NBER News and Research Archive

Proposition 209 banned the use of racial preferences in admissions at California's public colleges and was followed by a 4.4 percent increase in minorities' graduation rates. Peter Arcidiacono, Esteban Aucejo, Patrick Coate, and Joseph Hotz find that Proposition 209 was associated with better matches between students and campuses, which explain about 20 percent of the overall graduation rate increase. They estimate that changes in the selectivity of enrolled students explain between 34 and 50 percent of the increase, while schools doing more to help retain students and see them graduate can explain between 30 and 46 percent.

The property tax is the least popular tax in the United States and the only major tax whose revenues have declined as a share of income, perhaps because it is so salient. Marika Cabral and Caroline Hoxby show that in areas in which the property tax is less salient -- because a larger proportion of property owners are having taxes paid from escrow accounts, rather than being directly involved in the process of writing checks -- property taxes are higher, but property tax revolts are less likely to occur.

Alberto Alesina, Stelios Michalopoulos, and Elias Papaioannou combine linguistic maps on the spatial distribution of groups within countries with satellite images of light density at night to construct estimates of inequality in well-being (and/or public goods provision) across ethnic lines for a large number of countries. They find that ethnic inequality is weakly correlated with the standard measures of income inequality and is strongly negatively correlated with per capita GDP across countries. When they analyze within-country, across-district variation in 17 Sub-Saharan countries, along with household survey data, they find that those from the same ethnic group report worse living conditions, lower levels of formal education, and inadequate access to basic public goods when they reside in districts characterized with a higher degree of ethnic group inequality.

Using unique data from a supplement to the Health and Retirement Study, Steven Haider and Kathleen McGarry find that parents appear to distribute college-based transfers relatively unequally across their children, typically contributing more to the education of older children than to their younger siblings. They find no evidence that parents use cash transfers later in life to offset these differences in college support.

Since NAFTA took effect on January 1, 1994, trade between NAFTA members has increased considerably. Lorenzo Caliendo and Fernando Parro find that Mexico had the largest increase in exports and imports, followed by the United States and Canada. They estimate that 93 percent of the increase in Mexico's total trade-over-GDP ratio was a consequence of tariff reductions attributed to NAFTA.

Between 1991 and 2011, gross interstate migration in the United States declined by about 50 percent. Greg Kaplan and Sam Schulhofer-Wohl suggest that this is because of a combination of two factors: 1) a reduction in the geographic specificity of returns to different types of skills -- that is, it now matters less where your job is than what your job is; and 2) an increase in workers' information about how much they will enjoy living in alternative locations -- it's easier to use information technology and less expensive to travel, so workers can learn about other locations before they move there.

The U.S. military retirement system underwent substantial changes in 1986 with the passage of the Military Retirement Reform Act, and again in 2000 when key provisions of that bill were repealed. Jeff Smith and Jim West study rates of officer retention under more and less generous pension benefit regimes brought about by those changes and find that reductions in benefits significantly reduce the probability of officers remaining in the military. They estimate that a 20-percent reduction in the generosity of retirement benefits has roughly the same effect on the probability of remaining on active duty as a 0.27 percentage point reduction in the unemployment rate, or a roughly 2 percent increase in the GDP growth rate.

With a sample of more than 1,100 retractions of scientific papers, Pierre Azoulay, Jeff Furman, Josh Krieger, and Fiona Murray study how closely related articles that were published prior to these retractions are affected by this taint of "false science." They find that related articles experience a lasting 5-to-10 percent decline in the rate at which they are cited, although citations in for-profit publications respond less to the retractions than citations in academic publications. The decline in citations of related papers is more severe when the retracted article involves fraud or misconduct, rather than an honest mistake.

Almost all 18 year-old males in Sweden who enlisted in the military between 1980 and 1994 were required to take a battery of cognitive tests in preparation for military service, but test dates were assigned randomly, so those of similar age and graduating class had accumulated slightly different days of schooling when tested. Magnus Carlsson, Gordon Dahl, and Dan-Olof Rooth find that an extra ten days of school instruction raises the cognitive scores of these young men on intelligence tests.

Hunt Allcott and Todd Rogers study the short-run and long-run effects of an energy conservation initiative that involved mailing "Home Energy Reports" featuring personalized feedback, social comparisons, and energy conservation information to households monthly or every few months. This program has been running continuously since October 2008, and includes one randomly-selected group of households that stopped receiving the reports after two years. When the authors analyze over 200 million observations of daily electricity use over a six-year period, they find a pattern of "action and backsliding": consumers reduce electricity use markedly within days of receiving each of their initial reports, but their responses decay relatively quickly. However, there is a durable treatment effect that remains: for the group that continuously receives reports over a four-year period, the effects continue to grow. Even in the group whose reports were discontinued after two years, the backsliding happens more slowly, indicating that they had already formed habits of conserving.