Macro Roundup

Evidence from the GI Bill passed after the 2001 attacks finds that the money did expanded both college enrollment and BA completion, but reduced average annual earnings particularly for individuals with lower test scores and fewer occupational skills who the opportunity costs of enrolling in college as opposed to working is high. Results suggest talent is the limiting constraint, "...For the full set of Army veterans separating during this period,we find positive impacts on educational attainment but negative impacts to earnings. This result suggests that the housing subsidies are so generous as to bedrawing veterans into education and out of the labor market - even for veterans with low marginal returns to education.We explore whether this negative impact is driven by changes in college type, heterogeneity in the returns to college, or reductions in labor force experience. Our results suggest that some combination of lost experience and enrollment choices of veterans from military occupations that have lower predicted returns explains some of the effect. Veterans who were already enrolled when the new policy arrives see very modest degree attainment gains and suggestive evidence of a negative effect on earnings seven years after initial enrollment....Overall, our results suggest caution in providing education subsidies of this generosity to non-traditional students. While educational attainment increased modestly as a result of the benefit expansion, subsequent earnings did not. Furthermore, our analyses suggest that it is unlikely that earnings will ever grow enough to recover the earnings lost during the first seven years following separation….”

Note the negative earnings could be driven by the low value add of many of the schools (the for profit education sector)

Bottomline:“…We find that the PGIB raises the educational attainment of veterans at both the B.A. and A.A. levels. Considering the “overall” impacts of PGIB, the policy raised B.A. attainment by 2 percentage points and associate’s degree attainment by 0.4 percentage points. The impacts for already enrolled veterans are smaller in magnitude and may be zero. These impacts are not large given the generosity of the program and the responses found in studies of earlier GI Bills. Angrist (1993) found that veteran’s benefits on average raised schooling by 1.4 years. We are finding average increases of 2 months!Angrist and Chen (2011) find that veteran status (appropriately instrumented) raises average years of schooling by 0.24 to 0.27 years, raises B.A. attainment by 5 percentage points and associate’s attainment by 6 to 9 percentage points. Despite the large benefit amounts, PGIB delivers smaller educational gains than earlier GI Bills. Given the increase in bachelors attainment of 2.75 percentage points and an average increase in benefits of $8600, this implies a cost per additional bachelor’s degree of around $313,000. This calculation uses the differences in differences estimates. If we instead use the geographic variation to estimate impacts of total benefits used and increased bachelors attainment from a $1000 increase in annual benefits offered, we obtain $1460/.0046 = $317,000 per additional bachelor’s degree. These estimates are on the higher end of the range of estimates summarized by Dynarski (2003) and Dynarksi Hyman Schazenbach (2013). Our most remarkable findings are the negative impacts on labor income for the sample of veterans exposed to the program. Seven years after their separation from the Army, the PGIB lowered wages by about 1.8 percentage points. There are numerous possible explanations for this finding. One possibility is that this estimate constitutes a medium-term effect caused by the reduced labor market experience reflective of the schooling gains that we find. This is not our preferred explanation because the negative impacts survive all specification checks in which we control for years of experience, and the additional months of schooling are small relative to the average reduction in earnings. Our hypothesis is that the negative earnings returns stem from (a) the low value-added of many of the schools chosen by veterans under the PGIB combined with (b) the generosity of the BAH, which may be inducing veterans to forego valuable labor market human capital accumulation in favor of very marginal school enrollments.By pursuing schooling opportunities of marginal value, veterans may be missing out on opportunities to build occupation- or firm-specific human capital or to immediately put their Army-taught skills to work in the labor force. Admittedly, this is a more complex version of the experience story, rather than a rejection of the experience hypothesis. Our findings motivate ongoing work such as Barr et al, et al. (2019), who are providing additional information to separating veterans about the earnings returns to various frequently attended colleges and chosen majors. The contrast between PGIB impacts and that of earlier GI Bills may be driven by the changing composition of the Army. The Korea and Vietnam era GI Bills studied by Angrist and Chen (2011) involved conscription, which likely included large groups of young men who were well suited to college-going and further education. The PGIB was implemented in an all-volunteer Army, and so is offered to a group of young women and men who have already chosen the military over college after leaving high school…”

Dataset: "...We use detailed administrative data for a large set of individuals to examine the effect of increased benefits for higher education. We merge the full set of U.S. Army veterans separating between 1994 and 2016 with National Student Clearinghouse (NSC) data and population-based administrative tax return data. The NSC provides measures of college-going and degree attainment. Information from the 1098-T tax form provides supplementary information on college-going. The 1098-Ts are particularly useful because they have complete coverage of forprofit institutions while the NSC does not. Tax return data also contain information on the utilization of federal tax benefits for higher education, labor income, total income, and demographic characteristics such as marital status and family size. We supplement these with third-party reported information on labor income from Forms W-2 and 1099-MISC. These information returns provide individual-level measures of earnings that are available even if an individual does not file a tax return. Because these administrative tax records are available through 2017, we are able to track income and demographic outcomes for eight or more years after separation from the Army and initial enrollment in college..."
Evidence Related To Enrolment:"...Figure 6 uses “ever enrolled” as the outcome, as measured using NSC data, and plots the coefficients on the interactions between eligibility status and separation year. The estimate for the 2009 cohort is that PGIB raised enrollment by 5 percentage points relative to the 2002 cohort. In Figure 7, we perform a similar exercise using data from Form 1098-T to measure enrollment across various windows post-separation. Using the largest window (within 6 years of separation), we again find enrollment impacts of about 5 percentage points for the 2009 cohort. In addition to confirming the estimates from the NSC data, this figure provides further support for treating the 2007 and prior cohorts as being in the pre-period. The vertical line in each panel indicate the first cohort whose enrollment could have been affected by the higher benefit levels. For example, in panel (a), enrollment within a year of separation could have been affected for those separating in 2008 (as the benefits were available in fall 2009), but would not have been affected for those separating in 2007. The figure illustrates that even though the 2007 and prior cohorts were eligible for the PGIB benefits when they became available in 2009, there was no effect on their decision to enroll. This is consistent with the general pattern of post-separation enrollment and benefit use which tends to occur quickly after separation or not at all…”

“…Figure 19 presents our standard event study for labor income measured seven years after separation, illustrating a reduction in labor income of about $1,000 for cohorts separating in 2008 and after. In Table 7, we show PGIB’s impacts on wages, log wages, labor income, log labor income, and indicator variables for positive earnings and labor income, each measured seven years after separation from the Army. The PGIB has consistently negative impacts on wages and labor income for veterans. In Panel A, the PBIG reduces annual earnings seven years after separation by about $900. In the log wages specification (which exclude zeroes), the negative impact is roughly 2 percentage points. The impacts of the PGIB on whether a veteran “has W-2 wages” is small, negative and not statistically significant…”

Andrew Barr, Laura Kawano, Bruce Sacercote, WIlliam Skimmyhorn and Michael Stevens, "You Can't Handle The Truth: The Effects Of The GI Bill On Higher Education And Earnings" National Bureau Of Economic Research, July 2021, https://www.nber.org/papers/w29024

Tracy Alloway, "One Simple (Political) Reason to Explain Bond Yields," Bloomberg, July 21, 2021, https://www.bloomberg.com/news/articles/2021-07-21/one-simple-political-reason-to-explain-bond-yields

One Simple (Political) Reason to Explain Bond Yields

Joe and I have pointed out before that on a longer-term chart the move lower in U.S. Treasury yields doesn’t look so odd.

The Bear Traps report points out that higher yields as politicians coalesce around an agenda isn’t necessarily a new dynamic. In fact they say the yield curve has a history of steepening as control of the U.S. government shifts from two parties to one party. That’s exactly what happened in late 2016, when President Donald Trump won the presidency with Republicans in control of both the Senate and the House of Representatives.

You can see the dynamic from the (rough) chart below, which shows bond yields moving up when there’s a big spike in Republican or Democrat confidence, according to the Michigan survey of consumers (which I’ve used as a proxy for the parties securing greater control). Of course, the opposite move can also happen when a party loses control — with bond yields heading lower after the GOP lost control of Congress in late 2018.

“As you move from two-party to one-party control, the curve steepens (higher long-end yields) as global investors digest the political risk of having one party with their hands on the purse strings,” Bear Traps wrote in its report published on Tuesday. “In the summer of 2017, tax cuts were deemed ‘dead’— bond yields crashed, by November/December of that year, the GOP finally delivered of the fiscal juice, and the curve re-steepened.”

"...The world’s unicorn herd is multiplying at a clip that is more rabbit-like. The number of such firms has grown from a dozen eight years ago to more than 750, worth a combined $2.4trn. In the first six months of 2021 technology startups raised nearly $300bn globally, almost as much as in the whole of 2020. That money helped add 136 new unicorns between April and June alone, a quarterly record, according to cb Insights, a data provider. Compared with the same period last year the number of funding rounds above $100m tripled, to 390. A lot of this helped fatten older members of the herd: all but four of the 34 that now boast valuations of $10bn or more have received new investments since the start of 2020....The torrent of cash can be explained by two factors. The first is a divestment spree by the startups’ early venture-capital (vc) backers. These stakes command top dollar from investors desperate for exposure to the pandemic-era digitisation wave. Exits, via public listings and acquisitions, more than doubled globally year on year, to nearly 3,000. The proceeds are flowing back into new vc funds, which have so far this year raised $74bn in America alone, nearing the record $81bn in 2020 in half the time. The venture capitalists cannot spend the dough fast enough. In the three months to June Tiger Global, a particularly aggressive New York investment firm, made 1.3 deals on average every business day. The second reason for soaring valuations is more competition among investors. Relative newcomers to tech-investing, such as pension funds, sovereign-wealth funds and family offices, are encroaching on the private markets that used to be dominated by vc firms from Sand Hill Road in Palo Alto. In the past quarter “non-traditional” investors in America took part in nearly 1,800 deals that together raised $57bn. Many may have been encouraged by the success of earlier forays by dabblers from outside the vc world. Their annual returns from exited investments in a first round of financing have averaged 30% in the past decade, reckons PitchBook, another data firm. That is more than double the 10-15% for veteran vcs...."

Economist Staff, "Technology unicorns are growing at a record clip,"The Economist, July 22, 2021, https://www.economist.com/business/2021/07/19/technology-unicorns-are-growing-at-a-record-clip

Technology unicorns are growing at a record clip

Aileen lee, a venture capitalist who founded an investment firm called Cowboy Ventures, coined the term “unicorn” in 2013 to refer to what was then a rare, mythical species: privately held startups valued at $1bn or more. Any magical attributes aside, today they are commonplace—and becoming ever more so. Consumers, who stand to benefit from an array of novel, often cheap products and services, can expect to enjoy the ride. Investors betting on the unicorn derby should tread more carefully.

The world’s unicorn herd is multiplying at a clip that is more rabbit-like. The number of such firms has grown from a dozen eight years ago to more than 750, worth a combined $2.4trn. In the first six months of 2021 technology startups raised nearly $300bn globally, almost as much as in the whole of 2020. That money helped add 136 new unicorns between April and June alone, a quarterly record, according to cb Insights, a data provider. Compared with the same period last year the number of funding rounds above $100m tripled, to 390. A lot of this helped fatten older members of the herd: all but four of the 34 that now boast valuations of $10bn or more have received new investments since the start of 2020.

The latest tech darlings are no longer primarily Uber-esque marketplaces for matching services with consumers. Instead, they offer, or are developing, sophisticated products, often in more niche markets. Some 25% of the funding in the second quarter went to financial-technology firms, with lots also flowing into artificial intelligence, digital health and cyber-security (see chart).

The torrent of cash can be explained by two factors. The first is a divestment spree by the startups’ early venture-capital (vc) backers. These stakes command top dollar from investors desperate for exposure to the pandemic-era digitisation wave. Exits, via public listings and acquisitions, more than doubled globally year on year, to nearly 3,000. The proceeds are flowing back into new vc funds, which have so far this year raised $74bn in America alone, nearing the record $81bn in 2020 in half the time. The venture capitalists cannot spend the dough fast enough. In the three months to June Tiger Global, a particularly aggressive New York investment firm, made 1.3 deals on average every business day.

The second reason for soaring valuations is more competition among investors. Relative newcomers to tech-investing, such as pension funds, sovereign-wealth funds and family offices, are encroaching on the private markets that used to be dominated by vc firms from Sand Hill Road in Palo Alto. In the past quarter “non-traditional” investors in America took part in nearly 1,800 deals that together raised $57bn. Many may have been encouraged by the success of earlier forays by dabblers from outside the vc world. Their annual returns from exited investments in a first round of financing have averaged 30% in the past decade, reckons PitchBook, another data firm. That is more than double the 10-15% for veteran vcs.

This winning streak may yet end in tears. That is what happened two years ago, when richly valued firms with shaky business models either fizzled after their initial public offerings (like Uber and Lyft, two ride-hailing rivals) or never got that far (WeWork, an office-rental firm whose flotation was shelved after investors got cold feet). Many recently listed unicorns continue to bleed cash. According to The Economist’s calculations, those that went public in 2021 made a combined loss of $25bn in their latest financial year.

Assessing whether the remaining ones are worth their lofty valuations looks harder than ever. Like their predecessors, they do not disclose financial results. At the same time, extrapolating from the earlier unicorns, which tended to pursue growth at all costs in winner-takes-all markets, offers little help because today’s lot often aim to capture good margins by selling genuinely unique technology. This could be a more sustainable strategy—if the technology works. But it is harder for non-experts to evaluate, especially based on what is often little more than a prototype. Nikola and Lordstown, two electric-vehicle companies that listed in 2020 via reverse mergers with special-purpose acquisition companies (spacs), are under investigation by American authorities over allegedly exaggerating the viability of their technology.

Another risk comes from politics. Authorities around the world are growing warier of letting tech firms get too big or entering regulated markets such as finance or health care. As part of a broader crackdown against big tech firms China’s government recently sabotaged the operations of Didi, by banning its app from Chinese app stores days after the firm’s $68bn initial public offering in New York, ostensibly over misuse of users’ data. Such moves have chilled investors’ appetite for Chinese startups, funding for which has actually declined in the past two quarters. In America the Securities and Exchange Commission is beginning to scrutinise the use of cryptocurrencies. Many crypto-exchanges set investors’ pulses racing in last year’s bitcoin rush. Now the market capitalisation of Coinbase, one of the biggest, has shrunk by half, or $56bn, since peaking after its listing in April.

Investors, then, had better beware. For everyone else, the unicorn stampede is a boon. Because venture investments involve mostly equity and little debt, even flops such as WeWork or cautionary tales like Didi pose little risk to the financial system. So long as venture capital is bankrolling lossmaking startups while they offer subsidised services or develop clever new products, consumers have no reason to look the gift horned horse in the mouth.

Matt Klein offers up evidence that German fiscal policy needs a sea change as the German economy is likely under its potential, "...Germany’s massive trade surpluses—which reflect insufficient consumer spending and business investment relative to domestic production—and its poor track record of creating full-time jobs since the end of reunification both suggest that the German economy is operating far below its potential....Replace the “non-accelerating wage rate of unemployment”, which was 3.4% in Germany in 2019, with the actual 2019 unemployment rate after subtracting the number of long-term jobless (1.2%)...Assume that the employment gap between German women and German men (about 9 percentage points in 2019) should be comparable to the gaps in Scandinavian countries (about 3 percentage points)...Assume that many part-time workers would be willing and able to work an extra 1-2 hours each day…”

How to Improve German Fiscal Policy Without Removing the Schuldenbremse

Much of Europe has been locked into decades of stagnation thanks to misguided ideas about how much governments should be allowed to spend and borrow. Instead of investing in growth and supporting household consumption, the public sector has systematically taxed too much and spent too little.1

Those harmful ideas have supporters everywhere, but the most influential proponents of needless austerity have been Germany’s business, academic, and policy elites. They were the ones who:

Led the push for public borrowing limits and prohibitions on “excessive deficits” in the Maastricht Treaty that created the European Union in 1992

Demanded that those limits be permanently enforced through regular monitoring and a system of penalties with the 1997 Stability and Growth Pact (SGP)

Inserted the Schuldenbremse, or “debt brake,” into Germany’s Basic Law in 2009

Exported that law to Germany’s neighbors by convincing them to agree to the Treaty on Stability, Coordination and Governance in the Economic and Monetary Union (the Fiscal Compact)

Kept German fiscal policy even tighter than required since 20142

While there were always Germans who opposed this consensus, they have long been in the minority. Ironically, German leaders were among the first to realize that the macroeconomic stance imposed by the European Central Bank and the SGP in the early 2000s would mire the country in stagnation, with Chancellor Gerhard Schröder calling for lower interest rates and exemptions to the strict limits on budget deficits.

But Schröder failed, and his successors—some of whom had helped designed those institutions in the first place—convinced themselves and the general public that “fiscal discipline” was helpful rather than harmful. Thus it was the center-left Social Democrat Peer Steinbrück who developed the Schuldenbremse when he was finance minister in 2005-2009.

Ahead of September’s federal elections, only the Greens have committed explicitly in their platform to altering the debt brake. They want to carve out an exemption for investments “in modern infrastructure and mobility, in energy conservation and energy-efficient construction, in education, research and innovation.”

So it’s worth paying attention to an intriguing new paper from Philippa Sigl-Glöckner, Max Krahé, Pola Schneemelcher, Florian Schuster, Viola Hilbert, and Henrika Meyer published by the Dezernat Zukunft’s Institute for Macrofinance. The whole thing is worth reading in full for its treatment of everything that’s wrong with the establishment consensus, as well as where Germany ought to end up.

Unsurprisingly, their preferred solution involves rewriting Germany’s constitution and the European treaties. That’s not an impossible hurdle—and is clearly the optimal solution—but it’s a big challenge. Fortunately, the authors show that it’s possible to shift to a less austere fiscal posture without altering Germany’s federal constitution or the treaties. A lot can be accomplished with a few technical tweaks to “the inputs to the calculation of potential output.”

The Potential of “Potential”

Article 109 of Germany’s Basic Law states that “the Federation and Länder may introduce rules intended to take into account, symmetrically in times of upswing and downswing, the effects of market developments that deviate from normal conditions,” while Article 115 gives the government discretion over “the procedure for the calculation of the yearly limit on net borrowing, taking into account the economic cycle.”

The big question is how to determine what’s “normal.” It’s more art (or pseudoscience) than science. But the ambiguity also provides an opportunity to reshape German—and perhaps even European—fiscal policy.

German law gives the task of estimating the Konjunkturkomponente (the cyclical component) to the finance ministry, which in turn uses a methodology developed in consultation with the European Commission. If the German economy is running below the official estimate of “potential,” Germany’s federal deficit can be bigger than the constitutional maximum of 0.35% of national income, while the states can also borrow more than their target deficit of zero.

The self-imposed budget limits across the rest of Europe are similarly based on estimates of each country’s “output gap,” with governments in economies with large output gaps having more flexibility than those that are ostensibly close to potential.

This flowchart from the paper explains how Germany’s borrowing rule is applied in practice:

When I started The Overshoot, I pointed out the massive gap between what forecasters had reasonably expected on the eve of the financial crisis and what ended up happening. Rather than treating this as an ongoing catastrophe, most official forecasters have responded by progressively lowering their expectations of the global economy’s “potential.” See, for example, the Congressional Budget Office’s continual downgrades of what the U.S. economy can do.3

Perhaps the most egregious model—which thankfully isn’t used anywhere to calibrate monetary or fiscal policy—is the one developed by Federal Reserve economists Kathryn Holston, Thomas Laubach, and John Williams (currently the President of the New York Fed) to estimate what they call R*, or the “natural rate of interest.”4 The HLW model implies that many rich countries were running “above potential” for years after the financial crisis.

As Sigl-Glöckner and her co-authors put it, “ does not provide a well-founded direction for fiscal policy.” Estimates and models are repeatedly modified to generate “a minimum level of economic and political plausibility,” with the European Commission essentially turning off the computer during the pandemic to “to avoid implausibly large reductions in potential output.”

But it doesn’t have to be that way. Sigl-Glöckner and her co-authors suggest three small tweaks that would immediately help:

Replace the “non-accelerating wage rate of unemployment”, which was 3.4% in Germany in 2019, with the actual 2019 unemployment rate after subtracting the number of long-term jobless (1.2%)

Assume that the employment gap between German women and German men (about 9 percentage points in 2019) should be comparable to the gaps in Scandinavian countries (about 3 percentage points)

Assume that many part-time workers would be willing and able to work an extra 1-2 hours each day

Those modest changes would give the German government a bit of extra fiscal space once the debt brake turns back on in 2023, although not very much. Sigl-Glöckner and her co-authors think that “a full capacity utilisation fiscal policy requires, in the medium term, that the very architecture of rules built around output gaps, however calculated, should be abandoned.”

I’m not as pessimistic as they are about what could be accomplished solely by adopting a more expansive view of economic potential, although getting through my preferred changes might be more contentious than their proposed tweaks. But even their approach could have a substantial effect if adopted more widely across Europe.

More than half of all the unemployed in Italy are long-term unemployed, as are more than 70% of the unemployed in Greece. More than 20% of the “extended labor force” in Greece, Italy, and Spain were counted as unemployed or underemployed at the end of 2019. In France, the number was 17%. Even in Austria, the Netherlands, and Scandinavia, the “slack” in the job market represented more than 10% of the potential workforce. Those numbers are all far higher than in Germany.

There’s no substitute for changing minds

Europe’s harmful fiscal rules were the product of a misguided elite consensus. So while creatively reinterpreting those rules could accomplish a lot, replacing the underlying consensus that drove the continent into years of needless austerity is probably a necessary prerequisite.

There are hints that this could already be happening. Klaas Knot, the head of the Dutch central bank—normally a bastion of European conservatism—recently argued that “the SGP should be sufficiently flexible to allow for sizable and sustained expansionary fiscal policy, beyond normal automatic stabilization, if economic circumstances so dictate” and that governments “should also increase potential economic growth that ultimately generates the debt repayment capacity…by maintaining a sufficient level of public investment.”

Other officials at the European Central Bank have been calling for more expansive fiscal policy, and the ECB as a whole just pledged to underwrite a looser approach by committing to keep interest rates “at their present or lower levels” for years.

The laggards are the political leaders in the major economies, most notably in Germany. Hopefully the new paper from Sigl-Glöckner and her co-authors will help move the needle there.

1 The problem afflicts much of the rich world, not just Europe, but this piece is primarily about Europe and the problem is worst in Europe.

2 If you want to learn a lot more about those policy choices and their impact on Germany and the world, I recommend reading Trade Wars Are Class Wars.

3 In my previous piece, I looked at how some simple assumptions could lift the size of America’s latent labor force by more than 16% from what’s currently assumed by the Congressional Budget Office. That would go a long way to closing the gap between their forecasts of potential GDP from 2007 and more recently.

4 Whether such a thing exists is a topic for another time.