Macro Roundup

Research that mirrorsMcKinsey'swork on superstar cities and Simon Johnson recentbookfinds that "innovation industries" (defined here sectors whose R&D expenditures exceed $20,000 per worker and have a STEM-worker share of 45%) growth has increasingly consolidated in just five metro areas, four of which are on the west coast (Boston, San Francisco, San Jose, Seattle, and San Diego). 90% of the nation’s innovation-sector growth during the years 2005 to 2017 took place in these five regions, in contrasts the bottom 90% of metro areas (343 regions) lost share. The net result is one-third of innovation jobs now reside in 16 counties, and more than half are concentrated in 41 counties.

Note they note that Simon Johnson would accept "hypothetical" reductions in the efficiency of R&D spending in exchange for re-balancing regional economic gains, "...Similarly, Gruber and Johnson would accept a “hypothetical” reduction in the efficiency of R&D spending in exchange for a national campaign to expand the pipeline of new ideas and concomitant regional economic gains. They urge a “big push” to spread federal R&D dollars “wider than the existing superstar cities to which they flow today,” albeit to places where they can nevertheless be employed productively...."

Robert Atkinson, Mark Muro and Jacob Whiton, "The Case for Growth Centers: How to spread tech innovation across America," Brookings and Information and Technology & Innovation Foundation, December 2019, https://www.brookings.edu/wp-content/uploads/2019/12/Full-Report-Growth-Centers_PDF_BrookingsMetro-BassCenter-ITIF.pdf

"....The innovation sector... encompass America’s 13 highest tech, highest-R&D industries. Selected from among the 50 advanced industries, the 13 innovation industries represent a cohort whose R&D expenditures exceed $20,000 per worker and have a STEM-worker share of 45%.....—comprised of 13 of the nation’s highest-tech, highest-R&D “advanced” industries—contributes inordinately to regional and U.S. prosperity. Its diffusion into new places would greatly benefit the nation’s well-being. However, far from diffusing, the sector has been concentrating in a short list of superstar metropolitan areas. Most notably, just five top innovation metro areas—Boston, San Francisco, San Jose, Seattle, and San Diego—accounted for more than 90% of the nation’s innovation-sector growth during the years 2005 to 2017. In this fashion, they have increased their share of the nation’s total innovation employment from 17.6% to 22.8% since 2005. In contrast, the bottom 90% of metro areas (343 of them) lost share..... Fully one-third of the nation’s innovation jobs now reside in just 16 counties, and more than half are concentrated in 41 counties....."

Note attached WSJ and Porter in NYT reporting on the research.

New NBER looks at the effect of immigration restrictions in the United States during the 1920's and finds that a) wages for us workers in areas receiving fewer immigrants did not rise. b) US workers responded by moving to (higher productivity presumably) urban areas replacing European immigrants and drove down native wages in urban areas c) farmers responded to higher labor cost by shifting towards more capital intensive agriculture. The writers suggests that the American experience during the 1920's implies that restrictions on immigration will result in sustainability such as automation and off-shoring before wages climb.

Effects A and B, areas getting fewer immigration did not see wages rise, and wages fell in urban areas as higher skilled labor moved in (which implies immigrants had been sedating wages)"....In the United States, the era of open immigration with Europe ended abruptly in the 1920s. A series of restrictive acts introduced immigration quotas that were particularly targeted at immigrants from Southern and Eastern Europe. The quotas effectively limited the annual number of immigrants admitted to the United States by more than 75 percent. Given the substantial reduction of immigrant labor, a simple model would predict that wages for the existing workforce would increase. Yet, we find that the occupation-based earnings of US-born workers in labor markets exposed to the quota policy fell after the border closure.We then document how the economy adjusted to the decline in immigrant workers to explain this puzzling result. Once the immigrant flow into US cities declined, US-born workers and unrestricted immigrants from Mexico and Canada started entering in larger numbers, replacing the immigrant workers at a nearly one-to-one rate. The new arrivals into cities were more skilled, on average, than the immigrants that they replaced, making them closer substitutes to the existing US-born workers...We find that workers who were unrestricted by the quota policy (i.e., US-born workers and immigrants from the Western Hemisphere, including Mexicans and Canadians) replaced the loss of immigrant labor in exposed urban areas nearly one-for-one after the border closure. These new arrivals were more skilled than the immigrants that they replaced, filling only 50 percent of low-skilled blue collar positions previously held by immigrants. The arrival of higher skilled artisans and tradesmen, who may have been closer substitutes for existing US-born workers than the immigrant workers they replaced, can help explain why the earnings of existing residents fell after the border closure...."

Effect C, shift to capital intensive agriculture, "... By contrast, in rural areas, the loss of immigrant workers encouraged land owners to invest in more farm capital and to shift away from labor-intensive crops, which in turn discouraged US-born workers from moving into affected rural areas. We emphasize that our estimates reveal the total effect of the immigration restriction policy for local labor markets after corresponding flows of labor and capital occurred. Thus, the “puzzle” of why the earnings of US-born workers fell in cities after the border closure, despite a decline in immigrant labor supply, can be understood as the combined effect of a decline in the (primarily low-skilled) immigrant workforce and a corresponding rise in (a more mixed-skill) US-born workforce....Table 8 provides suggestive evidence that farmers adapted to the loss of immigrant farm labor by shifting into more capital-intensive production, thereby reducing employment opportunities for domestic workers as well. We measure the share of cultivated land planted in labor-intensive (hay and corn) versus capital-intensive (wheat) cereals... We find that rural areas with more quota exposure were more likely to plant capital-intensive wheat and less likely to plant labor-intensive cereals after the policy. Farmers also shift away from the use of draft animals (horses and mules), which are direct substitutes for new gasoline-powered tractor technology. Consistent with a decline in the demand for farm labor, farm wages decline by around 3 percent after the border closure for a one percentage point shift in quota exposure. However, we see no effect of the shift away from laborintensive production on average farm values indicating that the quota system did not impede the profitability of farming....."

Ran Abramitzky, Philipp Ager, Leah Platt Boustan, Elior Cohen, Casper Hansen, "The Effects of Immigration on the Economy: Lessons from the 1920s Border Closure," National Bureau of Economic Research, December 2019, https://www.nber.org/papers/w26536

Implications, "... It is rare to find such drastic changes in immigration policy and so this historical episode has important lessons for contemporary policy. Substantially walling off the US economy to new immigration did open up some employment opportunities for US-born workers who moved to urban areas to take jobs previously held by immigrant workers. However, using immigration restriction to raise the earnings of US-born workers more broadly is unlikely to be effective given the many factors that can substitute for immigrant workers. In the early twentieth century, restricting immigration from Europe encouraged labor flows from Mexico and Canada into urban areas, and the investment in new capital in rural areas. Today, these sources of substitutability may be automation in the manufacturing sector or the off-shoring of high-skilled tasks like computer programming or legal services....."

Burkhauser uses a Full-Income Poverty Measure and finds that poverty rate based on President Johnson’s standards fell from 19.5 percent in 1963 to 2.3 percent in 2017. His FPMis “..matches President Johnson’s initial standard. Second,these initial 1963 thresholds must be held constant in real terms over time by updating them each year using an unbiased measure of inflation. Third, the metric for evaluating success should include the full array of anti-poverty programs utilized by the government, including in-kind transfers and transfers administered through the tax code. No existing poverty measure meets all three of these conditions. We create a poverty measure, which we refer to as the Full-Income Poverty Measure (FPM), which meets all three of these conditions…”

"... Based on our FPM, we find that the poverty rate fell from 19.5 percent in 1963 to 2.3 percent in 2017, suggesting that President Johnson’s War on Poverty as he defined it based on absolute 1960s living standards is largely over and a success.That should not be surprising in light of substantial improvements in full-income throughout the distribution and the creation and expansion of major safety net programs, with median income more than doubling between 1963 and 2017. It also would not have been surprising to President Johnson and his advisers, including Robert Lampman who predicted in 1971 that poverty based on these initial standards would be eliminated by 1980. The dramatic reduction in poverty by 2017 based on President Johnson’s standards suggests that policymakers should consider setting new poverty thresholds that reflect modern-day expectations for what it means to be impoverished. For example, one could consider the current 12.3 percent poverty rate under the OPM to be the appropriate poverty rate based on today’s living standards. But even if today’s standards are preferred, this should not discount the progress made in reducing poverty over the past half-century. Under this approach, Figure 9 shows a version of the FPM that is anchored to the 12.3 percent poverty rate under the OPM in 2017. By this standard, the poverty rate under the FPM fell from 66.2 percent in 1963 to 12.3 percent in 2017. Thus, 54 percent of the U.S. population has been lifted from poverty since 1963 based on this modern poverty standard. Hence, even if policymakers reconsider the standards for poverty, it will still be true that poverty using an absolute standard is far below where it stood in the 1960s at the start of President Johnson’s War on Poverty...."

Brad Setser points out a key fact about the global economy - the absence of any growth in China's imports of manufactures.

In short the world's fastest growing economy over the last six years isn't importing any more manufactures (in dollar terms) than it did six years ago. Or as he put it, "... It increasingly looks like China’s post global crisis import income elasticity is well below 1.That’s the internalization of supply chains—with more parts for electronics exports produced in China..."Import substitution working?

Brad offers a few hypothesis for what is driving this trend, though he notes there isn't a definitive answer:

1)Is it a function of the direct effect of Trump’s tariffs and China’s retaliation on trade with the United States, and the indirect effects of tariff uncertainty on investment in China’s tradables sector?
2)Is it a function of Yi Gang’s apparently successful effort to limit China’s stimulus to the minimum necessary—and China’s broader decision not to sacrifice deleveraging to the exigencies of the trade war?
3)Is it a function of China’s version of peak “autos”?
4)Is it a function of Xi’s campaign to “localize” production in China—a push symbolized by import substitution goals embedded in “Made in China 2025”?
5)Or is it something else?

Brad Setser, "China, More Than Trump, Has Been Driving the Slowdown in Global Trade," Council on Foreign Relations, December 10, 2019, https://www.cfr.org/blog/china-more-trump-has-been-driving-slowdown-global-trade

"... One of the key stylized facts about the global economy in 2019 is that everyone should know China’s trade surplus in manufactures has increased by just under a percentage point of China’s GDP on the back of falling imports (the overall surplus is also up, but by a bit less).Even with the fall in trade with the United States. Yet it is something that a lot of people haven’t quite internalized. The IMF’s 2019 assessment of China was built around the assumption that the 2018 fall in China’s trade surplus would be sustained.And most of the financial press—the Wall Street Journal, the Economist—and many investment banks were forecasting that 2019 would be the year when China’s current account moved into deficit. Yet China's current account surplus is on track to top $200 billion this year on the back of a rising trade surplus…It isn't just a direct function of the trade war. China’s exports to the United States are on pace to fall by about 75 billion this year. That automatically would reduce China’s imports by say $25 billion (fewer exports mean fewer imported inputs). China’s tariffs will reduce imports from the United States by something like $20 billion. That isn't enough to explain the broader fall in China’s imports. But it is hard to differentiate between the impact of domestic weakness on investment—weak auto demand and auto oversupply should be combining to reduce investment in new Chinese capacity—with the impact of trade uncertainty, which also would deter investment in new manufacturing in China. We do though know that China's investment has been weak and investment tends to have a disproportionate impact on trade, as the imported content of investment is higher than the imported content of consumption…. Imports of manufactures for China's own use (that is, excluding processing imports) have been sliding relative to China's economy for quite some time.This isn’t just something that I have noticed. The ECB and the IMF both were on the case back in 2016. As was the Banque de France, which was somewhat more direct than either the ECB or the IMF ("the recent trade deceleration is closely linked to the shift of China’s production towards domestic demand," updated graphs here). It increasingly looks like China’s post global crisis import income elasticity is well below 1. That’s the internalization of supply chains—with more parts for electronics exports produced in China. And, perhaps, that’s China’s industrial policy in action...."

Good explanation of the mechanics

Swanson pushes back on Phillippon's claims the US telecommunications industry is extracting rents - American per capita internet traffic is 43.8% higher than western Europe and 42.6% higher than Japan

"....In most of at advanced economies, consumers pay around $35 per month for broadband Internet connections. In the US, they pay almost double. How on earth did that happen?...It turned out that most of the comparisons were poorly conceived and executed. The studies mostly failed to adjust for quality, speed, demographics, population density, data usage, and/or proper price indexes. They often cherry picked some nations and left out inconvenient ones — Switzerland comes to mind....After reviewing the conflicting claims and doing our own analysis, we concluded that overall, “The US Leads the World In Broadband.” US speeds were fast, coverage was broad (although never broad enough!), prices were competitive, and Americans generated and consumed (both per capita and per user) more data than anyone (see nearby chart). The US actually generates three times more internet data traffic than Western Europe....In its 2019 mobile report, Ericsson estimated data traffic per smartphone and found that in 2018, the US generated 7.0 gigabytes (GB) per month, while Western Europe generated 6.7 GB per month. Ericsson projects that by 2024, US traffic per smartphone will grow to 39 GB per month and Western Europe will rise to 32 GB per month. If US prices are in fact high, they are not depressing smartphone uptake and data usage. Direct price comparisons are tricky. And it’s true that in some studies, US mobile prices appear high. In its most recent international report, the Federal Communications Commission found that, according to simple metrics, out of 28 advanced nations, US mobile prices ranked in the teens or twenties. But after making hedonic adjustments for speed and quality, the US ranked 12th and 10th, respectively. We hear that newer studies, with more advanced methodologies, are on the way...."