2001 rate would make the chart very hard to read); subsequently I realized it would be simpler and more consistent with the other charts to use the assault numbers, which are almost but not quite identical. Here they are by county urbanization status (I combined the two categories of nonmetro counties for the charts) and for New York City. Once you click through, you’ll have to agree to a couple of data use restrictions, then wait as CDC Wonder generates the data.

Car Crashes Change the Picture: This adds together assault plus transportation accident deaths as reported under “External causes of morbidity and mortality.” Here they are by county urbanization, and for New York City.

Safer in the Big City: This is all “External causes of morbidity and mortality,” by county urbanization, and for New York City.

Still Safer in the Big City: This is external causes minus the categories discussed in the text and the chart footnotes, by county urbanization and for New York City.

Safe States and Dangerous Ones: Selected external causes by state for 2020.

The Safest Counties in America, Sort Of, and The Most Dangerous Counties in America, Sort Of: Selected external causes by county for 2020.

In the aftermath of the Covid-19 pandemic, the U.S. equity markets have witnessed a surge in the number of publicly listed companies. Using data for the three major U.S. stock exchanges (AMEX, NYSE, and NASDAQ), we find that the number of publicly traded companies went from 4,144 at the end of August 2020 to 5,301 at the end of December 2021, a staggering increase of about 28 percent.Figure 1 reports the trend in the number of public companies since 1973, two years after the NASDAQ started its operations. The number of publicly traded firms peaked in 1997, reaching 7,842 at the end of November, and has since steadily declined through 2012, after which it remained nearly flat until about the first half of 2020. Given the importance of the SPAC phenomenon, we manually classify the newly listed companies into SPACs and non-SPACs, that is, companies that go public with a traditional IPO process, and report in Figure 3 the number of newly listed firms in each category. During the period from the end of August 2020 to December 2021, there were 736 newly listed firms classified as SPACs and 646 newly listed non-SPAC firms; that is, SPACs accounted for about 53% of total newly listed firms.The number of newly listed SPACs took off in the second half of 2020 and peaked in the months of March and April 2021 We document that the entry of SPACs has significantly modified the composition of U.S. equity markets over the period from August 2020 to the end of 2021. First, the fraction of small size stocks classified as SPACs had increased from about 8% to about 28%. Second, the fraction of ordinary common shares of foreign-incorporated companies has jumped from about 14 percent to about 18 percent. Third, at the end of 2021, companies in the Other Financial Investment Activities industry (4-digit NAICS code 5239) represented 15 percent of listed companies, up from about 4 percent in August 2020.

Christine Dobridge, Rebecca John and Berardino Palazzo, "The post-COVID stock listing boom,"Federal Reserve Board, June 17, 2022, https://www.federalreserve.gov/econres/notes/feds-notes/the-post-covid-stock-listing-boom-20220617.htm

Introduction

In the aftermath of the Covid-19 pandemic, the U.S. equity markets have witnessed a surge in the number of publicly listed companies. Using data for the three major U.S. stock exchanges (AMEX, NYSE, and NASDAQ), we find that the number of publicly traded companies went from 4,144 at the end of August 2020 to 5,301 at the end of December 2021, a staggering increase of about 28 percent.1

Figure 1 reports the trend in the number of public companies since 1973, two years after the NASDAQ started its operations. The number of publicly traded firms peaked in 1997, reaching 7,842 at the end of November, and has since steadily declined through 2012, after which it remained nearly flat until about the first half of 2020.2

Figure 2 reports the 12-month moving average of the monthly growth rate for the NASDAQ and NYSE+AMEX, respectively. As the figure shows, the recent increase in the number of companies is a phenomenon that is equally important for firms traded in the NASDAQ and firms traded in the NYSE and AMEX platforms.

What is responsible for the recent listing boom?

The obvious answer to the above question is the huge popularity of investing in Special Purpose Acquisition Companies (SPACs). Starting in the second half of 2020, U.S. equity markets have witnessed an explosion in the number of so-called blank check companies' (or SPACs') stocks. These companies are effectively holding cash to acquire private companies and take them public, bypassing the traditional IPO process.3

Given the importance of the SPAC phenomenon, we manually classify the newly listed companies into SPACs and non-SPACs, that is, companies that go public with a traditional IPO process, and report in Figure 3 the number of newly listed firms in each category.4 During the period from the end of August 2020 to December 2021, there were 736 newly listed firms classified as SPACs and 646 newly listed non-SPAC firms; that is, SPACs accounted for about 53% of total newly listed firms. The number of newly listed SPACs took off in the second half of 2020 and peaked in the months of March and April 2021.5

Newly listed firms during the post-Covid listing boom: Some stylized facts

In what follows, we compare the characteristics of the about 1,400 newly publicly listed companies that entered the U.S. equity markets during the SPAC listing boom, between the end of August 2020 and December 2021, with the approximately same number of newly publicly listed companies that entered over the period December 2014 to August 2020.

Table 1 shows that companies that entered during the SPAC listing boom tended to have the same size distribution of the companies that had entered since the end of 2014; that is, about 30% of newly listed firms are small in terms of their market capitalization.7 However, Table 1 highlights that there is a significant difference in the size distribution of SPACs versus non-SPACs, with the former tending to be significantly smaller: out of 10 newly listed SPACs, 9 belong to the small size category.

Table 4 also highlights that SPAC stocks are more volatile after the announcement (but before De-SPAC). Figure 4 helps us understand why this is the case. We consider only the subset of SPACs that entered between the end of August 2020 and December 2021, have become an operating company, and depict some of the quantities reported in Table 4 around the period that goes from 60 days prior to 180 days after the day of the deal announcement. The shaded area is the interdecile range, the solid blue line is the median daily return, and the red dashed line is the fraction of SPACs that have become an operating company.

Concluding remarks

The number of publicly listed firms in the three major U.S. stock exchanges increased by about 1,100 in the period August 2020 to December 2021. This surge has been driven both by an unprecedented increase in the number of SPAC listings and by a large inflow of more traditional companies, thus suggesting that the post-COVID financial market environment has been particularly favorable for new IPOs amid low interest rates and ample market liquidity.

We document that the entry of SPACs has significantly modified the composition of U.S. equity markets over the period from August 2020 to the end of 2021. First, the fraction of small size stocks classified as SPACs had increased from about 8% to about 28%. Second, the fraction of ordinary common shares of foreign-incorporated companies has jumped from about 14 percent to about 18 percent. Third, at the end of 2021, companies in the Other Financial Investment Activities industry (4-digit NAICS code 5239) represented 15 percent of listed companies, up from about 4 percent in August 2020.

These compositional changes are very likely to be temporary since most SPACs are required to enter into a merger usually within 2 years from the IPO (or the SPAC must return shareholders' investments). As a consequence, we expect the fraction of firms in the Other Financial Investment Activities industry sector to decline as SPACs merge with companies operating in different industries. At the same time, we expect the fraction of ordinary common shares of foreign-incorporated companies to decline as more SPACs will become operating companies incorporated in the United States.

The fraction of SPACs in the small size stock category will also decrease. However, in the near future, having a large fraction of SPACs in small size portfolios might have implications for cross-sectional asset pricing given the peculiar properties of SPACs' returns before these investment vehicles become operating companies.9

References

Bulley, David and Dean Bennet (2020)," The use of Cayman Islands Structures for SPAC IPOS," webpage:https://www.applebyglobal.com/publications/the-use-of-cayman-islands-structures-for-spac-ipos/.

Doidge, Craig G., Andrew Karolyi, and René M. Stulz (2017), "The U.S. listing gap," Journal of Financial Economicshttps://www.sciencedirect.com/journal/journal-of-financial-economics, vol. 123 (3),pp. 464-487.https://doi.org/10.1016/j.jfineco.2016.12.002

Doidge, Craig G., Andrew Karolyi, and René M. Stulz (2004), "New lists: Fundamentals and survival rates," Journal of Financial Economicshttps://www.sciencedirect.com/journal/journal-of-financial-economics, vol. 73 (2),pp. 229-269.https://doi.org/10.1016/j.jfineco.2003.04.001

Gahng, Minmo, Jay R. Ritter, and Donghang Zhang (2021), "SPACs." Available at SSRN:https://ssrn.com/abstract=3775847.

________________________________

1. We measure the number of traded firms as the number of unique PERMCOs in the CRSP database with a non-missing value for price and number of shares outstanding. We only consider ordinary common shares of companies incorporated in the U.S. and outside the U.S. (share codes in CRSP equal to 10, 11, and 12) and traded in the three major U.S. stock exchanges: NYSE, AMEX, and NASDAQ (exchange codes in CRSP equal to 1, 2, and 3, respectively). If a SPAC changes its PERMCO after an acquisition, we do not consider the new PERMCO to be a new traded entity. Return to text

2. Fama and French (2004) suggest that the increase in the number of publicly listed firms in the period 1980-2001 is due to a decline in the cost of equity that has prompted smaller and more financially fragile firms to go public. Doidge, Karolyi, and Stulz (2017) calculate that about 46% of the decline in the number of publicly traded firms from 1996 through 2012 was due to delisting, while the remaining fraction was due to a lower listing rate. Return to text

3. Santilli, Peter and Amrith Ramkumar (2021), "SPACs Are the Stock Market's Hottest Trend. Here's How They Work," The Wall Street Journal, March 29, https://www.wsj.com/articles/spacs-are-the-stock-markets-hottest-trend-heres-how-they-work-11617010202. With the total number of publicly listed firms increasing by 1,157 from the end of August 2020 to the end of December 2021 and 1,382 new entrants over the period, we estimate that 225 firms delisted over the period. Return to text

4. We used the NASDAQ IPO calendar to manually classify newly listed firms into SPACs and non-SPACs (https://www.nasdaq.com/market-activity/ipos). We also use data from the website spacresearch.com (https://www.spacresearch.com/) to identify SPACs that have finalized an acquisition by the end of February 2022. In this note, we focus on the first time a SPAC appears on one of the three major U.S. stock exchanges. Return to text

5. Gahng, Ritter, and Zhang (2021) provide one of the first analysis of SPACs in U.S. equity markets. Their sample of SPACs ends in December 2019, prior to the recent SPAC listing boom. Return to text

6. Specifically, the number of non-SPAC firms increased from 3,952 in August 2020 to 4,377 in December 2021, an increase of about 11 percent. Return to text

7. We measure size using market capitalization at the end of the first month of trade. Small firms belong to the bottom quintile of the market capitalization distribution. Market capitalization quintiles are calculated each month using only companies traded on the New York Stock Exchange. Return to text

8. Many SPACs choose to incorporate in the Cayman Islands to take advantage of their more favorable statutory merger regime (Bulley and Bennet, 2020). Return to text

9. Excluding SPACs from portfolio formation before they become an operating company could be a potential solution to mitigate their effect on the return properties of small size portfolios. Return to text

The old wisdom that the more women pursue a career, the fewer children they bear, has long been shown to be false. In the 1980s, it was true that the higher a country’s female labour force participation rate, the lower its fertility rate. But a decade later, the relationship had flipped: the higher a country’s female labour participation rate, the higher its fertility rate. In some countries, the probability of having a second child is now highest among the professional classes and lowest among the lowest-skilled. In societies where male contributions are minimal, large numbers of women say they do not want another child despite their partner’s wishes. The result is that in countries where men do a third of childcare and housework, total fertility rates are roughly a third higher than in those where they do just a fifth.In the US, between May and December 2020, fathers spent 50 minutes per day in which their primary activity was childcare. During the first year of the pandemic this fell to 46 minutes, while mothers’ daily childcare commitments rose from 88 to 101 minutes. And the “new normal” beginning to take shape has seen US mothers’ labour force participation rates fall, reversing decades-long trends. The participation rate for fathers of young children dipped from 95 to 93.4 per cent in 2020, before recovering slightly to 93.9 last year. Mothers, meanwhile, have fallen from 66.4 to 65.8 per cent in 2020, labour participation rates slid further to 65.6 per cent last year, while in the UK female participation has also dipped slightly.

Change is needed on multiple fronts. Government and company policies must offer more flexibility for all parents. Britons need far more help meeting the country’s world-leading (in a bad way) childcare costs. And men must take their own step back from the office and join their partners in pulling their weight at home. Everyone must shoulder the burden of raising the next generation.

And the “new normal” beginning to take shape has seen US mothers’ labour force participation rates fall, reversing decades-long trends. The participation rate for fathers of young children dipped from 95 to 93.4 per cent in 2020, before recovering slightly to 93.9 last year. Mothers, meanwhile, have fallen from 66.4 to 65.8 per cent in 2020, labour participation rates slid further to 65.6 per cent last year, while in the UK female participation has also dipped slightly.

What’s worrying is that in parts of the west, men have not just been slow to act on this, but the trend has been in the opposite direction. In the US, between May and December 2020, fathers spent 50 minutes per day in which their primary activity was childcare. During the first year of the pandemic this fell to 46 minutes, while mothers’ daily childcare commitments rose from 88 to 101 minutes.

For decades, the assumption has been that when it comes to the workplace, out of sight is out of pocket, but the pandemic has shown otherwise. The new era of flexible working makes the equalisation of parental childcare more feasible than ever.

John Burn-Murdoch, "Men stepping up at home is key to boosting birth rates,"Financial Times, June 16, 2022, https://www.ft.com/content/6a6a7695-4756-42d1-b03d-283ad5e57d5f

Men stepping up at home is key to boosting birth rates

When we talk about how to address falling birth rates, it’s not surprising that the conversation usually centres on young women. After all, it’s quite tricky to influence birth rates without them. But according to a growing body of research, this demographic is already doing about as much as they can. A more fruitful policy target may in fact be men.

In the year 2000, there were 22.5 people aged 65 and older for every 100 people of working age in OECD countries. This year, the ratio is estimated to have climbed by almost half to 32.4. Over the same period, the female labour force participation rate climbed steadily from 59 to 65 per cent, and the gender pay gap for workers in OECD countries shrunk from 18 per cent to just under 12 per cent.

The coinciding trends present a conundrum: at a time when some governments are fretting about declining birth rates, the number of women navigating the fraught trade-off between flourishing professionally and having children has never been larger.

A comprehensive new study suggests it is time we stop defaulting to the assumption that job vs family is a trade-off solely for women, and start paying more attention to the role men can play in facilitating simultaneous advances in both birth rates and female employment.

The old wisdom that the more women pursue a career, the fewer children they bear, has long been shown to be false. In the 1980s, it was true that the higher a country’s female labour force participation rate, the lower its fertility rate. But a decade later, the relationship had flipped: the higher a country’s female labour participation rate, the higher its fertility rate. In some countries, the probability of having a second child is now highest among the professional classes and lowest among the lowest-skilled.

Several factors have contributed to this reversal, including improvements in state provision of and support for childcare, but among the biggest shifts has been the increase in paternal contribution to childcare and housework. This makes the trade-off for mothers much less steep.

Having a child is usually dependent on both partners wanting one. Research shows that agreement tends to be much higher where men do relatively high shares of childcare. In societies where male contributions are minimal, large numbers of women say they do not want another child despite their partner’s wishes. The result is that in countries where men do a third of childcare and housework, total fertility rates are roughly a third higher than in those where they do just a fifth.

Does screening applicants using exams help or hurt the chances of lower-SES candidates? Because individuals from lower socioeconomic backgrounds fare, on average, worse than those from richer backgrounds in standardized tests, a common concern with this "meritocratic" approach is that it might have a negative impact on the opportunities of lower-SES individuals. However, an alternative view is that, even if such applicants underperformed on exams, other (potentially more discretionary and less impersonal) selection criteria might put them at an even worse disadvantage. We investigate this question using evidence from the 1883 Pendleton Act, a landmark reform in American history which introduced competitive exams to select certain federal employees. Using newly assembled data on the socioeconomic backgrounds of government employees and a difference-in-differences strategy, we find that, although the reform increased the representation of "educated outsiders" (individuals with high education but limited connections), it reduced the share of lower-SES individuals. This decline was driven by a higher representation of the middle class, with little change in the representation of upper-class applicants. The drop in the representation of lower-SES workers was stronger among applicants from states with more unequal access to schooling as well as in offices that relied more heavily on connections prior to the reform. These findings suggest that, although using exams could help select more qualified candidates, these improvements can come with the cost of increased elitism.

The Pendleton Act

“…The act established that employees in certain “classified” positions would need to be selected through open, competitive and anonymously graded exams (United States Civil Service Commission, 1883). The act divided the classified (that is, subject to exams) civil service into three branches: the “classified departmental service” for employees in the executive departments in DC, the “classified Customs Service” for Customs Service employees, and the “classified Postal Service” for postal workers. Figure 1 shows the total number of workers in the Executive Departments in DC, as well as the share who worked in positions that would be covered by exams after 1883. Here, we define a position as “covered by exams” if it became part of the classified service at any point between 1883 and 1893. The total number of employees grew in the decade prior to the Pendleton Act, reflecting the expansion of government functions (Libecap & Johnson, 2007). Note that, since our analysis includes office fixed effects, offices created after the reform do not contribute to the identification of the effects of exams.17 Growth in the number of employees seems to level off after 1883, both in exempted and non-exempted positions (as reflected by the stability of the share of workers in exempted positions, which fluctuated around 60% throughout the period)….”

Measuring Employees’ Social and Professional Backgrounds

“…We collected information on employees’ socioeconomic backgrounds by linking the Registers to US population censuses, using name-based matching techniques (Abramitzky et al., 2019). Specifically, we used workers’ names, birthplaces and approximate ages to link each of the 1871-1893 Registers to the 1850, 1860, 1870 and 1880 censuses. Through this procedure, we obtained information about: (1) employees’ family backgrounds, including characteristics such as race, parental wealth and parental occupations; and (2) employees’ occupations prior to joining the federal government. We provide further details on the linking strategy and sensitivity checks in Online Appendix Section A. However, we note here that: (1) employees hired through exams are not more (or less) likely to be matched to the census (Table A1), (2) the results that do not require the linked data are very similar when estimated in this linked sample (Table A2), and (3) the results are similar when we reweight the sample to account for differences in the matching probability across different individuals (Table A3)…”

Core finding

“…Our identification strategy exploits the fact that not all federal positions were initially subject to exams. Specifically, among positions in the Executive Departments in DC, the reform exempted those at the bottom (such as laborers) and at the top (such as bureau chiefs) of the state hierarchy. We use this feature of the reform to estimate a difference-in-differences model, comparing the characteristics of employees hired before and after the reform, in exempted and non-exempted positions. In other words, we ask if individuals hired to do the same job in the same office were of a different social background when hired through exams rather than through patronage. We find that the reform led to an immediate decline in the share of employees from disadvantaged backgrounds, which then persisted for at least 10 years after its implementation. First, employees hired through exams came from families that were 6.5 percentile ranks higher in the national wealth distribution. This increase was driven by a reduced representation of applicants with parents at the bottom of the distribution, together with an increase in the share of workers from middle- and upper-middle class families. Second, the reform increased the share of employees with higher-status parental occupations: we find a 5 percentage points increase in the proportion of children of professionals (a nearly 50% increase), together with a similar decrease in the pro-portion of children of blue-collar workers. Finally, the reform led to a 4 percentage points decline in the share of foreign-born employees, as well as a 7 percentage points decline in the share of employees with foreign-born parents. Interestingly, this increased elitism occurred despite the exam was based on content that should have in principle been accessible for applicants with only a “common school” education. Why did the introduction of exams improve the prospects of middle-class applicants (at the expense of applicants from disadvantaged backgrounds)? Our interpretation is that, by increasing the relative weight of formal education in hiring decisions, the reform helped “educated outsiders”: individuals with high levels of education but low levels of connections. As middle-class applicants were overrepresented in this group, the reform increased their chances of obtaining a government job….”

The Evidence

“…We present three pieces of evidence that support this interpretation. First, we show that the reform indeed increased the representation of “educated outsiders”. In particular, employees hired through exams were 10 percentage points more likely to have held a professional occupation - such as lawyers and accountants- prior to joining government than patronage employees.7 Exam based hires were also more likely to lack the personal and political connections that would have facilitated access to patronage jobs: they were less likely to have a father who was himself a bureaucrat, less likely to have grown up in DC, and less likely to come from a county in which a majority of voters supported the incumbent party (suggesting a decline in political favoritism). Second, we document that middle-class individuals were likely overrepresented among the “educated outsiders": Although middle-class applicants were more educated than those from poorer backgrounds, in the pre-reform period they represented a similarly small fraction of workers in the federal positions that would become subject to exams. Moreover, after the reform, the representation of middle-class individuals in these positions became closer to their representation in comparable private sector white-collar jobs. Third, we show that exams had the most negative effects on the chances of applicants from disadvantaged backgrounds when applicants came from states with high levels of inequality in access to schooling- presumably the places in which the children of the poor were the least represented among “educated outsiders”. We discuss two additional explanations for the reduced representation of applicants from disadvantaged backgrounds. First, the reform mandated exams to be held throughout the country, not just in Washington, DC. This expanded geographical coverage could have per se affected the selection of candidates. However, the increase in employees’ social status remains even when we restrict the comparison to employees who grew up in the same location. Second, exam-based recruitment might have gradually changed the appeal of government jobs, thus changing the pool of individuals interested in such positions. Although this effect is inherent to any move towards “meritocracy” (and the combined effect of changes in screening and changes in the applicant pool is still policy-relevant), we find evidence that such change in preferences is unlikely to drive of our findings: we observe a rapid change in bureaucrats’ social backgrounds, which seems inconsistent with plausibly slower to change perceptions about the prestige of public employment….”

“…Employees hired through examinations had higher average parental wealth than those hired through patronage. Such increases could be compatible with increases in the representation of the middle class at the expense of the children of the poor, or with increases in the representation of the upper class at the expense of the middle (or by some combination of the two). To investigate which groups increased and which groups decreased their representation as a result of the reform, we split individuals based on the wealth quintile of their parents. Panels (a) and (b) in Figure 4 show, for employees in positions that became subject to exams after 1883, their distribution across family wealth quintiles in the pre- and post-reform periods. Panels (c) and (d) show the same distribution but for employees in positions that did not become subject to exams. Among workers in positions subject to exams, workers from the top quintile were overrepresented prior to the reform. Specifically, individuals from the top 20% accounted for about 35% of workers in such positions. However, there were no large differences in the relative representation of individuals from the bottom four quintiles: each of these groups accounted for about 15% of workers prior to the reform. After the reform, in contrast, we observe a sharp increase (from 15 to 25%) in the proportion of workers from the 60-80 quintile. This increase seems to come mostly at the expense of workers in the bottom two quintiles, with more limited changes in the share of workers from the top quintile. Among those in positions exempted from exams, both the top and the bottom family wealth quintiles were overrepresented in the pre-reform period (Panel (c)). This bimodal distribution likely reflects the fact that exempted positions included highly-paid leadership positions (such as bureau chiefs) as well as low-paid positions (such as laborers). However, this overrepresentation of both the bottom and the top of the parental wealth distribution appears to continue in the postreform period (Panel (d)). Panel (e) in Figure 4 confirms this pattern when we run separate regressions in which the dependent variables are indicators for belonging to the different quintiles of the parental wealth distribution. First, we find no change in the likelihood that an employee would belong to the top 20%. Second, we find an increased representation of the middle and upper middle class (that is, families between the 40 and the 80 percentiles of the wealth distribution), which comes at the expense of a reduced representation of families in the bottom two quintiles (particularly the bottom 20%)….”

“…Summary of Results. Our findings indicate that workers appointed through exams were of a higher social class than those appointed through patronage: they had higher levels of parental wealth, were more likely to be the children of professionals, and were less likely to be the children of immigrants (or immigrants themselves). This increase in social status occurred immediately after the passing of the reform and persisted at least 10 years after its implementation..”

Diana Moreira and Santiago Pérez, "Who Benefits from Meritocracy?,"National Bureau Of Economic Research, June 2022, https://www.nber.org/papers/w30113

"... An audit released on June 16 of the Illinois Department of Employment Security found that half of federal Pandemic Unemployment Assistance (PUA) benefits the agency paid out during the height of the pandemic were stolen. The report by Illinois’ auditor general states that the Illinois agency, which administers both state and federal unemployment benefits, paid $3.6 billion in federal PUA benefits during state fiscal year 2021. The report describes “overpayments associated with ID theft and traditional fraud within the PUA program” as “unprecedented.” Specifically, the audit states that ID theft overpayments under the PUA program totaled over $1.8 billion—or 50.4 percent of the $3.6 billion in PUA payments...."

Matt Weidinger, "Audit Finds Half of Pandemic Unemployment Assistance Benefits in Illinois Were Stolen,"American Enterprise Institute, June 23, 2022, https://www.aei.org/poverty-studies/audit-finds-half-of-pandemic-unemployment-assistance-benefits-in-illinois-were-stolen/

Audit Finds Half of Pandemic Unemployment Assistance Benefits in Illinois Were Stolen

An audit released on June 16 of the Illinois Department of Employment Security found that half of federal Pandemic Unemployment Assistance (PUA) benefits the agency paid out during the height of the pandemic were stolen.

The report by Illinois’ auditor general states that the Illinois agency, which administers both state and federal unemployment benefits, paid $3.6 billion in federal PUA benefits during state fiscal year 2021. The report describes “overpayments associated with ID theft and traditional fraud within the PUA program” as “unprecedented.” Specifically, the audit states that ID theft overpayments under the PUA program totaled over $1.8 billion—or 50.4 percent of the $3.6 billion in PUA payments.

Federal officials regularly confirm that the temporary and now-expired PUA program was uniquely subject to fraud and abuse. US Department of Labor Inspector General (IG) Larry Turner noted in his March 2022 Congressional testimony that PUA “posed the greatest risk to the UI system,” continuing warnings the IG issued about the program starting in May 2020, just weeks after its creation. Turner stated the elevated fraud risk was driven by “PUA’s expanded coverage” to groups not traditionally eligible for unemployment benefits, as well as the program’s allowing claimants to “self-certify their eligibility” for benefits. Earlier this month, the nonpartisan Government Accountability Office similarly raised concerns over the self-certification process used by PUA, while arguing the unemployment benefits system overall remains at “high risk” for fraud and abuse.

As state reports like this start to emerge, official national estimates of PUA misspending remain unavailable. Turner’s testimony qualified his estimate of a shocking $163 billion in misspending across all pandemic unemployment benefit programs by adding that the real total is “likely higher” since that figure does not include elevated abuse under the PUA program. Counting such elevated abuse, outside experts have estimated pandemic unemployment benefit losses to fraud could total as high as $400 billion. For comparison, all benefits paid out under the nation’s unemployment benefits system in calendar year 2019 totaled under $30 billion.

News accounts suggest the Illinois Department of Employment Security has blamed “insufficient and flawed federal guidance” as well as “a hastily constructed program put together by the Trump administration” for their extraordinary PUA misspending. They are right about the “hastily constructed” part, but that ignores the reality that the PUA program was designed by senior Democratic lawmakers. On March 12, 2020, current Senate Majority Leader Chuck Schumer (D-NY) and Finance Committee Chairman Ron Wyden (D-OR) introduced the “Pandemic Unemployment Assistance Act,” which proposed creating the unprecedented PUA program. That legislation was subsequently included in the massive CARES Act President Donald Trump signed into law two weeks later on March 27, 2020. Upon Senate passage, Schumer stated he “conceived” of the unemployment benefit expansions in the new law, which he dubbed “unemployment insurance on steroids.” Those expansions included the unprecedented benefits for “the self-employed and workers in the gig economy” offered by the PUA program.

A review of the legislative text reveals that the Schumer/Wyden bill is almost identical to the eventual CARES Act PUA policy. Indeed, the only major difference involves the proposed duration of the PUA program. Under the CARES Act, PUA was originally authorized through December 26, 2020, and subsequent extensions continued the temporary program until it expired nationwide on Labor Day 2021. Under the Schumer/Wyden bill, PUA would have operated until six months after the COVID-19 public health emergency expired; the most recent renewal of that determination was issued in April 2022. That means that, under the Schumer/Wyden bill, PUA—and presumably its extraordinary losses to fraud—would still be continuing even today.

I quantify and track the impact of supply- and demand-related factors on personal consumption expenditures (PCE) inflation. I assess inflation rates by spending category. I divide categories in the PCE basket into supply- and demand-driven groups. Demand-driven categories are identified as those where an unexpected change in price moves in the same direction as the unexpected change in quantity in a given month; supply-driven categories are identified as those where unexpected changes in price and quantity move in opposite directions. This methodology accounts for the evolving impact of supply- versus demand-driven factors on inflation from month to month. My analysis highlights that both supply and demand factors are responsible for current elevated inflation levels. Supply factors explain about half of the difference between current 12-month PCE inflation and pre-pandemic inflation levels, and the effects appear to be rising more recently. Demand factors are responsible for about a third of the difference, and those effects appear to be diminishing more recently. The remainder is due to factors that cannot be definitively labeled as supply or demand.The large impact of supply factors implies that inflationary pressures will not completely subside until labor shortages, production constraints, and shipping delays are resolved. Although supply disruptions are widely expected to ease this year, this outcome is highly uncertain. Figure 1 shows the contributions to 12-month headline PCE inflation from supply- (green) and demand-driven (blue) inflation between 1998 and 2022. Yellow sections show the ambiguous portion of PCE inflation that is not labeled as either demand or supply driven. I calculate the monthly contributions to inflation by multiplying the change in price for a given category by its most recent respective weight in the PCE index. The monthly contributions for demand-driven, supply-driven, and ambiguous categories are then calculated as the sum of contributions from all items within that category. The contributions for demand-driven and supply-driven inflation are then calculated as the 12-month trailing sum of their respective monthly contributions. This generates a series that can be directly compared to the year-over-year PCE inflation rate. Figure 2 compares the most recent contributions of supply-driven inflation in panel A and demand-driven inflation in panel B against their respective 2010-2019 averages.Supply-driven inflation is currently contributing 2.5 percentage points (pp) more than its pre-pandemic average, while demand-driven inflation is currently contributing 1.4pp more. Thus, supply-driven inflation explains a little more than half of the 4.8pp gap between current levels of year-over-year PCE inflation and its pre-pandemic average level.Demand factors explain a smaller share of elevated inflation levels, accounting for about one-third of the difference. The ambiguous category, which is not shown, explains the remainder of the difference.Figure 3 breaks down these contributions over past five years, showing the one-month changes used to construct the 12-month trailing sums shown in Figure 1.

Adam Hale Shapiro, "How Much Do Supply and Demand Drive Inflation?,"Federal Reserve Board Of San Francisco, June 21, 2022, https://www.frbsf.org/economic-research/publications/economic-letter/2022/june/how-much-do-supply-and-demand-drive-inflation/

How Much Do Supply and Demand Drive Inflation?

Inflation declined rapidly at the onset of the pandemic in the spring of 2020 before taking a dramatic turn upward in early 2021, rising to levels that remain well above the Federal Reserve’s longer-run goal of 2% on average. Researchers and policymakers have pointed to both supply and demand factors as being responsible for elevated inflation. For instance, Barnichon and Shapiro (2022) showed the impact of supply-related factors such as labor shortages, while Barnichon, Oliveira, and Shapiro (2021) and Jordà et al. (2022) demonstrated the importance of heightened demand stemming from pandemic-related fiscal relief. The extent to which either supply or demand factors are responsible for higher inflation levels has important implications for monetary policy. As Fed Chair Powell stated in a recent interview, “What can control is demand, we can’t really affect supply with our policies…so the question whether we can execute a soft landing or not, it may actually depend on factors that we don’t control” (Marketplace 2022).

In this Economic Letter, I quantify and track the impact of supply- and demand-related factors on personal consumption expenditures (PCE) inflation. Similar to the methodology introduced in Mahedy and Shapiro (2017), and outlined in Shapiro (2022), I assess inflation rates by spending category. I divide categories in the PCE basket into supply- and demand-driven groups. Demand-driven categories are identified as those where an unexpected change in price moves in the same direction as the unexpected change in quantity in a given month; supply-driven categories are identified as those where unexpected changes in price and quantity move in opposite directions. This methodology accounts for the evolving impact of supply- versus demand-driven factors on inflation from month to month. To help monitor these changes, the San Francisco Fed has launched a new Supply- and Demand-Driven PCE Inflation data page with monthly data updates.

My analysis highlights that both supply and demand factors are responsible for current elevated inflation levels. Supply factors explain about half of the difference between current 12-month PCE inflation and pre-pandemic inflation levels, and the effects appear to be rising more recently. Demand factors are responsible for about a third of the difference, and those effects appear to be diminishing more recently. The remainder is due to factors that cannot be definitively labeled as supply or demand. The large impact of supply factors implies that inflationary pressures will not completely subside until labor shortages, production constraints, and shipping delays are resolved. Although supply disruptions are widely expected to ease this year, this outcome is highly uncertain.

Separating supply and demand drivers

This analysis uses the more than 100 goods and services categories in the PCE index. For each month of data, I separate the categories where prices moved due to a change in demand from those where prices moved due to a change in supply. To do so, I rely on simple microeconomic theory: Shifts in demand move both prices and quantities in the same direction along the upward-sloping supply curve, meaning prices rise as demand increases. Shifts in supply move prices and quantities in opposite directions along the downward-sloping demand curve, meaning prices rise when supplies decline.

Each month I estimate the changes in the price level and quantity level of each category. It is important to isolate the unexpected components of the changes in prices and quantities, as opposed to the simple change itself. This is because prices and quantities generally continue an existing trend. These expected trend components are not likely to represent a shift in demand or supply, but instead reflect longer-run factors such as technological improvements, cost-of-living adjustments to wages, or demographic changes like population aging.

Extracting the unexpected components of the monthly changes in price and quantity for each category is an iterative process. I run 10-year-window rolling regressions for both price and quantity. For example, the first window begins in January 1988, the first period PCE data are available at the detailed level, and ends in December 1997. This generates predicted values for price and quantity in January 1998, which I then compare to the actual values of price and quantity in that same month. If the actual values of price and quantity are both above or both below their predicted values, the category is labeled as “demand-driven” in January 1998. If the difference between the actual and predicted values are of opposite signs, the category is labeled as “supply-driven” in that month. If either of the actual values is close enough to its predicted value that the difference is statistically indistinguishable from zero, the category is labeled as “ambiguous” in that month. I then roll the data window forward one month and repeat the process. I iterate this process for each month until I reach the last window of data, which for this Letter begins in May 2012 and ends in April 2022.

Categories that experience frequent supply-driven price changes include food and household products such as dishes, linens, and household paper items. Categories that experience frequent demand-driven price changes include motor vehicle-related products, used cars, and electricity. However, assessing the past two years of data reveals some changes in which categories experienced demand and supply shocks in the post-pandemic period. Specifically, categories with extraordinarily frequent supply-driven price changes in 2021 and 2022 include products with known supply constraints during the pandemic, such as new automobiles, fuel, and repair services. Categories with extraordinarily frequent demand-driven price changes during this period include many goods consumed at home—for example, furniture, clothing, toys, video equipment, and cookware—as well as services related to reopening from pandemic-related closures, such as restaurants and museums.

Figure 1 shows the contributions to 12-month headline PCE inflation from supply- (green) and demand-driven (blue) inflation between 1998 and 2022. Yellow sections show the ambiguous portion of PCE inflation that is not labeled as either demand or supply driven. I calculate the monthly contributions to inflation by multiplying the change in price for a given category by its most recent respective weight in the PCE index. The monthly contributions for demand-driven, supply-driven, and ambiguous categories are then calculated as the sum of contributions from all items within that category. The contributions for demand-driven and supply-driven inflation are then calculated as the 12-month trailing sum of their respective monthly contributions. This generates a series that can be directly compared to the year-over-year PCE inflation rate.

Inflation drivers over the pandemic period

To get a sense of whether supply or demand factors are responsible for current elevated inflation levels, I compare current supply- and demand-driven inflation to their average levels from the 10 years before the pandemic. During the pre-pandemic period, PCE inflation averaged 1.5%, considerably below the April 2022 rate of 6.3%. Figure 2 compares the most recent contributions of supply-driven inflation in panel A and demand-driven inflation in panel B against their respective 2010-2019 averages. Supply-driven inflation is currently contributing 2.5 percentage points (pp) more than its pre-pandemic average, while demand-driven inflation is currently contributing 1.4pp more. Thus, supply-driven inflation explains a little more than half of the 4.8pp gap between current levels of year-over-year PCE inflation and its pre-pandemic average level. Demand factors explain a smaller share of elevated inflation levels, accounting for about one-third of the difference. The ambiguous category, which is not shown, explains the remainder of the difference.

Finally, I assess the contributions to annualized monthly changes in headline PCE inflation. This provides a higher frequency depiction of how supply and demand factors affected inflation over the pandemic period. Figure 3 breaks down these contributions over past five years, showing the one-month changes used to construct the 12-month trailing sums shown in Figure 1.

Meanwhile, supply factors began to arise in April 2021, indicating a slightly delayed response from the economy reopening. Supply-driven inflation has remained elevated since then and has accelerated more recently. This acceleration is attributable to food and energy supply disruptions, including those associated with the invasion of Ukraine.

Conclusion

Analysis in this Letter shows that supply factors are responsible for more than half of the current elevated level of 12-month PCE inflation. This in part reflects supply constraints from continued labor shortages and global supply disruptions related to the pandemic and the war in Ukraine. While demand factors played a large role in the spring of 2021, they explain only about a third of recent elevated inflation levels. Factors that cannot be labeled as either demand or supply are also playing a nontrivial role.

These results showing that factors other than demand account for about two-thirds of recent elevated inflation highlight some risks for the economy. Because supply shocks raise prices and suppress economic activity, the prevalence of supply-related factors raises the risk of entering a period of low growth and elevated inflation levels. This risk depends crucially on how long labor shortages and global supply disruptions persist. While supply disruptions are widely expected to ease this year, this outcome is highly uncertain.

References

Jordà, Òscar, Celeste Liu, Fernanda Nechio, and Fabian Rivera-Reyes. 2022. “Why Is U.S. Inflation Higher than in Other Countries?” FRBSF Economic Letter 2022-07 (March 28).

Barnichon, Regis, Luiz E. Oliveira, and Adam Hale Shapiro. 2021. “Is the American Rescue Plan Taking Us Back to the’60s?” FRBSF Economic Letter 2021-27 (October 18).

Barnichon, Regis, and Adam Hale Shapiro. 2022. “What’s the Best Measure of Economic Slack?” FRBSF Economic Letter 2022-04 (February 22).

Mahedy, Tim, and Adam Hale Shapiro. 2017. “What’s Down with Inflation?” FRBSF Economic Letter 2017-35 (November 27).

Powell, Jerome. 2022. “Inflation, Soft Landings and the Federal Reserve.” Interview by Kai Ryssdal, Marketplace Business News Podcast, NPR, May 12. Audio, 27:38.

Shapiro, Adam Hale. 2022. “A Simple Framework to Monitor Inflation.” FRB San Francisco Working Paper 2020-29.

Our baseline forecast assumes that the fading effects of fiscal transfers and pandemic bonuses will lower wage growth by 0.75pp; inflationary shocks will gradually diminish, bringing down short-term inflation expectations by enough to slow wage growth by another 0.5pp; 1 million workers will return to the labor force, slowing wage growth by another 0.2pp; and the job openings rate will fall 0.5pp due to natural normalization, reducing wage growth by just over 0.1pp (Exhibit 9). To return wage growth the rest of the way to 3.75%, policymakers would need to further reduce labor demand by enough to lower the job openings rate by another 1pp and raise the unemployment rate by 0.3pp. We previously estimated that this would require GDP growth of about 1-1.5%, below potential but not in recessionary territory. The labor force participation rate remains depressed by 1.1pp relative to its pre-pandemic rate, equivalent to a 2.85mn reduction in the labor force. We estimate that 700k of this is due to natural population aging, 850k is due to early retirements that are less likely to reverse, and 1.3mn is due to a combination of other reasons including Covid fears, a larger financial cushion due to pandemic savings and house and stock price appreciation, and changes in lifestyle (Exhibit 5). Our baseline forecast assumes that 1mn people will return to the labor force in the next year, including a modest share of the early retirees and most of the others who left. Encouragingly, most prime-age workers who left the labor force during the pandemic still intend to search for work in the next 12 months (Exhibit 6).

Jan Hatzius, Alec Phillips, David Mericle, Spencer Hill, Joseph Briggs, Ronnie Walker and Manuel Abecasis, “Prospects For A Soft Landing: What Could Make The Fed’s Job Easier Or Harder,” Goldman Sachs, June 5, 2022

We use the price effects caused by the passage of rent control in St. Paul, Minnesota in 2021, to study the transfer of wealth across income groups. First, we find that rent control caused property values to fall by 6-7%, for an aggregate loss of $1.6 billion.A calibrated model of house prices under rent control attributes a third of these losses to indirect, negative externalities. Second, leveraging administrative parcel-level data, we find that the tenants who gained the most from rent control had higher incomes and were more likely to be white, while the owners who lost the most had lower incomes and were more likely to be minorities. For properties with high-income owners and low-income tenants, the transfer of wealth was close to zero. Thus, to the extent that rent control is intended to transfer wealth from high-income to low-income households, the realized impact of the law was the opposite of its intention.Table IV presents estimates of Equation 3 using data from the Twin Cities area that control for building size, age, type (multi-family, single-family, etc.), year-month fixed effects, and geographic location fixed effects. Throughout the paper, standard errors are double-clustered by year-month and by the geographic level of the fixed effects. Across the three specifications, we find that rent control caused a statistically significant decline in transaction prices of about 6%. This estimate does not vary across the three types of geographic fixed effects. Additionally, the magnitude of the results is similar to the univariate estimates. This suggests that the negative effect of rent control is not confounded by location or the traits of the properties. In Internet Appendix Table III, we find similar results when the control sample is restricted to the cities neighboring St. Paul, with prices falling between 4.5% and 5% in St. Paul following the reform. The slightly muted response in the closer cities is consistent with a spillover effect from St. Paul onto neighboring areas

Kenneth Ahern and Marco Giacoletti, "Robbing Peter to Pay Paul? The Redistribution of Wealth Caused by Rent Control,"National Bureau Of Economic Research, May 2022, https://www.nber.org/papers/w30083

Core Findings

“…Using a sample of nearly 150,000 real estate transactions over the period January 2018 to January 2022 in the five counties surrounding St. Paul, we first estimate the effect of St. Paul’s rent control on property values. These difference-in-difference tests identify the change in transaction prices of residential real estate in St. Paul following the passage of rent control, relative to the change in prices during the same period in cities adjacent to St. Paul. The tests control for i) year-month fixed effects to absorb common time-series variation in prices in the St. Paul area, ii) detailed location fixed effects to absorb time-invariant cross-sectional variation in prices across granular geographic regions, and iii) property-level attributes, including building age and size, and whether the property is a multi-unit or single-family residence. We find that the introduction of rent control caused an economically and statistically significant decline of 6-7% in the value of real estate in St. Paul. Because we control for year-month fixed effects, these results do not reflect seasonal changes caused by declining volume in winter months. Second, we run additional tests to account for changing preferences for suburbs over city centers using transactions from five Midwestern cities comparable to St. Paul. In triple-differences models that exploit variation in city centers versus suburbs, we still find that rent control caused a 6- 8% decline in property values. Third, we find that rental properties in St. Paul experienced an additional 6% decline in value compared to owner-occupied properties in St. Paul, for a total loss of about 12%. These results imply that the value loss is caused by rent control rather than a spurious variable that affected rentals and non-rentals equally. Finally, we verify that our results are unlikely to be caused by selection bias….”

“…To decompose the observed value loss into direct capitalization effects and indirect negative externalities, we derive a simple model of rent control that allows for stochastic growth rates and probabilistic transitions between owner-occupied and rental housing. Matching the model’s parameters to the St. Paul market, we estimate that about two-thirds of the value loss is driven by capitalization effects and one-third is driven by externalities. These results suggest that capitalization effects of rent control can have a large impact on prices even for owner-occupied properties with a small likelihood of switching to the rental market. The large decline in property values caused by rent control has significant consequences for St. Paul’s economy. Assuming that the market transactions we observe represent the average residential property in St. Paul, rent control would have caused an aggregated loss of $1.57 billion in property value and a 4% expected shortfall in property tax revenue. Given that property taxes are the main form of revenue for the city and the school district, the shortfall in tax revenue is likely to lead to tax increases to maintain city services. Next, we investigate our second research question: how does rent control redistribute wealth? The intention of St. Paul’s rent control is to reduce the burden of housing costs for low-income renters. To study whether the law achieved its intended goal, we test whether the wealth transfers caused by the law are larger when owners have higher incomes and renters have lower incomes….”

“…Next, we use a hedonic pricing model to predict the change in property value following rent control for over 60,000 residential parcels in St. Paul. Due to limitations in the administrative data, we focus on properties with three or fewer units owned by small landlords. This accounts for 90% of rental properties and 54% of all rental housing units in St. Paul. The large majority of residential parcels in St. Paul are single-family residences (89% of all parcels), of which 17% are rental properties. Of the rental properties in our sample, 43% are owned by small landlords. To measure the traits of renters and owners, we use highly granular Census data. We proxy for the traits of renters based on Census data corresponding to the property address. To proxy for the traits of owners, we collect their addresses from the county assessor’s office. To verify that the address of the owner is residential, rather than commercial, we match owners’ addresses to the US Postal Service’s residential delivery indicator (RDI). We then classify rental properties owners as small landlords if their listed address is residential and different than the property address and as large landlords if their listed address is commercial. For small landlords, we proxy for their demographic traits using the granular Census data that correspond with their home address. To test whether transfers are larger when renters have lower incomes and owners have higher incomes, we create a ‘high disparity’ subsample of properties in which owners have incomes above the median-owner’s income and renters have incomes below the median-renter’s income. We also create a ‘low disparity’ subsample in which owners have below-median income and renters have above-median income. The differences between the two subsamples are stark. In the high disparity subsample, the median owner’s income is more than double the income of the median renter, while in the low disparity subsample, owners’ and renters’ median incomes are statistically equivalent. Likewise, the fraction of minority renters is roughly 50% in the high disparity subsample, compared to 25% in the low disparity subsample. Similar variation exists for age and education. In contrast to the stated goals of the rent control law, we find that the largest transfer of wealth occurred in the low disparity subsample (8.52%) in which renters are relatively wealthier, while the smallest transfer occurred in the high disparity subsample (0.89%), in which renters are relatively poorer. This pattern persists in cross-sectional regressions. Wealth transfers are positively related to renters’ income and negatively related to owner’s income.

“…The MoM increase in Case-Shiller was at 2.09%. This is the largest MoM increase in the Case-Shiller index on record (since 1975), and stronger than any month during the housing bubble or during the late ‘70s with high inflation. In March, the months-of-supply was at 1.9 months, and the Case-Shiller National Index (SA) increased 2.09% month-over-month. The black arrow points to the March 2022 dot. In the April existing home sales report, the NAR reported months-of-supply increased to 2.2 months. This month was very likely the peak YoY growth rate - just above the peak last August. Since inventory is now increasing year-over-year (but still low), we should expect price increases to slow. The normal level of inventory is probably in the 4 to 6 months range, and we will have to see a significant increase in inventory to sharply slow price increases, and that is why I’m focused on inventory! Since Case-Shiller is a 3-month average, and this report was for March (includes January and February), this included price increases when mortgage rates were significantly lower than today. In January, the Freddie Mac PMMS averaged 3.4% for a 30-year mortgage, and 3.8% in February. Currently mortgage rates are around 5.25%....”

Bill McBride, "Case-Shiller National Index up 20.6% Year-over-year in March; New Record Monthly Increase,"Calculated Risk, May 31, 2022, https://calculatedrisk.substack.com/p/case-shiller-national-index-up-206

Case-Shiller National Index up 20.6% Year-over-year in March; New Record Monthly Increase

Both the Case-Shiller House Price Index (HPI) and the Federal Housing Finance Agency (FHFA) HPI for March were released today. Here is a graph of the month-over-month (MoM) change in the Case-Shiller National Index Seasonally Adjusted (SA).

The Case-Shiller Home Price Indices for “March” is a 3-month average of January, February and March prices.

FHFA House Price Index

On the FHFA index: U.S. House Prices Rise 18.7 Percent over the Last Year; Up 4.6 Percent from the Fourth Quarter

U.S. house prices rose 18.7 percent from the first quarter of 2021 to the first quarter of 2022 according to the Federal Housing Finance Agency House Price Index (FHFA HPI®). House prices were up 4.6 percent compared to the fourth quarter of 2021. FHFA’s seasonally adjusted monthly index for March was up 1.5 percent from February.

"High appreciation rates continued across housing markets during the first quarter of 2022," said William Doerner, Ph.D., Supervisory Economist in FHFA's Division of Research and Statistics. "Strong demand coupled with tight supply have kept prices climbing. Through the end of March, higher mortgage rates have not yet translated into slower price gains, but new home sales have dropped during the last few months, with a significant falloff in April."
emphasis added
This is the monthly and quarterly indexes. Here is a graph from the FHFA report showing the annual change by region for March 2022 compared to March 2021. Prices have increased sharply everywhere, but especially in the Mountain, South Atlantic, East South Central and Pacific regions.

From S&P: S&P Corelogic Case-Shiller Index Reports Annual Home Price Gain Of 20.6% In March

the S&P CoreLogic Case-Shiller U.S. National Home Price NSA Index, covering all nine U.S. census divisions, reported a 20.6% annual gain in March, up from 20.0% in the previous month. The 10-City Composite annual increase came in at 19.5%, up from 18.7% in the previous month. The 20-City Composite posted a 21.2% year-over-year gain, up from 20.3% in the previous month.

Tampa, Phoenix, and Miami reported the highest year-over-year gains among the 20 cities in March. Tampa led the way with a 34.8% year-over-year price increase, followed by Phoenix with a 32.4% increase, and Miami with a 32.0% increase. Seventeen of the 20 cities reported higher price increases in the year ending March 2022 versus the year ending February 2022.
...
“Those of us who have been anticipating a deceleration in the growth rate of U.S. home prices will have to wait at least a month longer,” says Craig J. Lazzara, Managing Director at S&P DJI. “The National Composite Index recorded a gain of 20.6% for the 12 months ended March 2022; the 10- and 20-City Composites rose 19.5% and 21.2%, respectively.For both National and 20-City Composites, March’s reading was the highest year-over-year price change in more than 35 years of data, with the 10-City growth rate at the 99th percentile of its own history.

“The strength of the Composite indices suggests very broad strength in the housing market, which we continue to observe. All 20 cities saw double-digit price increases for the 12 months ended in March, and price growth in 17 cities accelerated relative to February’s report. March’s price increase ranked in the top quintile of historical experience for every city, and in the top decile for 19 of them. emphasis added
This graph shows the nominal seasonally adjusted Composite 10, Composite 20 and National indices (the Composite 20 was started in January 2000).

House Prices and Inventory

This graph below shows existing home months-of-supply (inverted, from the NAR) vs. the seasonally adjusted month-to-month price change in the Case-Shiller National Index (both since January 1999 through March 2022).

Note that the months-of-supply is not seasonally adjusted.

In March, the months-of-supply was at 1.9 months, and the Case-Shiller National Index (SA) increased 2.09% month-over-month. The black arrow points to the March 2022 dot. In the April existing home sales report, the NAR reported months-of-supply increased to 2.2 months.

This month was very likely the peak YoY growth rate - just above the peak last August. Since inventory is now increasing year-over-year (but still low), we should expect price increases to slow.

The normal level of inventory is probably in the 4 to 6 months range, and we will have to see a significant increase in inventory to sharply slow price increases, and that is why I’m focused on inventory!

Since Case-Shiller is a 3-month average, and this report was for March (includes January and February), this included price increases when mortgage rates were significantly lower than today. In January, the Freddie Mac PMMS averaged 3.4% for a 30-year mortgage, and 3.8% in February. Currently mortgage rates are around 5.25%.

Note: I’ll have more on real prices, price-to-rent and affordability tomorrow.

“…The truth is, an analysis of the 10 best-known, widely cited pre-K programs makes clear that the research around these programs shows not “that ‘pre-K works’” but that “some early childhood programs yield particular outcomes, sometimes, for some children.” Indeed, the most credible research has made clear that “the most meaningful, far-reaching effects” are the product of “carefully designed, well-implemented programs”—not sweeping mandates. And even the studies that do indicate pre-K efforts were beneficial don’t necessarily point at a clear path forward for policies on pre-K. Consider the famed Perry Preschool project, a small pilot preschool and home-visiting program for 3- and 4-year-olds that ran from 1962 to 1967 in Ypsilanti, Michigan. The program provided 58 low-income Black children with two years of a research-based preschool curriculum and weekly home visits. It was found to have positive impacts on participants’ educational and life outcomes, reducing teenage pregnancy, and increasing lifetime earnings. All good stuff. But we’re talking about less than 60 students, in an intensive boutique pilot program, more than a half-century ago. The notion that Perry offers a clear blueprint for broad public policies is just fantastic….”

Frederick Hess, "The Not-So-Certain Science of Pre-K,"American Enterprise Institute, May 25, 2022, https://www.aei.org/op-eds/the-not-so-certain-science-of-pre-k/

The Not-So-Certain Science of Pre-K

There’s a lot to like about preschool. Over the years, my kids have used several different preschools—from a cheerful church basement to a hard-core Montessori—and had a number of really positive experiences. You don’t need to convince me that good, affordable, reliable, and convenient preschool is something we should value.

But it’s a big leap from there to concluding that preschool “works” or that universal pre-K is a good idea. Now, when I offer such qualifications, preschool aficionados immediately leap to patiently (or not-so-patiently) explain that I’m ignoring the science.

As for that science. Earlier this year, researchers at Vanderbilt University released the outcomes of a new gold-standard study of Tennessee’s statewide pre-K program, launched back in 2009. The study has been touted as the only randomized “thorough, ongoing investigation into the impacts of a statewide pre-K program for economically disadvantaged children.” The findings of the initial 2015 report were regarded as highly disappointing, with the benefits of the pre-K program fading before 1st grade. The follow-up results aren’t any better. The pre-K students fared worse on state assessments than their peers and had worse outcomes “for disciplinary infractions, attendance, and receipt of special education services.”

Readers who regularly hear that pre-K “works” might be surprised. They shouldn’t be. The truth is, an analysis of the 10 best-known, widely cited pre-K programs makes clear that the research around these programs shows not “that ‘pre-K works’” but that “some early childhood programs yield particular outcomes, sometimes, for some children.” Indeed, the most credible research has made clear that “the most meaningful, far-reaching effects” are the product of “carefully designed, well-implemented programs”—not sweeping mandates.

And even the studies that do indicate pre-K efforts were beneficial don’t necessarily point at a clear path forward for policies on pre-K. Consider the famed Perry Preschool project, a small pilot preschool and home-visiting program for 3- and 4-year-olds that ran from 1962 to 1967 in Ypsilanti, Michigan. The program provided 58 low-income Black children with two years of a research-based preschool curriculum and weekly home visits. It was found to have positive impacts on participants’ educational and life outcomes, reducing teenage pregnancy, and increasing lifetime earnings. All good stuff. But we’re talking about less than 60 students, in an intensive boutique pilot program, more than a half-century ago. The notion that Perry offers a clear blueprint for broad public policies is just fantastic.

And yet, some years back, Nobel Prize-winning economist James Heckman savaged a newspaper columnist for even questioning the benefits of expanded pre-K. Heckman insisted that it clearly works. How did he know this? Heckman pointed to his “evidence-based analysis of more than 30 years of data” from the Perry Preschool project. “It is as good a trial for effectiveness as those we currently rely on to evaluate prescription and over-the-counter drugs,” he wrote.

While Heckman is a very smart guy, he was wrong here. There’s a difference between a medical therapeutic and a public-policy response. Unlike a preschool program, the ingredients of an over-the-counter drug can be replicated exactly and administered to new patients with precise directions. In the case of Perry, it’s not especially clear just what it means to replicate the program. Was the intervention simply attending a preschool? Doubtful. Was it attending a “high quality” preschool? OK, but just what made Perry “high quality”? Was it the student-teacher ratio? Home visits? The curriculum? All of the above? How exactly does one know?

Unless we know, it’s tough to be confident that policies will deliver the desired results. Worse, the bigger the policy, the bigger the risk that getting things wrong will have adverse consequences. And yet much of the support for universal preschool proceeds with a blind assurance that leaves all such difficult questions aside. For instance, last fall, the Biden administration’s Build Back Better proposal for universal preschool contained sweeping new federal requirements that would’ve governed preschool education standards, credentials, and pay. There was little evidence behind any of this and actually more cause to fear the ways in which it might bureaucratize early-childhood education, squelch small church-based and neighborhood providers, drive up costs, and aggravate staffing challenges. We would’ve been far better served not by assertions of “the science says” but by serious discussions of uncertainty, trade-offs, and sensible compromise.

Helping ensure that families have access to reliable, trusted child-care options is a good idea. But the claims made about preschool—and especially about universal pre-K—tend to be overstated, the benefits are far from certain, and fail to consider that massive new programs can cause real harm. We should keep working to expand access to reliable, convenient, trusted preschool. But we should do so guided by good judgment, not by overreliance on not-so-certain science.

This paper has addressed a classic topic: the optimal accumulation of capital and the welfare effects of government debt in neoclassical growth models. Over the past several decades, the thinking of many economists has been shaped by the Solow growth model and the Diamond overlapping-generations model, which assume certainty and competitive markets. In these models, the real interest rate reflects the marginal product of capital. As a result, the low real interest rates experienced in recent years have tempted some economists to conclude that capital accumulation now has only small social value at the margin and, therefore, that the crowding out of capital by government debt is of little concern. The innovation in this paper is to generalize these two canonical modelsto include firms with market power. In this environment, the real interest rate earned by savers is below the net marginal product of capital. For a plausible calibration of the extended Solow model, the wedge between them is about 4 percentage points. The calibration also suggests that the U.S. economy is dynamically efficient. When market power is introduced into the Diamond model, government Ponzi schemes can have different implications for welfare than they do under competition. Even if a perpetual rollover of government debt and accumulating interest is feasible, the crowding out of capital may still reduce steady-state welfare.Previous work has established a close connection between the feasibility of government Ponzi schemes and the possibility of rational speculative bubbles. This equivalence suggests that in an overlapping-generations model with market power, rational bubbles are possible whenever government Ponzi schemes are. But unlike under competition, these bubbles may reduce welfare because they could occur even in dynamically efficient economies. By diverting saving away from capital accumulation, bubbles may depress aggregate consumption. This topic could be addressed in future research. Much previous work on dynamic efficiency and Ponzi schemes has stressed the role of uncertainty about capital returns. Under uncertainty, the risk-free real interest rate is below the expected marginal product of capital, and attempted Ponzi schemes by the government are riskyThis paper abstracts from these issues by considering models with certainty. A more realistic analysis of the U.S. economy would include both uncertainty and market power. We also leave that topic for future work.

Laurence Ball and Greg Mankiw, "Market Power In Neoclassical Growth Models,"Working Paper, April 2022, http://www.restud.com/wp-content/uploads/2022/05/MS30597manuscript.pdf

Impact of firm market power"...If typical prices are marked up 20% over marginal cost, as some of the literature suggests, the real interest rate is about 4 percentage points below the social return on capital. Correcting for the effects of market power yields an estimate of the social return of more than 8%. Because this estimated social return on capital is much higher than the economy’s growth rate, it indicates that the U.S. economy is dynamically efficient...."
Relationship with "Government Ponzi Scheme""...The impact of a government Ponzi scheme on welfare depends on two effects. The first we call the aggregate effect. Because government debt crowds out capital, it depresses aggregate consumption if the net marginal product of capital exceeds the growth rate and increases aggregate consumption in the opposite case. (This is parallel to the classic Phelps, 1961, result.) The second we call the generational effect: the introduction of government debt improves the allocation of aggregate consumption between the young and the old. This effect is similar to the Pareto improvement that can result from pay-as-you-go social security when the growth rate exceeds the interest rate (Samuelson, 1958; Weil, 2008). In a competitive economy, the aggregate and generational effects of debt work in the same direction. In an economy with market power, when the economy’s growth rate is greater than the real interest rate but less than the net marginal product of capital, the two effects conflict. These findings show that, in the presence of market power, the real interest rate by itself is a poor guide for judging the welfare effect of increased government indebtedness..."

For several years after the 2007-09 recession, roughly a million people moved to the U.S. annually. That pace started to slow during the Trump administration and fell to a trickle after the Covid-19 pandemic started. The slowdown has left the U.S. with 2.4 million fewer immigrants of working age—about 1% of the working-age population—than if pre-2017 immigration trends had continued, according to Giovanni Peri, a labor economist at the University of California, Davis....The slowdown in immigration began in 2017, when the Trump administration adopted a broad range of policies to curb both illegal and legal immigrants, including refugees, foreign recruits, international students and family members of American citizens. Those already-reduced flows were then cut in half in 2020 by the onset of the pandemic and the related restrictions.

Michelle Hackman, "Add Declining Immigration to Problems Weighing on the Labor Market,"Wall Street Journal, April 5, 2022, https://www.wsj.com/articles/add-declining-immigration-to-problems-weighing-on-the-labor-market-11649174837

Add Declining Immigration to Problems Weighing on the Labor Market

DALLAS—In an economy where nearly every industry is short of labor, nursing homes represent an extreme. Their employment has fallen 16% since 2019, and 99% of homes said last fall they didn’t have enough staff.

Burnout and stress are leading causes, but so is a less-known contributor: fewer immigrants.

Emeka Nwaokolo, 36 years old, came to the U.S. from Nigeria for a business trip and met a woman from Dallas he eventually married. He ended up working as a manager for Manchester Care Homes, which runs a network of nursing homes around Dallas.

Over the years, he has referred other Nigerian immigrants from his church to Manchester. In the past few years, though, it has become much tougher for Nigerians to emigrate to the U.S. because of more-stringent visa screening at American consulates abroad. The situation has worsened with pandemic-related backups in immigration processing.

For several years after the 2007-09 recession, roughly a million people moved to the U.S. annually. That pace started to slow during the Trump administration and fell to a trickle after the Covid-19 pandemic started.

The slowdown has left the U.S. with 2.4 million fewer immigrants of working age—about 1% of the working-age population—than if pre-2017 immigration trends had continued, according to Giovanni Peri, a labor economist at the University of California, Davis. The change is being felt as the economy rebounds and many employers struggle to replace workers who were laid off or quit since early 2020, contributing to wage pressure and inflation.

Labor shortages stem from a wide variety of factors, including early retirement, Covid-19 sickness or fear of it, lack of child care and a desire to start a business. Additionally, Mr. Peri noted, fewer immigrants are a factor. Industries with above-average levels of fore

The accommodation and food sector has a high vacancy rate, a high share of foreign-born workers and relatively low wages.

About 21% of truckers are foreign born and the vacancy rate in transportation, warehousing and utilities, which includes trucking, is 6.6%. In construction, 28% of workers are foreign born and 4.8% of jobs are vacant.

In health care and social assistance, which includes nursing homes, 17.2% of workers are foreign born, and 9% of jobs are vacant.

San Antonio home builder Ed Berlanga used to be able to call up his framer or painter and ask if they could assemble an additional crew of two or three workers to take on a new project. “They would say, ‘I have three cousins in Mexico, I’d like to get them to come work with us,’” he said.

Now, no one is coming, Mr. Berlanga said, and he faces more competition for his subcontractors’ time as home demand booms. He is also paying more for almost every service. Paint jobs, which routinely cost him $6 a square foot before the pandemic, now cost $7.50.

The smaller number of immigrants appears to be one reason wages are growing briskly in industries where vacancies are high, from hospitality to long-term care. “The numbers suggest it’s a small contributing factor to what’s happening with wage growth,” though likely dwarfed by factors such as the opportunities low-wage employees have to switch to different lines of work, said Daniel Costa, director of immigration law and policy research at the Economic Policy Institute, a left-leaning think tank.

The slowdown in immigration began in 2017, when the Trump administration adopted a broad range of policies to curb both illegal and legal immigrants, including refugees, foreign recruits, international students and family members of American citizens.

Those already-reduced flows were then cut in half in 2020 by the onset of the pandemic and the related restrictions. The U.S. closed consulates around the world, stopped processing refugees and began turning away most adult migrants at the Southern border under a new public-health authority known as Title 42.

In the 12 months ended last June 30, about 247,000 people moved to the U.S., less than a quarter of the 2016 level and half that of 2019, according to U.S. census data. Those figures don’t distinguish between people who came to the U.S. legally and those who didn’t.

Further crimping the availability of foreign-born workers are pandemic-related delays in processing work permits for those already in the U.S. There were 1.6 million pending work-permit applications at the U.S. Citizenship and Immigration Services at the end of December, up from 646,000 in 2019. It takes an average of nine to 11 months to process applications, versus three to four months at the end of 2019, according to government data.

Guito Tatawho fled violence in Haiti and now lives in Indianapolis, received a work permit in 2017 as part of his asylum application and became certified as a commercial truck driver. In 2020 he bought his own truck and began working for an Illinois trucking company, where he estimates he made $7,000 a month after truck payments.

When he applied to renew his work permit in February 2021, the immigration agency didn’t get to his case before his permit—and driver’s license—expired in August.

“I’m sitting at home right now doing nothing,” Mr. Tata said in February. “My truck is just parked somewhere. I’m losing money.”

On March 18, some seven months after the expiration, he finally got his new work permit.

A spokesman for the immigration agency said it couldn’t comment on an individual case but is aware of unprecedented processing delays and is working to fix them. Last week, the Biden administration said it plans to make expired work permits valid for many immigrants who rely on them.

The administration, in a move it said was to address labor shortages, this year made 55,000 additional visas available for temporary seasonal employers such as landscapers, hotels and ski resorts, nearly doubling the total available.

A political backlash to the administration’s handling of illegal immigration at the Southern border this year has made officials reluctant to take bigger steps to increase legal immigration, according to people familiar with the administration’s thinking.

About a fifth of consulates still aren’t processing most visas, and even those that are open are buried under a backlog of 7.5 million applications, according to State Department data.

As global travel rebounds, the department aims at “reintroducing appointments for the full range of nonimmigrant visas, including for tourism and business travel,” said a spokesman for the department’s Bureau of Consular Affairs. “We will do so in a manner consistent with science and public health.”

In long-term care, the outlook is especially striking. Demand for home health aides and nursing home care is increasing as baby boomers age. Over the past decade, an average 18% of the sector’s workforce has been foreign born. That is close to the national average, but a disproportionate share of those with hands-on jobs such as orderlies are foreign born, according to nursing home operators.

Employment in nursing-care facilities fell to 1.34 million in March from 1.6 million in mid-2019. While the U.S. has some visa programs to help foreign doctors or engineers immigrate, operators of long-term-care homes can’t recruit from abroad because the jobs they are filling wouldn’t qualify for any existing visa program.

A September survey by the American Health Care Association and the National Center for Assisted Living, sister trade groups, found 99% of nursing homes and 96% of assisted-living facilities reporting staffing shortages.

Because of those shortages, 58% of nursing homes and 28% of assisted-living facilities limited how many new clients they could take on. With fewer caregivers, industry officials say, people looking for homes or health aides sometimes can’t even get their calls to agencies returned.

Adam Lampert, the chief executive of Mr. Nwaokolo’s Dallas employer, Manchester Care Homes, and a sister company employing home health aides, said that African immigrants make up roughly 80% of his workforce, many referred to him through a church.

Recently, when his team has tried to recruit through employee networks, it has come up short. There are fewer new arrivals, the people likeliest to take caregiving jobs.

“We’re not trying to discriminate against citizens when we look for employees—quite the opposite,” Mr. Lampert said. “We’re just looking for people to fill the roles that we need. And time and again, we find it is the immigrant population that tends to respond to us.”

Mr. Nwaokolo said caregiving can be demanding, such as when he needs to change clients’ adult diapers or help them go to the bathroom. It can be emotionally taxing, too, particularly working with those who have dementia.

“You have to learn how to be patient,” he said. “You have to watch your anger, because sometimes they can hit you or maybe spit in your face.”

Mr. Nwaokolo, who had been working as a local health official in Nigeria, came to the U.S. in 2015 for a health conference in Boston, where he met his future wife. That meant finding a new job. He wanted to work in healthcare but knew his experience likely wouldn’t transfer. Caregiving appealed to him, he said, because all the children in his village were taught to care for their elders.

Among Mr. Nwaokolo’s colleagues at the Dallas company, many came to the U.S. on tourist visas and ended up staying, typically by marrying an American. Others were sponsored by immediate family members who are U.S. citizens.

His cousin Anthony Chidera Nwakor, who lives in Lagos, applied for a visa to the U.S. in 2019 and was rejected. He applied again last year and is still waiting for an appointment at the U.S. consulate in Lagos. “I want to come there and make some connections, maybe work or go to school,” he said.

Mr. Nwaokolo said he has received offers to work as a traveling home nurse for as much as triple the $15 an hour he currently earns. He has decided to stay put for now because he feels some loyalty to Mr. Lampert, the chief executive of Manchester Care Homes, who paid for his nursing training and has lent him money for his family’s medical care in Nigeria.

“This place is like a family to me,” Mr. Nwaokolo said.

“…Trump's CEA showed, based on credit transactions among manufacturers, that vehicle standards to abate a ton of CO2 cost about $163 on the margin, while even Obama said the abatement was worth only $50. i.e., tightening emissions regulations fails a cost-benefit test by a wide margin. Now Biden claims that new stricter standards pass a cost benefit test. Although this will be cast as an environmental issue, the changes have nothing to do environmental economics: (1) Consumer fuel savings get (mostly) double counted because "behavioral economics."...(2) Biden says that some tightening comes for free because 5 manufacturers had already signed a pledge with California EPA to so tighten...(3) When the above are enough to tilt the scale, all costs and benefits are discounted 3%/yr. When an extra push is needed, Biden discounts environmental benefits at 2.5% per year while everything else is discounted 3%/yr…”

More on coercive paternalism

Trump's DOT and EPA spoke forcefully against paternalism as a justification for fuel standards. If people lack knowledge, give them the knowledge rather than imposing a decision on them. Here is how they said it

"the idea that regulating fuel economy and CO2 emissions can mitigate the consequences of inadequate access to information by placing decisions that depend on access to complete information in the hands of regulators rather than buyers has superficial appeal. Yet commenters do not establish that such a drastic step is necessary to overcome any inadequacy of information, or that requiring manufacturers to supply higher fuel economy will be more effective than less intrusive approaches such as expanding the range of information available to buyers." (85 FR 24608, italics added)

In contrast, Biden's DOT and EPA say nothing like this, but instead extol the purported virtues of "behavioral economics." They do not mention less intrusive approaches, let alone show why they would have fewer net benefits.

Casey Mulligan, "New Emissions Regulations are Coercive Paternalism, not Environmental Science, or even Benevolent Paternalism,"Supply And Demand (In That Order), April 5, 2022, http://caseymulligan.blogspot.com/2022/04/emissions-regulations-are-divorced-from.html

New Emissions Regulations are Coercive Paternalism, not Environmental Science, or even Benevolent Paternalism

Trump's CEA showed, based on credit transactions among manufacturers, that vehicle standards to abate a ton of CO2 cost about $163 on the margin, while even Obama said the abatement was worth only $50. i.e., tightening emissions regulations fails a cost-benefit test by a wide margin.

Now Biden claims that new stricter standards pass a cost benefit test. Although this will be cast as an environmental issue, the new conclusion is driven by assumptions unrelated to environmental economics or climate science:

(1) Consumer fuel savings get (mostly) double counted because "behavioral economics." Specifically,

"The agency’s analysis assumes that potential car and light truck buyers value only the savings in fuel costs from purchasing a higher-MPG model they expect to realize over the first 30 months they own it. Depending on the discount rate buyers are assumed to apply, this amounts to 25-30 percent of the expected savings in fuel costs over its entire lifetime." (p. 420 of DOT's final rule)

This double counting (100 - 27.5% = 72.5% of $98 billion in fuel savings) is more than quadruple the purported $16 billion net benefit shown in Table VI-11 of the final rule.

By comparison, the gross climate benefit is purportedly $27.5 billion. i.e., they would have to more than double their already inflated "social cost of carbon" to push their thumb on the scale as vigorously as they did with "behavioral economics." See below for more on paternalism.

(2) Biden says that some tightening comes for free because 5 manufacturers had already signed a pledge with California EPA to so tighten

But this ignores that California rules, when followed by just a subset of manufacturers, do not reduce the supply of federal credits, whereas changes in federal rules do even if the federal rules are not as strict as California's. The equilibrium credit price is built into the prices paid by purchasers of new cars.

(3) When the above are enough to tilt the scale, all costs and benefits are discounted 3%/yr. When an extra push is needed, Biden discounts environmental benefits at 2.5% per year while everything else is discounted 3%/yr.

"the use of the social rate of return on capital... inappropriately underestimates the impacts of climate change for the purposes of estimating the SC-GHG.... the consumption rate of interest is the theoretically appropriate discount rate in an intergenerational context." (p. 547 of the Technical Support Document. See also p. 573 of the final rule.)

Stock-market short-termism is that rare enemy that unites left and right...Neither is correct...as a compelling short book from Harvard Law School Professor Mark Roe shows...The trickier critique is that short-term thinking leads companies to invest less than they should in long-term projects and in research and development, and to prioritize dividends and buybacks instead. If widespread, the economy would suffer. Yet, the data don’t back up the idea of broad damage to the economy. True, U.S. corporate investment in factories, equipment and suchlike is much lower, as a share of gross domestic product, than at its peak shortly before the 1980 recession. But the same applies to companies in countries where shareholders have less influence, and U.S. companies actually cut investment less. The argument that R&D is lower is simply wrong. U.S. R&D spending reached a post-World War II high in 2020, as a share of GDP. Companies are spending twice as much as in 1980. If there’s an R&D problem, it lies in the public sector, where R&D has been slashed, not the private sector.

James Mackintosh, "Quarterly Capitalism Isn’t Ruining the World,"Wall Street Journal, April 6, 2022, https://www.wsj.com/articles/quarterly-capitalism-isnt-ruining-the-world-11649237479

Quarterly Capitalism Isn’t Ruining the World

Stock-market short-termism is that rare enemy that unites left and right, management and employees, and some of the world’s biggest fund managers and consultants. On the left, “quarterly capitalism” is shorthand for venal executives and greedy shareholders abusing workers and the environment in pursuit of unsustainable profits. On the right, critics fear short-term-oriented stockholders pressure management to make dumb decisions, depressing investment and the economy.

Neither is correct, as a compelling short book from Harvard Law School Professor Mark Roe shows. And that matters for public policy, lawmakers and the dominant narrative in how America’s corporations are run. There are plenty of problems, but they need different fixes.

Start with the left-wing criticism, since it is easiest to debunk. Short-term thinking is clearly not the cause of climate change, pollution or poor treatment of staff. Indeed, it is often the exact opposite: Oil companies have to plan decades ahead before building an expensive new refinery or drilling costly offshore wells. DuPont polluted the Ohio River in West Virginia despite internal warnings about it, not because of a desire for short-term profits but because it didn’t expect to be caught. It succeeded in hiding the problem for many years, a long-term outcome.

The same goes for corporate treatment of workers: Paying only the minimum wage isn’t a short-term strategy, but it was perfectly sustainable for decades for many low-skill jobs, at least until the recent economic boom. Whatever your views on how much labor unions might hurt shareholders, opposing them is definitely not a short-termist approach: Amazon successfully fought against unionization drives for almost a quarter of a century before the first vote for an Amazon union this month.

There are genuine problems, but they are due to companies being able to offload costs—pollution, global warming, societal difficulties—onto others, not short-termism. The incentives for such dumping need fixing, and taking aim at short-term thinking will, as Prof. Roe’s book title says, be “Missing the Target.”

The trickier critique is that short-term thinking leads companies to invest less than they should in long-term projects and in research and development, and to prioritize dividends and buybacks instead. If widespread, the economy would suffer.

Yet, the data don’t back up the idea of broad damage to the economy. True, U.S. corporate investment in factories, equipment and suchlike is much lower, as a share of gross domestic product, than at its peak shortly before the 1980 recession. But the same applies to companies in countries where shareholders have less influence, and U.S. companies actually cut investment less.

The argument that R&D is lower is simply wrong. U.S. R&D spending reached a post-World War II high in 2020, as a share of GDP. Companies are spending twice as much as in 1980. If there’s an R&D problem, it lies in the public sector, where R&D has been slashed, not the private sector.

Companies might turn out to have made a horrible mistake if the cost of borrowing soars and they can’t meet the interest payments, but it has been a long-term trend, not short-term pandering to shareholders. Companies overall haven’t been starved of cash they could have invested elsewhere, contrary to what politicians claim.

Yet, it’s true that executives themselves admit to the worst sort of short-termism. Prof. Roe revisits the most compelling evidence, a widely cited study by John Graham, Campbell Harvey and Shiva Rajgopal, which found almost 80% of executives willing to accept a longer-run cost to maintain short-term earnings. Dig into the study, and while still damning—quarterly numbers shouldn’t matter this much—it turns out only 2% would sacrifice a large amount of value to smooth earnings, and 23% a moderate sacrifice.

If that’s the case, the market may have fixed the problem by itself. Companies that give up profitable projects face more nimble competitors eager to step in, or might themselves start the project later. That could explain the lack of economywide evidence of problems, if the value-adding projects do eventually go ahead.

I support some small fixes that can be done at low or no cost, such as stopping quarterly earnings guidance. But politicians should focus elsewhere, rather than try to address a problem that mostly doesn’t exist—especially since so many of the supposed solutions would give more power to executives, creating other problems.

The book is a must-read for anyone interested in markets and policy, and a great overview of the state of the evidence. But it doesn’t address a linked concern that has always bothered me, that markets are fickle in their commitment to the long term, with damaging consequences.

Investors sometimes think far too long term, bidding up dot-com stocks in the late 1990s, encouraging miners to spend billions on new holes in the ground in the early 2010s, or pushing lossmaking and sometimes zero-revenue speculative growth companies to ridiculous prices early last year. Other times they think too short term, as when fearful investors demanded companies hold on to fortress balance sheets after the 2008 financial crisis in case of a repeat, rather than invest for growth.

I’m not sure there’s a fix for this capriciousness, and it may just be one of the many imperfections of capitalism that we have to live with.

“Evidence from both recent worker surveys and historical data on quits shows that the ‘Great Resignation’ is not as unusual as one might think,” wrote Bart Hobijn... latest Economic Letter from the Federal Reserve Bank of San Francisco... Hobijn argues that prior to 2000, high waves of quits had actually been quite common during rapid economic recoveries in the postwar period, based on the Manufacturing Labor Turnover Survey, which was discontinued in 1981. “The quits waves in manufacturing in 1948, 1951, 1953, 1966, 1969, and 1973 are of the same order of magnitude as the current wave,” he said. “All of these waves coincide with periods when payroll employment grew very fast, both in the manufacturing sector and the total nonfarm sector.”...The so-called quits rate, which measures voluntary job leavers as a share of total employment, remained near a record at 2.9% in February. Roughly 4.4 million Americans quit their jobs in the month, according to Labor Department data back to 2000"

Alister Bull, "‘Great Resignation’ Not That Great After All, Fed Study Says,"Bloomberg, April 4, 2022, https://www.bloomberg.com/news/articles/2022-04-04/-great-resignation-not-that-great-after-all-fed-study-says

Great Resignation’ Not That Great After All, Fed Study Says

High levels of workers quitting their jobs -- dubbed the “Great Resignation” -- may not be that rare after all, according to the latest Economic Letter from the Federal Reserve Bank of San Francisco.

“Evidence from both recent worker surveys and historical data on quits shows that the ‘Great Resignation’ is not as unusual as one might think,” wrote Bart Hobijn, a visiting fellow with the regional Fed’s research department and a professor of economics at Arizona State University.

The so-called quits rate, which measures voluntary job leavers as a share of total employment, remained near a record at 2.9% in February. Roughly 4.4 million Americans quit their jobs in the month, according to Labor Department data back to 2000.

But Hobijn argues that prior to 2000, high waves of quits had actually been quite common during rapid economic recoveries in the postwar period, based on the Manufacturing Labor Turnover Survey, which was discontinued in 1981.

“The quits waves in manufacturing in 1948, 1951, 1953, 1966, 1969, and 1973 are of the same order of magnitude as the current wave,” he said. “All of these waves coincide with periods when payroll employment grew very fast, both in the manufacturing sector and the total nonfarm sector.”

Still, while he suggests that employers will have more success filling open positions, which should ease upward pressure on wages, he doesn’t see that happening until late 2022.

We study the labor market effects of permanent 23-50% reductions in unemployment insurance benefits available in seven states. Leveraging linked firm-establishment data,we find that establishments based in reform states experience 1.5-2.4% faster employment growth relative to the same firm's establishments in other states. Using a similar multi-state firm design, starting salaries are 1.8-7.2% lower in reform states and posted salaries for the same job fall by 1.4-5.5%. These labor supply shocks yield an average labor demand elasticity of -1.0. Our results reveal a substantial decline in match quality and worker bargaining power as UI benefits become less generous.Specifically, we study the employment and earnings responses to reforms in 7 different states which sharply cut the generosity of their state UI programs in the 2010s. The largest of these state-level reforms occurred in North Carolina in 2013. This reform is well-suited for understanding the consequences of cutting UI benefits for two reasons. First, it was larger than any previous rollback implemented in the U.S. The change simultaneously reduced the maximum weekly benefit from $535 to $350 and the maximum duration from 26 to 20 weeks. The combined reductions permanently reduced the maximum value of UI benefits by 50%. Second, the cuts were implemented based on the insolvency of North Carolina’s state UI fund, rather than local labor market conditions. This allows us to compare workers in North Carolina to other states with similar labor market trends, but which had more prudent funding of their state UI programs. Six other states also enacted sizable reductions in UI generosity, but which were more modest compared to North Carolina. These “moderate reform” states (Florida, Georgia,Kansas, Michigan, Missouri, and South Carolina) cut maximum benefit durations by 6 weeks while holding fixed weekly benefit amounts, resulting in the maximum value of benefits falling permanently by 23%. Our first key finding is that following the reform, North Carolina-based establishmentsexperience 2.4% faster employment growth than do their same-firm counterparts in other states over the two years after the reform. For the six moderate reform states, employment grows 1.5% relative to same-firm counterparts in other states. These results suggest that any contractionary effect on consumer spending and aggregate demand, or increased competition for jobs are not large enough to overturn the incentive effects of finding a new job quickly. Using data from Glassdoor, our second key finding is that the earnings of new hires fall by an economically and statistically significant 7.2% in North Carolina establishments relative to the same firm’s establishments in control states. In the moderate reform states, the corresponding effect is a 1.8% decline in starting salaries. These drops combine not only the effect for new hires transitioning from unemployment and changing jobs, but also those entering the labor force. The drop in earnings is unlikely to be explained by negative worker composition effects, as new hires in the post-reform period are not negatively selected based on demographic characteristics in the CPS. Using the Glassdoor data, we estimate that lower match quality in the form of firm and occupational downgrading can account for approximately 40% of the wage effect. The remaining 60%-a 5.9% and 1.6% drop in starting salaries in North Carolina and the moderate reform states, respectively-is due to a decline in either unobserved match quality or worker bargaining power.One limitation of the Glassdoor data is that we cannot entirely rule out unobserved degradations in match quality or unobserved changes in the composition of new hires. Whether unemployment insurance cuts are desirable from a policy perspective depends on the benefits versus costs. On the positive side, these UI reforms stimulated employment growth and lowered benefit payouts. But counterbalancing this was a reduction in the wages of new hires, due to a combination of lower match quality and reduced bargaining power. This tradeoff adds a layer of complexity to debates on the optimal level of unemployment insurance benefits.

Core of paper, "...Specifically, we study the employment and earnings responses to reforms in 7 different states which sharply cut the generosity of their state UI programs in the 2010s. The largest of these state-level reforms occurred in North Carolina in 2013. This reform is well-suited for understanding the consequences of cutting UI benefits for two reasons. First, it was larger than any previous rollback implemented in the U.S. The change simultaneously reduced the maximum weekly bene t from $535 to $350 and the maximum duration from 26 to 20 weeks. The combined reductions permanently reduced the maximum value of UI benefits by 50%. Second, the cuts were implemented based on the insolvency of North Carolina's state UI fund, rather than local labor market conditions. This allows us to compare workers in North Carolina to other states with similar labor market trends, but which had more prudent funding of their state UI programs. Six other states also enacted sizable reductions in UI generosity, but which were more modest compared to North Carolina. These “moderate reform" states (Florida, Georgia, Kansas, Michigan, Missouri, and South Carolina) cut maximum bene t durations by 6 weeks while holding fixed weekly benefit amounts, resulting in the maximum value of benefits falling permanently by 23%. We analyze North Carolina's reform separately due to its more drastic nature, and combine the six moderate reform states to gain precision. In contrast to most studies, we estimate the effects of reductions in UI generosity, rather than expansions, during a period when the labor market was recovering, rather than languishing...."

Key Findings

"... Our first key finding is that following the reform, North Carolina-based establishments experience 2.4% faster employment growth than do their same firm counterparts in other states over the two years after the reform. For the six moderate reform states, employment grows 1.5% relative to same firm counterparts in other states.These results suggest that any contractionary effect on consumer spending and aggregate demand, or increased competition for jobs are not large enough to overturn the incentive effects of finding a new job quickly. Instead, our results are more consistent with workers' employment and reservation wages depending on their outside options, which decrease when UI benefits become less generous, as suggested by Mortensen and Pissarides (1994) and Mitman and Rabinovich (2015). We focus on the wages of new hires given downward wage stickiness for already employed workers (Kahn 1997; Pissarides 2009; Haefke et al. 2013)...."

"... Using data from Glassdoor, our second key finding is that the earnings of new hires fall by an economically and statistically significant 7.2% in North Carolina establishments relative to the same firm's establishments in control states. In the moderate reform states, the corresponding effect is a 1.8% decline in starting salaries.These drops combine not only the effect for new hires transitioning from unemployment and changing jobs, but also those entering the labor force.6 The drop in earnings is unlikely to be explained by negative worker composition effects, as new hires in the post-reform period are not negatively selected based on demographic characteristics in the CPS. Using the Glassdoor data, we estimate that lower match quality in the form of rm and occupational downgrading can account for approximately 40% of the wage effect. The remaining 60%{a 5.9% and 1.6% drop in starting salaries in North Carolina and the moderate reform states, respectively is due to a decline in either unobserved match quality or worker bargaining power. One limitation of the Glassdoor data is that we cannot entirely rule out unobserved degradations in match quality or unobserved changes in the composition of new hires. We overcome this limitation by estimating the reforms' effects on posted wages for the same job, within the same rm, but across treated versus non-treated establishments. We use the near-universe of posted wages in online job ads from Burning Glass Technologies, which as noted in Hazell and Taska (2020), has the key advantage that they are not contaminated by the compositional or match quality effects which could be present in the wages of new hires. We estimate that the reforms generated a 5.5% and 1.4% reduction in posted wages, effect sizes which are nearly identical to the Glassdoor estimates that account for occupational and firm fixed effects.

"... We find that the reductions in UI generosity result in an average 2.8 week decline (8% drop) in unemployment spell lengths...."

"...The unexpected UI reductions represent exogenous negative shocks to workers' outside options, which shift labor supply curves outwards. Marginal revenue product of labor declines for two reasons: increases in employment due to lower reservation wages and reduced match quality. With two points along rms' labor demand curves, we can calculate the labor demand elasticity. Combining our estimates of the changes in employment and in posted wages from the EEOC and Burning Glass analyses, respectively,we derive an average labor demand elasticity of -1.0 across all of the 7 reform states. These are in line with historical estimates of labor demand elasticities calculated based on data from British plants and coal mines, American women following World War II, and manufacturing labor in Germany. Our elasticity lies at the higher end of estimates based on changes to the minimum wage, which could be explained by UI covering low, middle, and higher wage workers...."

Evidence

"...In Figure 1, we plot the evolution of the short-term recipiency rate over time in North Carolina, starting after the official end of the Great Recession and continuing for 9 years. The blue line in panel (a) plots the rate in North Carolina at the monthly level, while the red line plots the corresponding rate for other Southern and Midwestern states which did not change their UI program over this time period. The dashed vertical line marks the date North Carolina passed their UI reform bill, while the solid vertical line denotes the date the reform became effective for newly led claims. Prior to the implementation of the reform, North Carolina (blue line) and other states (red line) exhibit a similar level and trend for short term UI receipt. Since the economy was recovering from the Great Recession, UI participation is gradually falling in the pre-period, with some seasonal patterns present as well. After the reform, the control states' recipiency rate continues its mild decline. In sharp contrast, North Carolina's rate drops shortly after the reform's implementation.20 Two years after the reform, North Carolina's recipiency rate is 10% compared to a rate of 30% in control states. Panel (b) graphs the difference between the blue and red lines, along with 95% confidence intervals. There is no evidence for differential pre-trends, with the UI recipiency rate difference between North Carolina and the control states bouncing around zero. After the reform, the gap becomes negative and widens to a 20 percentage point difference roughly two years after the reform, with this gap persisting until the end of our sample period. Estimates for the moderate reform states can be found in panel (c). Using a dynamic event study design, there is a 5.5 percentage point reduction in the short-term UI recipiency rate in treated states relative to controls in the post-reform period. This smaller effect for moderate reform states is expected given the more extreme cuts enacted by North Carolina. These analyses confirm that the UI reforms sharply curtailed the use of UI, which was the intent of the law changes. The reduction in use is partly mechanical, as individuals were eligible for 6 fewer weeks of benefits after the reform. But it could also be partly driven by the 35% reduction in weekly bene t levels in North Carolina, which could have caused individuals to exit UI earlier..."