Every Monday, the National Bureau of Economic Research, a nonprofit organization made up of some of North America’s most respected economists, releases its latest batch of working papers. The papers aren’t peer-reviewed, so their conclusions are preliminary (and occasionally flat-out wrong). But they offer an early peek into some of the research that will shape economic thinking in the years ahead. Here are a few of this week’s most interesting papers.
Title: “Veterans’ Labor Force Participation: What Role Does the VA’s Disability Compensation Program Play?”
Authors: Courtney Coile, Mark Duggan, Audrey Guo
What they found: The growth of disability compensation programs may be contributing to lower labor-force participation among veterans.
Why it matters: Veterans are much less likely than nonveterans to be active in the labor force, and their participation has declined since 2000. (“Participation” includes anyone who is either employed or actively looking for work.) At the same time, the share of veterans receiving federal disability benefits has risen significantly, in part because of liberalized eligibility requirements that make it easier to qualify for the program. (The average size of disability payments has also been rising.) In this paper, the authors argue that the two trends are linked and that increased availability of benefits may be discouraging some veterans from participating in the labor force. They find that veterans’ labor-force participation has become more sensitive to the state of the economy, suggesting the veterans may be using disability compensation as a safety net during periods of high unemployment.
Key quote: “Other factors may have contributed to the decline in veterans’ labor force participation over time, and these merit further research. Our discussion has focused on the liberalization of medical eligibility criteria as an explanation for the growth in the [disability compensation] program. But it is important to acknowledge that the program may be growing for the right reasons if veterans are increasingly disabled over time, due to injuries sustained in recent conflicts in Afghanistan and Iraq or to the effects of Vietnam service that are now recognized as being connected to Agent Orange.”
Data they used: The Annual Social and Economic Supplement to the Current Population Survey, 1980-2014.
Title: “Racial Disparities in Savings Behavior for a Continuously Employed Cohort”
Authors: Kai Yuan Kuan, Mark R. Cullen, Sepideh Modrek
What they found: African-American and Hispanic workers are less likely to participate in their companies’ 401(k) plans and invest less when they do, contributing to the racial wealth gap.
Why it matters: The wealth gap between white and minority families is large and growing. This paper looks at one piece of that gap: 401(k) retirement savings. The authors study the savings behavior of nearly 10,000 employees of Alcoa, a large metals company with workers across the country. They find that black and Hispanic workers are less likely to participate in the company’s 401(k) plan than white workers of similar backgrounds. When they do participate, they contribute a smaller share of their income and invest more conservatively, limiting their returns. They are also more likely to withdraw money or borrow from their accounts, which reduces retirement savings. Together, those factors mean that black and Latino workers tend to have lower retirement account balances than otherwise similar white workers.
Key quote: “There are other important factors that we would have liked to include in our analyses. For example, household income is an important variable for which we have no measure. If income shocks are distributed unevenly across groups and if such shocks affect savings behavior, then they are likely to mediate the observed relationship. Another variable of interest would be a measure of institutional distrust. If the legacy of contentious race relations has made certain ethnicities, such as African Americans and Hispanics, less likely to trust or invest in established/formal sector financial institutions, then that too could mediate the relationships observed. While these important constructs are missing in our mediation analyses, the primary results that there are substantial differences in savings behavior by race still hold.”
Data they used: 401(k) saving and withdrawal behavior for Alcoa employees over eight years.
Authors: Atif R. Mian, Amir Sufi
What they found: Mortgage lending surged in low-income, less creditworthy areas of the U.S. between 2002 and 2005. But systemic differences between incomes reported on mortgage applications and incomes reported to the IRS indicate that much of this “subprime” lending was reliant on borrowers fraudulently overstating their income.
Why it matters: Between 2002 and 2005, there was a tsunami of money for prospective U.S. homebuyers. This surge of mortgage credit was strongest in less creditworthy, low-income areas. But some economists have argued that incomes of homebuyers were increasing in these areas. After all, by looking at income as reported on mortgage applications, the areas with lower credit scores seem to have robust growth of homebuyers’ income. But new research from Sufi and Mian — the authors of “House of Debt” who have written for FiveThirtyEight — confirms that, no, economic improvement wasn’t behind these improving income numbers. It was fraud. Specifically, the fraud of homebuyers overstating income. Using ZIP code-level income statistics, the authors identify the highest fraud areas by comparing income as reported to the IRS versus income stated on mortgage applications. These areas experienced the strongest surge of mortgage credit while having lower income and credit scores; the areas were also where proven mortgage fraud was most common.
Key quote: “Looking past 2005, we find that zip codes with high overstatement perform terribly. Default rates in these zip codes skyrocketed from 2005 to 2007. Using a longer horizon, the zip codes with high overstatement from 2002 to 2005 experienced lower IRS income and wage growth from 2005 to 2012. They also saw lower median household income growth from 2000 to 2010 according to the Census. Finally, there was a jump in both poverty and unemployment rates from 2000 to 2010. Recall, these zip codes already had higher poverty and unemployment rates in 2000, and they increased further through 2010. These patterns are inconsistent with gentrification, but consistent with fraudulent income overstatement on mortgage applications.”
Data they used: IRS data for ZIP code-level income; database of mortgage fraud from Interthinx.
Title: “Childhood Medicaid Coverage and Later Life Health Care Utilization”
Authors: Laura R. Wherry, Sarah Miller, Robert Kaestner, Bruce D. Meyer
What they found: Low-income African-Americans who had Medicaid health insurance in childhood had fewer hospitalizations in adulthood, possibly offsetting 3 to 5 percent of the program’s annual cost. The effects were especially pronounced for those with chronic diseases and those from low-income areas.
Why it matters: One argument for publicly subsidized health insurance is that if people have coverage early in life, they will demand less treatment later and potentially offset some of the cost. Yet it’s tricky to empirically test whether this is true. One study from the Oregon Medicaid experiment concluded that expansion of insurance coverage did not lead to less utilization, in the form of hospitalizations or emergency department visits. But that study was over a relatively small time span. The authors of this paper take a broader view: They look at the expansion of Medicaid in the early 1980s. Specifically, they look at children born just before and just after Sept. 30, 1983. That was the cutoff set for Medicaid eligibility — those born after that date got more years of coverage in their lifetime than those born just before. African-American children who got health insurance did seem to have less health care utilization later in life; they had 8 to 13 percent fewer hospitalizations and 3 to 4 percent fewer emergency department visits by age 25.
Key quote: “Our results highlight the importance of evaluating these programs over a longer time period. Indeed, we find no impact of Medicaid coverage in our analysis of the ‘immediate’ effect at age 15, but do find effects later in life at age 25. These findings suggest that the benefits of insurance may only materialize over a long horizon.”
Data they used: Hospitalization data from state health agencies.