ABSTRACT. The African Swine Fever has caused a significant reduction in the Chinese swine herd. This has lowered the amount of Chinese imports of U.S. whey, a key input into swine production as piglet food. We use a partial equilibrium model of both intermediate and final goods markets to simulate the impacts of ASF and Chinese tariffs on Chinese imports of U.S. whey. We find large price impacts, and significant negative impacts to Chinese imports of U.S. whey, as remaining swine producers shift to other whey suppliers. We then run counterfactual analyses to analyze industry consolidation, where smaller pig farms close down from ASF and larger production plants dominate the market. We find that Chinese imports of U.S. whey rebound because larger production plants have a higher whey consumption rate.

ABSTRACT. As the country’s largest anti-poverty measure, the Earned Income Tax Credit (EITC) has effectively lifted millions of Americans out of poverty. Because credit amount is tied to family size, the EITC’s structure could inadvertently encourage childbearing. Previous literature considering average fertility effects finds mixed, but generally negative associations between the credit increase and fertility. However, no other work has considered heterogeneous effects due to preference. I find that among groups that have a revealed preference for larger family sizes, the 1994 expansion is associated with as much as a 15.5 percentage point increase in childbearing. Because EITC recipients qualify for the credit based on their low incomes, and because the credit might not completely offset the cost of childrearing, fertility responses to the credit’s expansion could result in an outcome that opposes the credit’s antipoverty intent.

ABSTRACT. Using data on full-time wage and salary workers from the 2017–2018 American Time Use Survey Leave and Job Flexibilities Supplement, this paper examines the characteristics of teleworkers, the effects of teleworking on wages, and whether differences in time-use patterns suggest that teleworking allows workers to experience positive work-life balance. Teleworkers earn more than office workers, on average. However, when we use econometric techniques that relate selection on observables to selection on unobservables to estimate bounds on the true causal effect of telework on wages, we cannot reject the hypothesis that the wages of teleworkers and office workers are the same. In terms of work-life balance, we find that teleworkers spend more time on home production and leisure activities and spend more time with their family on days they work at home. They can spend more time on these activities because they spend less time on commuting and grooming activities.

ABSTRACT. Brazil and Argentina are countries with two of the world’s most competitive agricultural sectors and leading world producers and exporters of oilseeds, grains and livestock. In more recent times, macroeconomic reforms and policies have become an engine of growth and development for their agriculture. Brazil’s macroeconomic and agricultural policies have played an important role in Brazil’s emergence as one of the top exporters of agricultural products, including soybeans, corn, cotton, sugar, coffee, orange juice, and meat. Fluctuations in exchange rates, macroeconomic growth, and policies encouraging credit and investment impact farm production and Brazil’s propensity to export agricultural commodities and to expand the country’s agricultural frontier.

Argentina is currently the world’s leading exporter of soybean oil and soybean meal. In addition, Argentina ranks third behind the United States and Brazil as a producer and exporter of soybeans and as an exporter of corn, and it is the world’s sixth largest wheat exporter. Argentina’s economic history includes multiple periods of economic instability featuring both currency devaluation and hyperinflation. The Argentine Government has introduced significant changes to the country’s exchange-rate policy. Other economic policy measures taken include the elimination of certain taxes on exports, fuel, and financial transactions; and lower import tariffs on fertilizers, herbicides, pesticides, and irrigation equipment. These other policy measures have provided a significant boost for agricultural growth. Expansion of the country’s crop frontier, mainly in the northeast and northwest regions where soybeans are the main crop, was accompanied by increased productivity for major field crops.

Given the large size and continued growth of the Brazilian and Argentine agricultural sectors, their share of world markets and impact on global supply chains affect U.S. sales of agricultural goods—including soybeans, grains, cotton, and meats—in foreign markets where the United States and these two South American countries directly compete for sales. Thus, it is strategic to understand the evolution of policies that Brazil and Argentina have used to influence their agricultural sector and their economy as a whole.

To understand the forces behind this development, this study examines the effects of changing macroeconomic conditions and agricultural policies on Brazilian and Argentine agricultural production and trade. This study explores the role of macroeconomic variables by evaluating the potential impact of a more rapid currency depreciation and stronger sustained growth on agriculture. The models of Brazilian and Argentine agriculture used to obtain USDA’s long-term baseline agricultural projections are used in this study to assess the impact of macroeconomic and agricultural policies on agricultural production and trade over 10 years for grains, oilseeds and products, cotton, livestock products, sugarcane, sugar, and ethanol. The models are combined with 40 other country/regional models to generate global commodity supply, demand, trade, and equilibrium prices for 2020-29.

ABSTRACT. Following a lengthy period of population decline ending in the late 1990s, the District of Columbia’s net population has increased by over 100,000 people between 2000 and 2018. Against this backdrop, the District has found itself in the midst of a housing affordability crisis which is impacting workers across the income spectrum.

Several studies have shown that undersupply contributes to strong housing price increases. This paper seeks to answer two questions concerning housing in the District of Columbia. First, what has been the impact of the substantial increase in the supply of rental housing in recent years on average apartment rents in the city. And second, what is the estimated impact of an additional planned substantial increase in rental housing units on apartment rents in 2025.

We answer these two questions by way of three economic model estimations. The first determines the effect of the actual increase in the number of apartment units in the city on rents for years 2000 to 2018. The second estimation simulates the average apartment rents in 2018 if the city had not doubled the rate of new additions beginning in 2012. And, the third estimation simulates average rents in 2025 under the Mayor’s recent plan to nearly double the rate of new additions again beginning in 2020.

Apartment data was sourced from CoStar, a real estate information firm. Data is a quarterly time series with citywide data from class A and class B mid- and high-rise apartment building with 20 or more units. It includes buildings within the District of Columbia, built as early as 1960. The data spans the period 2000 Q1 to 2018 Q1. Additional data was sourced from the DC Government’s Office of Revenue Analysis (ORA) and the Bureau of Labor Statistics.

This study finds that if the marked delivery rate of new apartment units in recent years had not occurred, average city apartment rents would have likely been 5.94 percent higher in 2018. That is, the average citywide monthly apartment rent is likely to have been $3,207 in 2018 rather than the actual average of $3,030. This study also finds that if the planned increase in new apartment units (under the Mayor’s 2019 Housing Initiative) does not occur, average city apartment rents are estimated to be 5.24 percent higher in 2025. That is, the average citywide monthly apartment rent is likely to be $3,261 in 2025 instead of $3,090 under the Initiative.

ABSTRACT. Theory suggests that when communication between different levels of government is impaired or there are political pressures, central planners may not be able to maximize social welfare and simply allocate projects to all interested local officials (universalism). We provide novel empirical support for these mechanisms using a recent place-based policy allowing governors to designate certain census tracts as opportunity zones (OZs), conferring tax incentives for investment. We collect communications between mayors (local officials) and governors (central planners) nominating which tracts to select. With these data, we find: First, even though mayors' nominations are by far the most important determinant of governors' OZ selections, nominations appear to be worthless for maximizing the policy's effect. This finding is consistent with an impairment of communication and governors practicing universalism. Second, governors rubber stamp nominations more when political pressures are greater. Finally, rubber stamping resulted in 40\% of selected OZs being in areas that were predictably expected to bring no additional investments.

ABSTRACT. The European Commission years ago adopted a policy of encouraging the substitution of motor carrier haulage of freight with rail and water carrier haulage, as part of its “green” agenda of reducing fuel consumption, emission of pollutants, carbon intensity, and road congestion. Regarding railway freight in particular, one policy tool that the Commission has emphasized for this purpose is the restructuring of the rail sectors of member countries through the creation of competition for the incumbents by new train-operating companies (TOC’s) – on its face a less obvious policy choice than alternatives such as Pigouvian pricing measures or infrastructure subsidies. This paper focuses on one important commodity group – grain – in three EC member states and one non-member state – Poland, the Czech Republic, Slovakia, and Ukraine – to examine the degree to which increased rail competition has been associated with increases in rail’s modal share, and more broadly to learn what appear to be the binding constraints to increases in rail’s share. Such constraints seem more closely related to shortages in infrastructure capacity than to a lack of competition among TOC’s. This suggests that other “models” of railway restructuring may be more effective in easing this constraint.

ABSTRACT. Competition law enforcers around the world have traditionally exhibited a strong preference for "structural" remedies over "behavioral" or "conduct" remedies when seeking to prevent the anticompetitive effects of an otherwise welfare-enhancing horizontal merger. However, ex post examinations by the US FTC and the EC have found that many structural remedies accepted by enforcers have proved ineffective in preserving post-merger competition. This paper discusses recent attempts by the US agencies to craft remedies that address these concerns, including non-traditional remedies accepted in the Bayer/Monsanto, Twin America, and USAir/American Air cases, as well as an arguable missed opportunity for enforcers seeking to impose post-merger access terms to learn from the extensive experience and insights of regulatory economics in this area.

ABSTRACT. This paper evaluates how a redesigned processing system, incorporating changes in data extraction and imputation, affected estimates of medical expenditures in the Current Population Survey Annual Social and Economic Supplement (CPS ASEC). Using data from the 2018 CPS ASEC Production and Bridge files, we compare family medical expenditures across two processing methods for the same surveyed respondent information. The first method, a legacy procedure used in the 2018 CPS ASEC Production file, imputed health insurance coverage and medical expenditures separately and included limited information about a family’s income. The second method, an updated procedure used in the 2018 CPS ASEC Bridge file, introduced a joint imputation for health insurance coverage and medical expenditures and considered income when imputing both topics. Findings suggest that the updated processing system did not change median family medical expenditures but did reduce the percentage of families with a high medical burden (those for whom medical expenses are at least 10% of their family income.) Furthermore, the processing system changes affected the association between family income and medical expenditures. Relative to legacy processing procedures, updated processing procedures showed that families with high incomes had higher medical expenditures and families with low incomes had lower medical expenditures.

ABSTRACT. High quality early childhood education programs provide young children with the knowledge, tools and resources they need to be successful in school and in life. Early childhood and pre-kindergarten education for disadvantaged children can significantly increase their cognitive abilities and lead to increases in educational achievement and school success in the long-term while reducing future participation in criminal and delinquent activity. Unfortunately, many low-income families cannot afford the cost of quality early childhood education programs. Therefore, children from areas with higher poverty levels tend to have lower enrollment percentages in early childhood care, lower test scores, high school graduation rates and college attendance and graduation rates than children in wealthier communities. Family income and poverty status influence the cognitive development and behavior of children, even after accounting for other differences between low-income and high-income families including family structure and parental education. In 38 states, the cost of infant care is greater than 10 percent of the state’s median income for a two-parent family. According to the National Association of Child Care Resource and Referral Agencies, the average cost of center-based day care in the United States is $11,666 annually or $972 per month. Head Start and the expansion of state-funded early education programs since the 1990s have significantly increased access to preschool for 3- and 4-year olds, though many 3-year-olds continue to be left out contributing to disparities in educational achievement. While child care subsidies are offered by state and federal organizations, there is not enough funding to cover all the families who need it. Further differences exist with respect to the quality of early childhood education programs. This study examines common causes of the racial disparity in early childhood educational opportunities. It then provides recommendations for funding early childhood care, increasing the number of teachers of color, and providing innovative learning methods for children and caretakers.

ABSTRACT. This paper studies the government spending multiplier when fiscal policy has supply-side effects. We show that in normal times, during which the nominal interest rate is free to adjust, the presence of supply-side effects makes fiscal policy more effective. However, in liquidity traps, the supply-side effects make fiscal policy less effective. We also show that fiscal policy is not necessarily more effective in liquidity traps. One implication of our findings is that spending on research and development and cutting corporate taxes should be avoided in liquidity traps.

ABSTRACT. Fundamental economic measures such as worker quality and income inequality are typically based on self-reported cash income. Over the past few decades, the value of health benefits earned by workers has consistently grown faster than cash income. As a result, those fundamental economic measures may have become increasingly mismeasured. Our paper uses health insurance claims data from MarketScan, health insurance coverage data from the Current Population Survey and other sources to estimate the value of health benefits by age, gender, education, and industry.

Our preliminary results suggest that older workers and less educated workers receive a much larger share of their compensation in the form of health benefits. As a result, the existing measures of worker quality underestimate the returns to experience and overestimate the returns to education. In future research, we plan to use our new measures of employee compensation to recalculate labor quality and productivity growth over time.

ABSTRACT. This paper investigates a new explanation for the long-run decline of sectoral-switching in the US--the increased prevalence of referral-networks in job-search. Current Population Survey (CPS) data show a rise in the use of referral-networks concurrent with the decline in sectoral-switching. Estimates for various industry aggregations show using referral-networks reduces the probability of switching sectors by 1.2%-3% percentage points. To better understand welfare implications, a sectoral-switching model is constructed using a search-and-matching framework with labor market referrals. The estimated model predicts a referral-switching elasticity of about -.12, which implies the rising use of referrals overtime can explain about 20% of the decline in switching. Welfare experiments suggest the decline in sectoral-switching is related to improved matching efficiency.

ABSTRACT. South Asian economies are major sources of international migrant labor and have experienced positive development impacts from the process. However, these countries have some of the highest migration costs in the world, and there is limited understanding of how these costs affect household welfare and behavior. This paper focuses on Nepal and Pakistan, to address this gap. In these countries, households that receive remittances spend 11% percent more on education and health than non-remittance receiving households. However, the high costs of migration tend to dampen both the propensity to remit and the amount remitted: a 1% increase in recruitment cost may decrease remittances by 0.05-0.15%. Specific interventions, such as the elimination of visa costs for Pakistani migrants could increase remittances by 2.9-5%. Policies that reduce migration costs may have large and important impacts on household investment behavior, and therefore the economic growth and development of Nepal and Pakistan.

ABSTRACT. In 2017, the Federal Deposit Insurance Corporation (FDIC) reported that 20.5 million Americans lived in households that lacked a checking or savings account, with many of these individuals living in low-income households. Without access to a checking or savings account, these individuals lack the ability to store and save their income in secure accounts, driving them to costly short-term alternatives which cut into their already low-incomes. To better understand the link between being unbanked and in poverty, we first exploit the survey design of the Current Population Survey and link respondents in the Annual Social and Economic Supplement (ASEC), the official source of poverty statistics for the United States, to the bi-annual FDIC Banked and Underbanked Supplement from 2009 to 2017. We use the linkage to understand the banking landscape for poor households, examining how households operate without formal relationships with banks and the cost of alternative financial services, such as check cashers or payday lenders. We then use data from the Survey of Income and Program Participation (SIPP) from 1993-2016 to show how historical unbanked rates have diverged by demographic characteristics and income-to-poverty ratios with a particular focus on race, households with children, and program participation. This is important to determining whether unbanked rates have improved or worsened over time for these groups of interest.

ABSTRACT. Aspiration is a fundamental determinant of one's effort and investments. Due to its consequences for individuals' future outcomes, understanding the process of aspirations formation helps to inform public policies. This work asks whether peers play a role in such a process. I use novel data on Brazilian students' networks, matched with administrative data, and investigate whether students' college aspirations spills over to their friends. The employed methodology acknowledges that social cliques are formed endogenously and addresses this challenge by modeling friendship formation based on homophily in predetermined characteristics and on students' as-good-as-random chances of interaction. Using the predicted adjacency matrix, I explore network structures and use predicted friends of friends' characteristics as instruments for friends' aspirations. The results show evidence of positive, significant, and quite large peer effects on aspiration - an extra friend aspiring to go to college increases on average 9.9\% the likelihood that a student will also aspire to it. Such an impact is driven mostly by boys. While peers' aspirations do not seem to influence students' proficiency, an extra aspiring friend decreases the likelihood of dropping out of high school and increases class attendance in reading. Compliance with social norms seems to play a role in explaining such an impact.

ABSTRACT. The main objective of the paper is to investigate the relationship between macroeconomic variables and economic growth in India. The research has used the macroeconomic data from the world bank’s database over 15 years from 2001 to 2015. This paper has employed correlation coefficient and multiple linear regression models to analyze the relationship among five macroeconomic variables GDP growth rate, foreign direct investment inflow, foreign direct investment outflow, inflation, and real interest rate. The result of the correlation coefficient establishes a negative and insignificant association between GDP growth rate, foreign direct investment net inflows and real interest rate. However, foreign direct investment net outflows and inflation are positively but insignificantly associated with economic growth. The regression analysis results indicate that inflation, interest rate, foreign direct investment inflow, and foreign direct investment outflow demonstrate a significant impact on the GDP growth rate of India. The result of the research recommends that policymakers should adopt a contractionary monetary policy to control the inflation and achieve long-term sustainable economic growth.

ABSTRACT. Los Angeles County's homeless population has increased by approximately 40 percent in the past five years. While county voters have supported the goal by approving billions of dollars in bonds that would provide tens of thousands of affordable housing units and services for the homeless, there remains a substantial gap in affordable housing for the homeless and low-income individuals who are at risk of homelessness, drive by fears and stigma in local communities. I investigate the effect of such housing sites on street homelessness, crime, and property values. I construct a comprehensive data that geocodes the locations of all designated homeless housing sites in Los Angeles County. Using spatial and time variation in homeless housing sites, I estimate the exposure of a community to designated homeless housing sites over time and use changes in this exposure to recover the causal relationship. I find that communities that had an increase in homeless housing in their boundaries and vicinity experience a sizable decline in homeless encampments, overall crime, and homeless-related crimes, and that housing values in these communities had increased.

ABSTRACT. The Current Population Survey Annual Social and Economic Supplement (CPS ASEC) is the source of the nation’s official household income and poverty statistics. In 2012, the CPS ASEC showed that median household income was $33,800 for householders aged 65 and over and the poverty rate was 9.1 percent for persons aged 65 and over. When we instead use an extensive array of administrative income records linked to the same CPS ASEC sample, we find that median household income was $44,400 (30 percent higher) and the poverty rate was just 6.9 percent. We demonstrate that large differences between survey and administrative record estimates are present within most demographic subgroups and are not easily explained by survey design features or processes such as imputation. Further, we show that the discrepancy is mainly attributable to underreporting of retirement income from defined benefit pensions and retirement account withdrawals. Using archived survey and administrative record data, we extend our analysis back to 1990 and provide evidence of underreporting from an earlier period. We also document a growing divergence over time between the two measures of median income which is in turn driven by the growth in retirement income underreporting. Turning to synthetic cohort analysis, we show that in recent years, most households do not experience substantial declines in total incomes upon retirement or any increases in poverty rates. Our results have important implications for assessing the relative value of different sources of income available to older Americans, including income from the nation’s largest retirement program, Social Security. We caution, however, that our findings apply to the population aged 65 and over in 2012 and cannot easily be extrapolated to future retirees.

ABSTRACT. Declining survey response rates in federal surveys has led to concerns of increasing nonresponse bias in key government statistics. A potential solution is to leverage administrative data from federal agencies when constructing survey weights. This project performs initial research on incorporating administrative data into the weighting algorithm for the Survey of Income and Program Participation (SIPP). Specifically, we match income data from IRS tax forms and demographic data from the Social Security Administration and the Decennial Census to both respondents and nonrespondents. We then use this matched data in the household nonresponse adjustment of the SIPP weighting algorithm, which adjusts the weights of respondents to account for differential nonresponse rates among subpopulations and reduce nonresponse bias in survey estimates. We show how these new weights affect estimates of wealth, income, poverty, health insurance coverage, and participation in government assistance programs and their impact on nonresponse bias compared to the traditional weights. Overall, the new weights are associated with a small increase in estimated economic wellbeing.

ABSTRACT. This paper evaluates trends in poverty using a new series of household-level data constructed from IRS tax return data. This series, from 2007 to 2017, is an extension of the Tax Household Sample from Larrimore, Mortenson, and Splinter (forthcoming), which uses addresses on tax filings to aggregate tax returns to the household level. This approach, which accounts for both filer and non-filer income, overcomes the traditional limitations that have prevented researchers from using tax records to analyze the lower tail of the income distribution. Recognizing that public assistance programs are increasingly administered through the tax code, including the Earned Income Tax Credit and Child Tax Credits, IRS data are uniquely suited to consider the effect of tax-based programs on poverty trends. In addition to analyzing trends in poverty, the panel nature of the data allow us to investigate the income dynamics of those in or near poverty, including the characteristics of individuals who experience transient and persistent poverty spells. In contrast to the Current Population Survey data which can only observe individuals for up to two years, the IRS data allow us to track the same individuals over a much longer period. The tax data-based panel remains nationally representative in each year, and allows for entry and exit due to birth, death, emigration, or otherwise falling out of the IRS’ vantage. This allows us to estimate the share of individuals in poverty at any point over 3-, 5-, 10-year periods; the share in poverty for one year relative to the share experiencing extended periods in poverty; and the characteristics of individuals experiencing transitory and persistent poverty.

ABSTRACT. Evaluations of the EITC, including its antipoverty effectiveness, are based on simulated EITC benefits using either the Census Bureau’s tax module or from external tax simulators such as the National Bureau of Economic Research’s TAXSIM or Jon Bakija’s model. Each simulator utilizes model-based assumptions on who is and who is not eligible for the EITC, and conditional on eligibility, assumes that participation is 100 percent. However, recent evidence suggests that takeup of the EITC is considerably less than 100 percent, and thus claims regarding the impact of the program on measures of poverty may be overstated. We use data from the Current Population Survey Annual Social and Economic Supplement (CPS ASEC) linked to IRS tax data on the EITC to compare the distribution of EITC benefits from three tax simulation modules to administrative tax records. We find that significantly more actual EITC payments flow to childless tax units than predicted by the tax simulators, and to those whose family income places then well above official poverty thresholds. However, actual EITC payments appear to be target efficient at the individual tax unit level, whether correctly paid or not. We then compare the antipoverty impact of the EITC across the survey and administrative tax measures of EITC benefits. We find that in the full CPS ASEC the tax simulators overestimate the antipoverty effects of the EITC by about 1.8 million persons in a typical year. Restricting to a harmonized sample of filers, we find that the antipoverty estimates derived from the TAXSIM and Bakija models align more closely to actual EITC payments compared to the CPS, suggesting a discrepancy in assignment of tax filers between the tax simulators.

ABSTRACT. Social norms can have a persistent influence on outcomes. Since the end of the second World War, men have been most households’ primary breadwinners in the developed world, and US data from the late twentieth century suggests violation of this norm stresses partnerships. Is this still true? We examine whether female breadwinning makes partnerships less healthy or less stable using more recent US and Australian data. We find a much more modest association in OLS models for mixed-gender couples in both countries, and one that primarily affects young people in cohabiting partnerships and men in less educated partnerships. Transitions into female breadwinning are problematic mainly for cohabiting couples. We interpret these results as partly reflective of changing social norms and partly a consequence of mating market dynamics, presenting initial evidence that cohabiting women in Australia who out-earn their partners subsequently re-partner with men who have higher earnings relative to themselves.

ABSTRACT. There has been increasing interest in the distribution of income and in income mobility over the past decade. However, most of the recent literature on measuring income inequality and mobility has focused on the national distribution of income, with very little work on changes in the distribution of income within states, counties or other subnational entities. We believe this is an important blindspot in the literature, which we hope to remedy in this project. By using the universe of tax returns from 2000 to present, combined with sociodemographic information from the American Community Survey and the 2000 and 2010 Decennial Census, we are able to produce aggregate income statistics broken down by demographic characteristics for subnational geographies. In addition to producing improved versions of standard distributional statistics, we introduce two novel measures: first, a regional divergence index which allows for inference both about between-region and within-region inequality in incomes, as well as a new series of statistics on the persistence of incomes at the top of the distribution, extending recent work on income mobility. For the current prototype statistics, our focus is at the state-level; in future work we plan to extend these statistics to commuting zone, county, and tract level measures.

ABSTRACT. Business growth has become increasingly geographically concentrated. Previous research has shown that in the recoveries following recent economic recessions a small number of counties accounted for an increasing share of growth in the number of business establishments. This analysis uses data from the Quarterly Census of Employment and Wages to explore changes in the geographic concentration of growth between 1976 and 2019. A measure of concentration suitable for both expansionary and contractionary periods is introduced. Concentration is shown to have increased over the time examined, and some significant deviations from that trend are identified. Relationships are explored between the rise in concentration and changes in the geographic distributions of the population and of business lending.

ABSTRACT. This paper studies the short- and long-run effects of each U.S. recession since 1973 on local economic activity. We analyze how economic activity evolves across local areas that are differentially affected by national recessions. For each recession, we find that employment, population, employment-to-population ratios, and earnings per capita experience persistent declines for at least a decade after recession’s end. While transfers also remain elevated for a decade or more, they are insufficient to fully offset earnings losses, leading to long-term declines in per-capita income as well. Changes in the composition of workers explain less than half of these persistent effects.

ABSTRACT. It is an open question how the cost of an adult child affects parental labor supply. College expenses are important for parents of adult children. We exploit changes in college costs after the roll-out of nine generous merit aid programs from 1993 to 2004 to analyze the difference in labor supply of mothers of college-age children before and after the implementation of these programs. Mothers of college-age children decreased their annual hours of work after the start of merit aid, while fathers did not adjust their labor supply. There is no strong evidence that mothers changed their employment status, as most of the decrease in hours of work happened among employed mothers. Mothers of college-going children are entirely responsible for the decline in hours of work, where mothers of children who did not go to college experienced no change in hours of work. A 10 percent increase in spending on merit aid programs per undergraduate student leads to a 1.3 percent decline in hours of work among mothers of college-going children. The decline in labor supply is mainly due to adjustments among married, highly educated, and white mothers.

ABSTRACT. Why do postsecondary institutions use merit-based admissions? These strategies allow institutions to select applicants that will add value along certain, but they also often rely on imperfect tools and measures, and may reduce the diversity of new cohorts. In recent years, there has been new interest in the role of postsecondary admissions as a tool of social policy, with the ability to affect socioeconomic mobility in particular. There is little evidence, though, on how merit-based admissions affects the composition of incoming cohorts and their later outcomes. In this paper I study a state-wide change in admissions rules. In 1991, the California Community Colleges restricted admissions at associate degree in nursing (ADN) programs, allowing only “non-evaluative” admissions policies such as waitlists and random lotteries. In 2007, the state legislature reversed this policy and allowed colleges to select students based on prior grades, work history, essays, and personal references. I leverage the incremental adoption of these new admissions rules after 2007 across the state's many ADN programs to provide estimates of their effect on the academic preparation, demographic composition, and academic outcomes of new ADN cohorts. The identification strategy relies on variation in the timing of the adoption of the new admissions across colleges in the state. I use individual-level administrative data on all California community college students between 1992 and 2017. I first study whether the policy brought in cohorts with higher past academic performance. I find that a switch to evaluative methods led to particularly large increases in the biology GPA of incoming students, as well as decreases in the number of remedial courses students had taken. I then examine the demographic composition of new cohorts. I find that colleges switching to evaluative admissions increased their enrollment of white students while decreasing their share of Latino students. I also find that the mean age of students decreased. Finally, I examine academic outcomes. I find no evidence of improvements in first-year GPA, completion rates, time to degree, or licensing exam pass rates for students in cohorts accepted under the new admissions regimes. I do find a large decrease in the number of years between when a student first started taking community college coursework, including prerequisites, and first enrolled in an ADN program. Taken together, I find that evaluative admissions for ADN programs brought in better-prepared students, as intended, but reduced the racial and ethnic diversity of incoming cohorts and did not have any meaningful change in academic outcomes. However, the new admissions dramatically decreased the amount of time students had to wait before entering the cohorts. This suggests that certain aspects of the new systems, like doing away with waitlists, were more efficient relative to the old systems. I conclude the paper by proposing ways that these programs, and others, could redesign their admissions in order to increase diversity without hurting academic outcomes.

ABSTRACT. High-achieving, low-income students attend selective colleges at far lower rates than upper-income students with similar achievement. Behavioral biases, intensified by complexity and uncertainty in the admissions and aid process, may explain this gap. In a large-scale experiment we test an early commitment of free tuition at a flagship university. The intervention did not increase aid: rather, students were guaranteed before application the same grant aid that they would qualify for in expectation after admission. The offer substantially increased application (67 percent vs 26 percent) and enrollment rates (27 percent vs 12 percent). The results suggest that uncertainty, present bias, and loss aversion loom large in students’ college decisions.

ABSTRACT. Policies and practices throughout the educational pipeline harm the educational attainment of black Americans, who are less likely than whites to obtain a bachelor’s degree. This report draws on detailed student-level data from Virginia and Connecticut to help institutional leaders and state policymakers better understand equity gaps in college completion rates at two- and four-year colleges. We examine the role played by factors including academic preparation, financial circumstances, access to individual institutions, segregation across colleges, and other structural factors. We take a statewide view concerning the longest-standing historical inequity in higher education systems—and American society more broadly—that harm the outcomes of black students relative to white students. We provide a statistical decomposition of the black-white gaps in graduation rates statewide into different components, including academic preparation, financial circumstances, college segregation, and other structural factors. Consider the nearly twenty percentage point black-white graduation rate gap at four-year colleges in Virginia. Equalizing the average difference in SAT scores and high school GPA across races would close this gap by 42 percent. Differences in student financial circumstances (measured using financial aid receipt) explain another 21 percent of the gap. In other words, slightly more than half of the racial disparity in success at the college level stem from pre-college racial inequality in college readiness and financial strain. Racial inequity in access to high quality colleges also plays a sizable role in the statewide graduation gap. Colleges do not enroll students of different races in equal proportions, especially in the 4-year sector. Our estimates suggest that, were black students to attend each college in the state at similar rates as whites, and black students were to receive the same benefit from attending these schools that white students do, the gap would close by an additional 26 percent. This finding is driven by white students being much more likely to attend higher-quality colleges, controlling for preparedness and financial need. Together with the estimates above, our results suggest that about 90 percent of the racial gap in two-year graduation rates in Virginia is associated with racial gaps in college readiness, financial strain, and access to high-quality colleges. We find similar results for four-year colleges in Connecticut and two-year colleges in both states, although college segregation is not a significant factor in the two-year sector. These approaches to understanding equity gaps in college graduation rates could be incorporated into equity audits conducted by both institutional leaders and state policymakers such as those currently being considered. Understanding the roles played by factors both inside and outside of higher education is critical to designing higher education policies that equitably serve all students.

ABSTRACT. This paper shows how measured investment and measured GDP may be distorted when investment purchases are bundled together with intermediate input purchases. The paper then illustrates the distortion with an example from the vehicle industry where measured capital investment is overestimated and an example from the vehicle industry where measured capital investment is underestimated. The paper also provides additional examples to show that the measurement problems associated with bundling apply to a broad range of industries and asset categories. We discuss why it is difficult to estimate the net impact of bundling on measured investment and measured GDP.

ABSTRACT. To what extent does states’ resource dependency affect inequality? The literature attempting to address this question using aggregated country-level data or subnational-level data from resource-rich countries yields highly inconclusive evidence. This paper contributes to the ongoing debate by estimating the causal impact of increased resource-based government revenue on inequality at the sub-national level using data on the 50 states of the US. Income and wealth inequality have been a subject of heated debate in the US as the economic gap between the top and bottom percentiles of income distribution has increased sharply over the past five decades. Yet, we observe that resource-dependent states have lower inequality (as measured by the share of income captured by the top 1%) than the national average. We combine state level data on government revenue from natural resources with detailed micro-level data on income and consumption from the Consumer Expenditure Survey to test the hypothesis that an increase in resource-based government revenue reduces consumption and income inequality. To address the issue of endogeneity of state revenue from natural resources with respect to inequality, we instrument for state resource revenue with reserves of natural resources.

ABSTRACT. Efforts to incentivize the entry of new and beginning farmers into agriculture have been ongoing since at least the 1990s, with funds and the number of programs meant to support beginning and socially disadvantaged farmers increasing through the 2002, 2008, 2014, and the most recent 2018 farm bills. With roughly a quarter of current operators in the United States above the age of 65 and given the difficulties in securing capital and land necessary to launch a new farm, policymakers have been concerned with increasing rates of retirement and the entry dynamics of farm operators. At the same time, there is a variety of different types of new farmers, with differences age, previous experience, financial standing, and education all likely to play important roles in determining their risk preferences, financial viability, and technology adoption behavior. Similarly, new farmers are often touted as more likely to participate in local food systems and seek out innovative practices, technologies, and crops, although this effect likely varies significantly by the age and experience of new entrants. As new farmers are incentivized to entry and existing farmers retire, then, there are likely to be changing patterns of farm behavior, which may have important consequences on the use of agricultural support policies, the food system, and farm incomes. This work addresses the impact of new farmer entry on two primary outcomes: farm capital investment (total machinery assets, new tractors, total tractors) and the participation of new farmers in government-sponsored conservation programs. To do so, it employs data from the 1992 to 2017 Censuses of Agriculture to evaluate the impact of the Land Contract Guarantee Pilot Program (LCGP) on the share of new farmers in counties across the United States. Estimates of shares of new farmers from a difference-in-differences model are then used to assess the impact of new farmers on aggregate measures of farm capital and federal program participation. In general, the LCGP has a significant and positive effect on the share of new farmers. Counties with higher shares of new farmers are less likely to participate in federal conservation programs and have lower total machinery assets. An event study approach that includes data from 2017, when the LCGP was expanded nationwide, confirms positive effects from the program, but offers conflicting views on farm capital and program participation. It is suggested that this is likely a result of additional programs put in place in the 2014 Farm Bill, and future research is proposed to address this conflicting sign.

ABSTRACT. The economy-wide impact of sector-specific energy efficiency investments has interested economists since Jevons (1865) first introduced the notion of a general equilibrium (GE) rebound effect. Economists have traditionally relied on computable general equilibrium (CGE) models to estimate GE rebound, but estimates based on these approaches are highly varied, ranging from negative rebound to backfire. Moreover, the ``black-box" nature of CGE models make the underlying mechanics of these predictions largely unverifiable. Consequently, numerical estimates of GE rebound are subject to a variety of criticisms, and ultimately, perceived as less dependable for policy design.

In this paper, we develop a novel GE model with non-unitary elasticities of substitution and input-output linkages to resolve these shortcomings of numerical approaches. We seek to explain (1) the mechanics underlying the general equilibrium rebound effect and (2) the reasons why numerical estimates are so varied in practice. Importantly, we design our theoretical model to reproduce the internal elements of CGE models used to estimate GE rebound. Using the model, we decompose the GE rebound effect into a price, scale, and composition channel, illustrate how input-output linkages determine the magnitude of these channels, and provide evidence that variation in the input-output network of the economy explains between 38 and 92 percent of the variation in numerical estimates of the GE rebound effect.

From a practical perspective, our results suggest that even if an analyst adopts a common model framework, with identical elasticities of substitution, and applies this framework to two similar economies, differences in the network of input-output linkages will generate variation in estimated rebound effects. Our findings also suggest that variation in estimates can emerge from applying the same energy efficiency improvement to different sectors within the same economy. These results are, therefore, purely based on supply chain relationships between sectors encapsulated by the network of input-output linkages. We refer to these relationships as ``percolation centralities." Percolation centrality, a new and important concept, captures how two sectors with the same observable features, such as total energy use or total revenue, but with different positions within the input-output network, would display different outcomes for the GE rebound effect.

Our work is unique in that it highlights the crucial role that input-output networks play in determining the effectiveness of energy efficiency policy. Accepting our results requires reimagining traditional energy policy. Unlike current policy approaches, energy policy in an interconnected world must better incorporate the impact of interregional and international supply chains on each sector’s energy performance. Our results also show that for policies to be most effective, we need to address not only sectors with the highest energy use, but also those sectors who occupy more central positions within the economy. Core sectors may substantially influence the propagation of energy efficiency shocks and lead to larger reductions in energy use than by focusing policy on traditionally energy intensive sectors, such as manufacturing. Moreover, our framework offers important lessons for the design of traditional policy instruments, such as carbon taxes or research subsidies, when applied in a GE setting.

ABSTRACT. The Research Division at the U.S. International Trade Commission (ITC) has built a website that offers industry-specific, partial equilibrium (PE) modeling tools that can be used to simulate the economic impact of changes in trade policies. This PE Modeling Portal is available online to other government economists and the public at https://usitc.gov/data/pe_modeling/index.htm. The models are based on economic theory and can be applied by the user to specific industry data and policy scenarios. They take complicated trade policy modeling issues – including foreign direct investment, oligopoly interactions, stockpiling and gradual supply adjustment, intellectual property right, and offshoring – and translate complex models from the academic trade literature into practical Excel spreadsheet models that are easy to run, without expertise in the underlying equations of the model or specialized mathematical software. We will be releasing a new revision to the PE Modeling Portal in March 2020, with substantial additional content, and propose to present how these various trade policy models are built and how they can be used.

ABSTRACT. Official measures of Bureau of Labor Statistics import and export price indexes are calculated from a business survey, but recent research calculating unit values based on administrative trade data for homogeneous products show that some unit values have the potential to replace or expand detail in published measures. The paper will discuss the opportunities that this research using 200 million data records provides, from improved methodology to expanded coverage. The paper will also describe the obstacles to bringing the data into production. These obstacles are more than logistical. There is a fundamental difference between the BLS approach, which selects a subset of homogeneous products, and the approach by many researchers to calculate price indexes for all product categories. The two approaches and their underlying assumptions are evaluated. Results from BLS research highlight the effect of unit value bias on price indexes of heterogeneous product groups.

ABSTRACT. Many OECD countries, including the U.S., have adopted research and development (R&D) tax credits to encourage innovation, especially for those small and medium enterprises (SMEs) that do not have relatively abundant financial resources like their counterparts, the industry incumbents. But countries are different in the design of tax mechanisms. Moreover, studies have shown that smaller firms are important job generators and more innovative than larger firms (Klette and Kortum, 2004; Michaelidou et al., 2011). However, both U.S. and OECD data show that large firms dominate the R&D investments not only domestically but also globally. For example, the U.S. National Science Foundation reports that more than 80% of manufacturing R&D are undertaken by large firms and the OECD Science, Technology and Industry Scoreboard of 2015 reports that more than 60% of global R&D is done by only 250 companies. Moreover, compared with their large incumbents, SMEs are more vulnerable in the increasing global competition environment. Therefore, it is important to investigate whether in the U.S., the R&D tax credit stimulates SMEs to invest more in R&D, whether firms in different industries exhibit different R&D investment patterns, how the differences relate to the degree of their response to the R&D tax credit and the degree of their exposure to import competition. To our knowledge, there is no research answering those questions. This research aims to fill in the gap. Our study shows some interesting findings: First, after the newly enacted R&D tax credit in the U.S. in 2009, more SMEs are eligible and qualified for R&D tax credits and the value of our R&D inequality index declined dramatically after 2009. Second, when examining the index by industry in detail, we find that the R&D tax credits can favor either large firms or SMEs depending on the industry that we study. Third, our panel regression analysis indicates that import competition can negatively affect U.S. innovation but the negative effect can be mitigated as the degree of R&D inequality increases. Fourth, the degree of R&D inequality has a statistically positive relationship with U.S. innovation measured by U.S. capital stock of R&D assets. However, when measured by U.S patents, U.S. innovation has a statistically negative relationship with the degree of R&D inequality. That is, lowering R&D inequality can increase U.S. patenting.

ABSTRACT. We examine the impacts of the 2018-2019 U.S. import tariff increases on U.S. export growth through the lens of supply chain linkages. Using 2016 confidential firm-trade linked data, we document the implied incidence and scope of new import tariffs. Firms that eventually faced tariff increases on their imports accounted for 84% of all exports and represented 65% of manufacturing employment. For all affected firms, the implied cost is $900 per worker in new duties. To estimate the effect on U.S. export growth, we construct product-level measures of import tariff exposure of U.S. exports from the underlying firm micro data. More exposed products experienced 2 percentage point lower growth relative to products with no exposure. The decline in exports is equivalent to an ad valorem tariff on U.S. exports of almost 2% for the typical product and almost 4% for products with higher than average exposure.

ABSTRACT. Developing countries rely more heavily on financial transfers between private households for economic welfare. Using data from three middle income and three high income countries in the Luxembourg Income Study Database, this paper examines the effects of such transfers on within country comparison of inequality. Deducting private transfer payments from disposable income increases inequality, but effects differ by the position of donor and receiving households in the distribution, by urban or rural location and by age of household members. We conclude that considering the role of private financial transfers is crucial to income inequality analysis.

ABSTRACT. We model limited commitment informal insurance networks among individuals whose impurely altruistic marginal gains to giving to others diminish with the number of transfers one makes, giving is costly, and stochastic income has both publicly observable and unobservable components. Contrary to the canonical informal insurance model, in which bigger networks and observable income are preferable, our model predicts that unobservable income shocks may facilitate altruistic giving that better targets the least well of individuals within one's network and that too large a network can overwhelm even an altruistic agent, inducing her to cease giving. We test the empirical salience of the model using a unique data set from southern Ghana. We analyze transfer flows among households by coupling observations of gift-giving networks with experimental cash windfall gains - randomized between private and publicly observable payouts - repeated every other month for a year. The empirical evidence supports the model predictions. The magnitude and progressive targeting of transfers precipitated by private income gains underscores the importance of altruistic, and not just insurance, motives underpinning interhousehold transfers.

ABSTRACT. Networks in village economies can provide insurance and facilitate use of labor and intermediate inputs in production. Ironically, as not everyone is engaged in all of these networks simultaneously, arguably due to various distinct frictions, networks can both provide insurance and yet propagate shocks. We show that when one household experiences a significant health shock with associated large expenses, this can propagate to other households linked to it via the production village network. This occurs when the shocked household is not in the insurance network and, as a result, adjusts production decisions - drawing down working capital, cutting input spending, and reducing labor hiring - hence affecting households who supply inputs and labor. Likewise upstream businesses close to the under-insured households in the supply chain network experience reduced local sales and increased inventories. Similarly, workers closer to the under-insured households in the labor network have lower probability of working locally and experience reduced earnings. In turn there is evidence of ex-post adjustments of these upstream households, shifting resources towards activities with lower exposure to local shocks. Policy remedies should be geared to mitigating the underlying frictions that limit the size of these various networks.

ABSTRACT. There has been considerable interest within the U.S. Census Bureau and in the research and policy communities in measurement error and mis-reporting of income on surveys. One possible avenue to correct for these issues is to use administrative data to replace or complement survey responses. However, combining survey and administrative data is not without its challenges. In this paper, we review the literature on how household survey estimates of various income statistics may be biased by issues such as mis-reporting and increasing non-response. We describe the administrative data currently available for research use at the U.S. Census Bureau and the data available to other federal agencies that we could potentially use in the future to improve income statistics. We detail the implementation opportunities and challenges to using administrative data. Finally, we outline our research agenda to address these challenges.

ABSTRACT. Until recently, the quinquennial Survey of Business Owners (SBO) was the only source of information for U.S. employer and non-employer businesses by owner demographic characteristics such as race, ethnicity, sex and veteran status. Now, however, the Non-employer Statistics by Demographics series (NES-D) will replace the SBO’s non-employer component with reliable, and more frequent (annual) business demographic estimates with no additional respondent burden, and at lower imputation rates and costs. NES-D is not a survey; rather, it exploits existing administrative and census records to assign demographic characteristics to the universe of approximately 25 million (as of 2016) non-employer businesses, and produce an annual series that will become the only source of statistics of U.S. businesses with no paid employment by business owner demographics. Using the 2014-2016 vintages of non-employer businesses and demographic information from the decennial census, the American Community Survey, the Census Numident and AR from the Department of Veteran Affairs, we discuss preliminary results, the challenges encountered along the way, and next steps.

ABSTRACT. The narrowing of the gender wage gap has slowed in recent decades, yet among full-time, year-round workers, women earn approximately 18 to 20 percent less than men at the median. Women’s human capital endowments increasingly resemble men’s, so remaining differences in characteristics like educational attainment and labor-market experience explain less of the gender wage gap now than in the past. As human capital wanes in importance, studies show that occupational and industrial segregation explain a growing proportion of the gender wage gap. However, a major limitation of these studies is that the longitudinal datasets that contain information about labor-market experience lack the large samples needed to analyze occupation and industry effectively. This study combines survey and administrative data to examine the contribution of labor-market experience to the gender wage gap, net of gender differences in occupation and industry. We use tax records to construct detailed work histories for a sample of full-time, full-year workers in the Current Population Survey Annual Social and Economic Supplement. We estimate and compare Oaxaca-Blinder decompositions of the average wage gap between men and women using 5-, 10-, 15-, and 25-year work histories. We find that experience continues to play a statistically significant role in the gender wage gap but is far less important than occupation and industry. Moreover, using the recent rather than the full work history does not substantially alter our estimates of the contributions of human capital variables to the gap. These results suggests that, for studies focused on the last two decades, substituting potential for actual labor-market experience introduces minimal bias in analyses of the gender wage gap using large, nationally representative datasets.

ABSTRACT. This paper outlines a generic framework for forecast evaluation and improvement; and, it illustrates that framework with empirical analyses of different U.S. government agencies' forecasts of U.S. federal debt, and of the Fed's forecasts of Chinese GDP growth. Techniques for forecast evaluation include comparison of mean squared forecast errors, forecast encompassing, tests of predictive failure, and tests of bias and efficiency. Recent extensions of these techniques utilize machine-learning algorithms to handle more potential regressors than observations---a characteristic common to big data. These techniques are generally applicable, including to both economic and non-economic forecasts. Evaluation of forecasts is fundamental to assessing the forecasts' usefulness; and evaluation often indicates ways in which the forecasts may be improved. The empirical examples show how to apply this framework for forecast evaluation and how to use it for forecast improvement.

ABSTRACT. I analyze damages from hurricane strikes on the United States since 1955. I select the most important drivers for damages using machine learning methods and show that large errors in a hurricane's predicted landfall location lead to higher damages. This relationship holds across a wide range of model specifications and when controlling for ex-ante uncertainty and potential endogeneity. I find that the cumulative reduction in damages from forecast improvements since 1970 is about $82 billion, which exceeds the U.S. government’s spending on the forecasts and private willingness to pay for them.

ABSTRACT. A duo of vital allies concerning US economic and foreign policy in East Asia, South Korea and Japan serve as two of the most dependable export markets for US farmers. Collectively, exports to these markets account for more than fifteen percent of US agricultural exports, offering robust and diverse consumer markets for key products. Domestically, shared inhibitive factors constraining agricultural output have saddled both countries with food insecurity, creating two of the world’s largest agricultural export markets. However, with the United States’ initiative to reshape its economic interests internationally, its retreat from liberalized trade has fomented uncertainty amongst many of its allies.

The aim of this presentation is to dissect the Japanese and South Korean agricultural markets regarding commodities vital to US interests. By establishing a thorough foundation of each country’s agricultural sector, as well as evolving dietary, demographic, and socioeconomic trends, this study will highlight the characteristics that afford US farmers these vital export markets. Using national-level monthly data from 2000 through 2019, this analysis will employ an additive triple exponential smoothing model to develop five-year forecasts for top US and its competitors for beef, pork, wheat, and dairy exports to both countries, respectively. This analysis will illuminate the significance and foreseeable obstacles to US market-share to inform US outlook in East Asia.

Compounded by demographic and socioeconomic transformations inhibiting self-sufficiency of most foodstuffs, agricultural sectors in Japan and South Korea are small and contracting. Several endogenous factors related to each agricultural sector are further exacerbated by exogenous challenges concerning geographical endowment and economic development. Arable land is scarce in both countries where mountainous terrain unsuitable for farming accounts for over seventy percent of total land area. Byproducts of rapid development led to a contraction of available farmland further compounded by rural flight, leaving what remained in the care of a dwindling and elderly rural population. In tandem with economic development, the diffusion of western dietary staples following conflicts in the region initiated a gradual diversion from historical dietary habits in each country.

As consumers favor foodstuffs not conducive to domestic cultivation, the Japanese and South Korean governments are forced to lower trade barriers on agricultural products. Additionally, resource deficiency plays an underlying role both economic policies, necessitating a reliance on trade to fuel export-oriented growth strategies. This is evident in each countries’ accelerated pursuit of trade partnerships in recent years. While the US reconfigures its global economic interests, resource deficiency and food insecurity compel both countries to seek expedient solutions to global economic volatility. As such, resultant partnerships endanger US market-share for key agricultural items that will be the focus of this paper.

This research portends a variety of significant economic and diplomatic implications for US interests. Emerging markets within the US defensive alliance system, such as the Philippines, Indonesia, and Vietnam, will conceivably follow similar development patterns as Japan and South Korea. Such emerging markets provide export opportunities to US farmers. However, continued geopolitical upheaval risks alienating these economic partners, pushing them closer to US competitors.

ABSTRACT. It is well known that random assignment guarantees unbiased estimation of policy effects by simply comparing mean outcomes, but essentially all policy experiments use at least one (if not several) strategies to adjust for covariates. The adjustment introduces a small amount of bias, but can potentially dramatically reduce variance, so that mean squared error is lower overall. However, these different methods face several different tradeoffs, and we illustrate these tradeoffs via a simulation where we know the true effect and attempt to recover it via alternative estimation strategies. One new approach is shown to have substantial advantages over alternatives, and several strategies can be seen to be strictly dominated.

ABSTRACT. This study explores impacts on compliers (those who are treated when randomly assigned to receive treatment) in the context of a randomized controlled trial (RCT) of a federally-funded sectoral job training program. The Health Profession Opportunity Grants (HPOG) Program experimentally tested the impact of three specific program enhancements: emergency assistance, non-cash incentives, and facilitated peer support. The study explores the extent to which these enhancements influenced educational progress among those who took up the enhancements as compared to a variety of comparison groups. The analysis also explores the performance of nonexperimental methods at recovering the experimental benchmark of the impact among compliers. Given constraints such as time, money, and logistics associated with RCTs, researchers might choose such nonexperimental approaches to estimating program impacts. While RCTs provide internally valid estimates of the impact of an offer of job training programs or enhancements, there is often interest in the impact among individuals who take up the program’s offer (compliers). The results here demonstrate that nonexperimental approaches fail to replicate the impacts from the experiment on compliers to differing extents across the three enhancements.

ABSTRACT. Permanently. I study the removal of an Australian tax incentive generating marginal tax rates of between 30 and 30,000%. Two-fifths of people responded to the incentive, raising average deductions by $500. Responders differed markedly from non-responders in observed variables. I uncover a new consequence of frictions attenuating the response to taxes: among those responding to the incentive, deductions remained elevated at least seven years after its removal. This was concentrated among those doing their own taxes, suggesting responders learnt about their taxes. My results suggest taxes have a larger welfare impact than previously thought, and caution against sharp tax incentives.

ABSTRACT. The U.S. Department of Veterans Affairs provides medical care to a large share of veterans. This report describes some of the distinctive features of the system including the services provided, funding mechanism, and cost to veterans. It examines the growth in the cost of providing services over the past decades and of expanding access to medical care by non-VA providers and facilities. The paper presents two projections of VA spending growth over the next decade—illustrating what VA’s spending would be if the Congress appropriated funding under varying assumptions. The first projection illustrates the sensitivity of spending based on anticipated higher-than-inflation increases in overall medical spending in the economy. The second projection shows how spending might increase if past spending trends continue. Both could raise concerns about affordability in a period of constrained federal budgets.

ABSTRACT. The Department of Defense operates hundreds of bases that support the daily operations of units in the Air Force, Army, Marine Corps, and Navy, providing services such as housing, utilities, and grounds maintenance, much as might be found in a town or city. In 2016, $25 billion—about 4 percent of the Department’s budget— was allocated to the costs of such services, called base operations support (BOS). This analysis explores certain characteristics of bases, such as their size, geographic location, and the mission of the units they serve, and uses statistical methods to assess the relationship between those characteristics and BOS costs.

ABSTRACT. The Department of Defense (DoD) and the Department of Veterans Affairs together expect to spend about $350 billion in 2019 on compensation and benefits for current and former military personnel. In this report, the Congressional Budget Office examines spending on military compensation— cash, current noncash benefits, and deferred benefits for enlisted personnel and officers—and its effects on recruitment, retention, and motivation. CBO found that cash compensation for service members, including nontaxable allowances for housing and food and the associated federal tax savings, exceeds in most cases DoD’s benchmark, which is equal to the 70th percentile of wages and salaries for comparable civilians, often by a significant amount. Relative to civilian compensation packages, total military compensation is weighted more heavily toward noncash and deferred benefits (such as health care, educational programs, and retirement pay). The military’s compensation package could be changed to slow the growth of government spending while still attracting a high-quality force or to strengthen the link between compensation and recruitment, retention, and motivation. CBO examined five approaches that would alter the way that DoD compensates military personnel.

ABSTRACT. I examine whether the ACA Medicaid expansion increases the supply of entrepreneurs as measured by self-employment. Using the 2003–2017 Current Population Survey, I apply difference-in-differences, propensity score weighting, and instrumental variable (IV) methods. I find that expanding Medicaid eligibility raises the self-employment rate by 10–14%, without increasing self-employment exit. IV estimates indicate that newly covered individuals have a higher propensity to become self-employed. Exploiting additional variation by spousal coverage or poor health of individuals or their spouse, I find evidence that the underlying mechanism of the effect was through increasing access to health insurance. Findings suggest that limited access to health insurance may be a barrier to entrepreneurship.

ABSTRACT. The ACA includes numerous provisions that expand health insurance coverage for women of reproductive age (15-44). In this study, we assess the impact of the ACA Medicaid expansion, which, in participating states, extended Medicaid eligibility to low-income women of reproductive age prior to conception, on fertility and the socioeconomic and demographic composition of new mothers. By reducing financial barriers to reproductive health services and improving access to care, these provisions may affect decisions surrounding pregnancy and the demographic and socioeconomic composition of women with a recent birth. Changes in fertility may, in turn, result in changes in school enrollment and labor market participation among women of reproductive age.

Measuring overall fertility alone may obscure underlying changes in the effect of ACA Medicaid expansions for different subgroups of women, which may further contribute to changes in the composition of women with a recent birth. By studying these characteristics directly, we assess how ACA Medicaid expansions affected different groups of women and provide insight into the pathway through which fertility was affected. For example, a decline in births to unmarried women may suggest a decline in unintended births attributable to increased access to family planning. Alternatively, an increase in births to married, educated women may suggest an increase in planned births, attributable to the income effect associated with Medicaid expansions.

We use a difference-in-differences strategy and a triple-difference strategy to investigate the extent to which ACA Medicaid expansions affected fertility, composition of recent mothers, and school enrollment and labor market participation among women of reproductive age. We use restricted-access natality records comprising the total universe of U.S. birth certificates and include information on date of delivery, maternal demographic characteristics, and prenatal maternal behaviors. Access to restricted-use records provides information on the mother’s state and county of residence, which is necessary to compare trends in fertility and composition of recent mothers across expansion and non-expansion states. We supplement our fertility and composition of new mother findings with analysis using the American Community Survey. Finally, we use information from the ACS to assess the impact of the ACA Medicaid expansions on school enrollment and labor supply of women of reproductive age.

Preliminary findings from the ACS indicate that, among ages 19-44, Medicaid expansion is associated with reducing births by 2.8 percent in expansion states relative to nonexpansion states. These reductions are larger for non-Hispanic black women (8.5 percent), and small and insignificant for non-Hispanic white women and Hispanic women. These early findings suggest that Medicaid expansion is associated improving access to family planning services to vulnerable populations. Finally, we observe no association between Medicaid expansion and employment or school attendance for women of reproductive age.

ABSTRACT. Whether the coverage provisions under the Affordable Care Act (ACA) would lead to adverse labor market consequences in the form of reduced employment, hours worked, earnings, and offers for employer-sponsored insurance (ESI) has been the subject of substantial debate and analysis. Using data from the American Community Survey and the Current Population Survey, we exploit variation in baseline occupation-specific uninsured rates to assess how the main ACA coverage provisions affected coverage, the number of workers in each occupation, weekly earnings, usual number of hours worked per week, and rates of ESI coverage. We find that occupations with the highest baseline uninsured rates in 2010 (and therefore most targeted by the ACA) experienced the greatest gains in coverage after ACA implementation – occupations with a 10 percentage point higher baseline uninsured rate experienced coverage gains of about 2.7 percentage points after the implementation of the ACA’s main coverage provisions in 2014. Despite these large effects on workers’ coverage, we find no evidence that the ACA was associated with reductions in the number of workers, weekly earnings, or the usual number of hours worked per week. Moreover, we find no evidence that occupations more targeted by the ACA saw reductions in employer-sponsored insurance rates. If anything, we find that the ACA is associated with modestly raising weekly earnings, hours worked, and the likelihood of ESI coverage, findings opposite from most predictions of the labor market impacts of the ACA.

ABSTRACT. Homeless individuals grapple with varied health problems but have limited access to health care. This study, extending the literature on health care and homelessness, provides the first causal evidence of the impact of the Medicaid expansion on the migration of homeless individuals from non-expansion to expansion states. The Medicaid expansion, adopted by 26 states and Washington DC and rejected by 24 states in January 2014, expanded coverage to previously uninsured homeless individuals. Besides health care, the expansion equipped homeless service providers with extra flexibility to tackle homelessness through several supportive services and housing-related activities, absent in non-expansion states. Using the state-level data on the homeless population from 2009-2018, released by the Department of Housing and Urban Development (HUD), the estimates from a difference-in-differences model suggest migration of homeless individuals from non-expansion to expansion states. After 2014, the expansion states saw a statistically significant 4.4 percent increase, while non-expansion states simultaneously experienced a 22.6 percent decline. A pre-treatment parallel trend establishes the identification, and the results are robust to several specifications, including controls for essential predictors of local homelessness and falsification tests. Furthermore, utilizing the difference in coverage status of homeless individuals vis-a-vis homeless people in families (who had pre-expansion coverage), estimates from a triple difference (DDD) model also confirm a significant migration. Previous studies documenting the positive impact of Medicaid expansion on several outcomes – improved coverage and health care utilization, reduced financial hardships and medical bills sent to collection, and increased household savings – complement the results of this study. This paper provides additional evidence that Medicaid expansion had no significant impact on the number of homeless veterans or indicators like bankruptcy, unemployment rate, and poverty rate.

ABSTRACT. This paper adds a spatial angle to economic growth. We create an endogenous growth model, where agents need to find specialists for their ideas. The closer together (further apart) agents live, the easier (harder) it is to find these specialists and thus it is cheaper (more expensive) to implement ideas. This lower (higher) cost of implementation causes more (less) ideas getting implemented which in turn translates into faster (slower) economic growth. We then create the new measure Spatial Population Proximity, for how close people live together and empirically show that areas where people live closer together tend to grow faster both for the US and the world.

ABSTRACT. Economic models of migration, domestic and international, typically begin with the assumption that a moving household's primary goal is to attain higher income than it would earn by staying. This article uses administrative records for almost all people earning formal market income in the U.S., 2001-2015, totaling about 1.7 billion household observations with 82 million long-distance moves, to develop a detailed match between movers and comparable stayers and thus a comparison of movers' income changes relative to stayers. In aggregate, movers see about a median 1% gain in income after moving relative to the counterfactual of staying, with wide variance. Even a decade later, about two out of five households have lower income relative to staying, with an overall median relative income gain of about 6%. Pecuniary benefits are not evenly distributed: movers leaving school and younger single households without children are likely to see higher income relative to staying, but other movers, most notably single parents, are roughly half as likely to see a relative income gain. The overall story is a bifurcated population of movers. Roughly half move to higher income relative to staying, and the rest do not, indicating for whom the hypothesis of income maximization is difficult to support, and where future research about the many motives for moving may focus.

ABSTRACT. Consumer interest in locally grown food has been increasing dramatically in the United States via food hubs, farmers markets, and other venues. The number of farmers markets has grown significantly from 1,755 in 1994 to more than 8,600 markets as currently registered in the USDA Farmers Market Directory. This increase can be attributed to the increased demand for fresh, locally produced products. We developed an IMPLAN-based SAM model of 22 counties surrounding and including the District of Columba to evaluate the direct, indirect, and induced economic impacts of farmers' markets on the study region. To supplement the input of IMPLAN, this research incorporates robust data collected from a consumer survey of shoppers in Maryland, Virginia, and the District of Columbia conducted in 2017. The empirical results from IMPLAN show the direct gross sales and income figures into an estimate of the number of jobs in the study region’s economy that were tied to farmers' market activities. The analysis also provided jobs created from indirect and induced effects. The average income multiplier is 1.51 indicating that a $1 increase in personal income for a farmers market translates to $1.51 in personal income across the economy of the study region.

ABSTRACT. Brexit will have major implications for trade Western European nations and the Eastern European nations. In the short term, the decision to withdraw from the European Union will begin to complicate trade patterns with all members of the European Union because is no actual trade deal in place after exiting Brexit.  A final deal is not expected until December 2020. If such a trade deal is reached, there are longer-term consequences to be considered, not just for the UK but with its trade relations with its Eastern European countries. Robinson (2015) stated that probably the most important consequence of Brexit would affect the movement of capital that will affect businesses by increasing uncertainty to firms in the UK and those in Eastern Europe. The increase in uncertainty would reduce investment to the UK and Eastern European countries.   This change would impact company revenues and profits and yield a negative impact on its share prices.  Given the potential uncertainty, this paper examines the impact of Brexit on the projected impacts of future trade patterns with the Eastern European members and the UK.  

ABSTRACT. What is the optimal temperature finance gravitating towards? While we have empirical evidence for the optimal cardinal temperature for GDP production and international development links external climate conditions to levels of societal development, no literature exists on climate-induced-finance flows. Research on temperature-dependent financial flows holds invaluable insights considering climate change. The connection between climate change and finance will be drawn via price mechanisms. Given the extinction potential of crops, industry and service production, price mechanism will be prospected with a hyperbolic tilt towards the end of durability and the closeness to extinction. In contrast to classical and standard approaches in economics to determine prices, the following research thereby considers that agents in their behavior can be constrained by shrinking timeframes for production considering global warming. The contemporary attention to global warming and climate shocks is thereby assumed to affect the price expectations and hence actual market prices of commodities. Paying attention to supply and demand side perspectives, inflated prices surrounding scarcity will be first modelled and then backtested on data about prices in commodities of food and beverages.

ABSTRACT. U.S. consumers, businesses, and government entities spent $1.71 trillion on food and beverages in 2018. Spending at food-away-from-home establishments—restaurants, school cafeterias, sports venues, and other eating places—accounted for 54.4 percent of these expenditures, and the remaining 45.6 percent took place at grocery stores, supercenters, convenience stores, and other retailers. The away-from-home market, which accounted for about one-third of total food expenditures 50 years ago (33.8 percent in 1968), has grown through the decades. The purpose of this paper is to attribute changes in food expenditure to various causes from 1997 to 2018 and assess the relative effects in periods before and after the Great Recession. In particular, we look at how the GDP, the share of retail spending, the share of retail and food spending, and the share of food-at-home has on the changes of food-away-from-home (FAFH). We use the food expenditure series (FES) created by the Economic Research Service of USDA, which measures the value of the U.S. food system over time, by outlet and product type, and by final purchasers and users. We combine FES with other census data to do the decomposition analysis. Preliminary results suggest that nominal FAFH grew at annual rate of 5 percent from 1997-2018. Out of the four components that contributed to the growth of FAFH, the share of retail spending is the one that contributed the most.

ABSTRACT. This study measures price elasticity of demand for 45 consumer products and services for the U.S. from 2005 to 2017 with the total expenditure method and uses average expenditure data per consumer unit household from the Consumer Expenditure Survey and retail price indicis from the Consumer Price Index. This method measures the impact of price changes on consumption by the consumer, and although it cannot provide an exact magnitude of elasticity, it nevertheless provides a proxy of elasticity. It provides a good comparison to other studies which uses traditional constant elasticity demand models that uses quantity and price data. This alternative method can be useful for instances when quantity demanded is not readily available. For example, while BLS does not publish data on quantity demanded for goods and services, it does publish extensive consumer expenditure data. This paper examines whether using expenditure data in place of quantity demanded data to calculate price elasticities of demand is an acceptable fallback measure and whether the elasticities calculated are meaningful, and what the limitations are with using this method. Usually the price of a good or service and the demand for that good or service are inversely related to each other. In other words, when demand is elastic, a fall in the price of a good results in increase in total expenditure. When price increases, the expenditure decreases. When a good is elastic, the price and expenditure will move in opposite directions. On the other hand, there are some goods which do not behave this way. Inelastic goods are those goods, the demand for which remains change constant and it is not effected by changes in price. We use over ten years worth of actual household consumer expenditures data to observe whether elasticities changed in recent years and to help form a better sense of recent general consumption patterns. What we have found is that more goods have become price elastic in the recent years, meaning consumers have become more price sensitive in recent years, even with goods that historically were considered necessities and inelastic. Other studies that do not focus on recent years or that utilize consumer preference surveys may come up with more inelastic estimates for the same goods. Consumer households have been faced with cumulative price increases and the stagnation of wages and have thus began to change household consumer behavior. More goods, even those that in the past were deemed inelastic or necessity goods, have become elastic.

ABSTRACT. We study the causal effect of mortgage rate changes on consumer spending, debt repayment, and defaults during an expansionary and a contractionary monetary policy episode in Canada. Our identification takes advantage of the fact that the interest rates of short-term fixed-rate mortgages (the dominant product in Canada's mortgage market) have to be reset according to the prevailing market interest rates at predetermined time intervals. Our empirical strategy exploits this exogenous variation in the timing of mortgage rate resets. We find asymmetric responses of consumer durable spending, deleveraging, and defaults. These results can be rationalized by the cash-flow effect in conjunction with changes in consumers' expectations about future interest rates. Our findings help to understand the responses of the household sector to changes in the interest rate, especially in countries where variable-rate, adjustable-rate, and short-term fixed-rate mortgages are prevalent.

ABSTRACT. The optional provision of Medicaid expansion, through the Affordable Care Act (ACA), has triggered a national debate among diverse stakeholders regarding the impacts of Medicaid coverage on various dimensions of public health, costs, and benefits. Randomized experiments like the Rand Health Insurance Experiment and the Oregon Health Insurance Experiment have generated some credible estimates of the average treatment effects of access to insurance. However, identical policy interventions can have heterogeneous effects on different subpopulations. This paper uses data from the Oregon Health Insurance Experiment to estimate the heterogeneous treatment effects of access to Medicaid on health care utilization, preventive care utilization, financial strain, and self-reported physical and mental health. I detect heterogeneous treatment effects using a cluster-robust generalized random forest, a causal machine learning approach. I find that the impact of Medicaid is more pronounced among relatively older non-elderly and poorer households, which is consistent with standard adverse selection theory. Furthermore, I implement the "efficient policy learning," another machine learning strategy, to identify policy changes that prioritize providing Medicaid coverage to the subgroups that are likely to benefit the most. On average, the proposed reforms would improve the average probability of outpatient visits, preventive care use, overall health outcomes, having a personal doctor and clinic, and happiness by a range of 2% to 9% over a random assignment baseline. These findings show efficient Medicaid eligibility criteria that reduce cost while improving Medicaid outcomes and, therefore, are useful as a Section 1115 waiver tool to design Medicaid.

ABSTRACT. The PPI currently uses a modified Laspeyres formula to calculate all indexes, including industry net output indexes, commodity indexes, industry net input indexes, and FD-ID indexes. The modified Laspeyres formula is used at the elementary level (or cell level) of index calculation, as well as at levels of aggregation above the elementary level. It was recently brought to the PPI program’s attention that, based on recommendations from their technical advisory committee, Canada plans to adopt a geometric mean formula for elementary level calculations. In addition, several other countries (including Italy, Chile, Norway, Denmark, Japan, and the Netherlands) are currently using some version of a geometric mean formula to calculate elementary level producer price indexes. Based on this information, the US PPI program decided to investigate whether it should change its formula for calculating elementary cells to one based on a geometric mean rather than an arithmetic mean. In the 1990s, the PPI had also considered using a geometric mean based formula to calculate elementary level indexes, but chose to remain with a Laspeyres formula. At the time, a geometric mean formula was largely being considered because the CPI had just moved from a modified Laspeyres formula to a weighted geometric mean formula for elementary level price indexes for most items in the CPI market basket. The CPI made this change to capture consumer substitution that occurs across specific goods and services within elementary items, as the geometric mean formula imposes substitution towards products with relatively lower price change. The PPI program decided not to make this formula change because, under the assumptions of the theoretical model of PPI output price indexes, producers do not substitute to relatively lower priced products. Instead, in an effort to maximize profits, producers substitute their outputs towards relatively more expensive products. After the PPI program became aware that Canada was being advised to adopt geometric mean formula, the program contacted Canada and learned that they were making the change primarily due to axiomatic reasons as opposed to reasons based on the economic model. The PPI program therefore decided to reinvestigate whether it should change its formula for calculating elementary cells. This paper considers both the axiomatic and economic approaches to assess whether PPI should adopt a geometric Young formula. The axiomatic approach evaluates index formulas by comparing mathematical properties of the indexes. Indexes with more favorable axiomatic properties are generally determined to be superior. The economic approach assesses observable indexes by comparing them to a theoretical index based on a model where prices and quantities are related to each other and are determined by constrained optimization problems. Formulas that more closely approximate the theoretical index are deemed superior.

ABSTRACT. The economic questions that can be addressed using many of the large administrative employer-employee linked data sets with workers’ earnings has been limited by the absence of information on workers’ base wages, variable compensation, hours or weeks worked, and other factors determining workers’ earnings. This paper presents a set of machine learning methods that identify each worker’s unobserved persistent base wages, paydays weeks, and annual bonuses from the worker’s quarterly earnings. I then implement and evaluate the quality of these methods using quarterly earnings data in the U.S. Census Bureau’s Longitudinal Employer-Household Dynamics (LEHD) dataset, an employer-employee linked dataset for the United States. Using the estimated nominal wages of workers in 30 U.S. states, I document four patterns of nominal wage adjustment: i) estimated persistent wage changes exhibit downward nominal wage rigidity, ii) optimal real wage cuts are suppressed by downward nominal wage rigidity, iii) workers’ nominal raises follow a Taylor-like pattern, with the probability of a wage raise spiking every four quarters, and iv) the timing of workers’ annual raises are synchronized within the firm.

ABSTRACT. The Supplemental Poverty Measure (SPM) was designed to more fully account for the resources that individuals have to meet their basic needs. As such, the SPM adds the value of in-kind benefits and deducts necessary expenses from total resources; these deductions include state and federal income taxes, child support paid, medical expenses, and work-related expenses such as commuting and child-care costs. Currently work-related expenses (excluding childcare) are derived from the Survey of Income and Program Participation (SIPP) and are set at 85 percent of median weekly expenses. This value is flat across all modes of transportation and geographic areas. This paper takes advantage of new data from the American Housing Survey (AHS) commuting module to describe and analyze alternate methods of estimating commuting costs for the SPM, allowing for geographic variation in commuting costs by Census division. Poverty estimates using the two methods are compared and evaluated.

ABSTRACT. The Supplemental Nutrition Assistance Program (SNAP) is a major piece of the social safety net. At peak participation during the Great Recession in 2013, an average of 47.6 million individuals received SNAP benefits each month with annual cost of $76.1 billion. However, research has found substantial and systematic under-reporting of the receipt of SNAP benefits. This makes it challenging to assess the impact of SNAP receipt on outcomes of policy interest. For example, estimates of the relationship of SNAP receipt and labor force participation or earnings as well as SNAP’s impact on material well-being could be biased by SNAP under-reporting. Unfortunately, because SNAP is administered at the state level, comprehensive nationwide administrative program data is not available. We address this issue by using administrative SNAP data from 8 states to impute “true” SNAP participation in the other 42 states and DC. We validate our approach by implementing a “leave-one-out” imputation, where we leave each state with administrative data out of the imputation model separately and compare our imputed SNAP benefits in those states to the actual observed administrative data. We then show how estimates of the Supplemental Poverty Measure (SPM), a measure of poverty that includes resources from SNAP and some other in-kind benefits, are affected by correcting for SNAP under-reporting.

ABSTRACT. This paper matches administrative data from Housing and Urban Development (HUD) to Consumer Expenditure (CE) data to investigate alignment between survey responses to questions about rental assistance receipt and monthly rent paid with administrative records data. We first link all CE sample units to the HUD data, and find that sampled HUD participants are six percentage points more likely to respond to the CE than are non-HUD participants. We compare the amount of monthly rent reported in the CE compared to HUD administrative records for CE respondents. The average difference between HUD reported total tenant payment (the amount less any subsidy) and CE reported rent is -$100, suggesting that CE rent is over-reported for the linked sample. This implies that the proportion of individuals experiencing rent burden is 7 to 11 percentage points lower when calculated using linked CE-HUD data than when using CE survey data alone. Additionally, Supplemental Poverty Measure (SPM) poverty thresholds are increased by $120 to $460 when replacing survey-reported rent with HUD administrative data. In contrast, when the survey-reported rents were replaced with imputed market rents, the SPM thresholds increased $356 for renters.

ABSTRACT. Official poverty statistics are used in the United States to evaluate economic well-being at the national level, and to distribute federal anti-poverty funds across states and urban areas. However, these statistics are based on poverty thresholds that do not take into account geographic differences in price levels. To provide an alternative estimate, beginning in 2011, the U.S. Census Bureau has issued a supplemental poverty measure (SPM). Unlike the official measure, the SPM adjusts the poverty thresholds for geographic differences in the cost of housing. This paper examines the impact of a change in the methodology for calculating geographic adjustments for the Supplemental Poverty Measure (SPM) that uses a comparable wage index modeled on the work done by the National Center for Education Statistics.

ABSTRACT. This paper explores the effect of a credit supply shock from small banks on small businesses. I construct exogenous credit supply shocks by exploiting the fact that small banks reduce lending when exposed to natural disasters elsewhere in their lending network. I find that when small banks reduce lending to non-disaster counties, large banks do not increase small business lending, and the number of small businesses diminishes. I conduct additional tests to examine the role of soft information, or information that is not easily quantifiable. The decline in businesses is more pronounced in industries that would benefit from lenders that collect soft information about the local market; an additional test using a proxy of market-wide adverse selection suggests that large banks rely on coarse market-level signals instead of borrower-specific soft information. Overall, these results suggest small banks are an important source of small business financing because of their ability to collect different types of soft information.

ABSTRACT. Increases in concentration across various sectors are a salient feature of industry dynamics in the U.S. and other developed countries during the past 30 years. This trend is particularly notable in the U.S. retail sector, which has changed over the past three decades from one with many small local firms to one dominated by large national firms. Existing work on concentration focuses on nationwide changes (the national Herfindahl-Hirschman index more than doubled between 1997 and 2007). Yet, less is known about the dynamics of concentration in local markets, and the relationship between local and national trends. We address this issue by providing a novel decomposition of national concentration into a local and a cross-market component. We use new data on store-level revenue for all U.S. retailers by 20 major categories of goods to measure concentration in local markets at the product level. Despite local concentration increasing during the last three decades (the Herfindahl-Hirschman index increased 50 percent), we show that the rise in national concentration is driven almost exclusively by the expansion of large firms into more markets, with changes in local concentration having little impact on national trends.

ABSTRACT. One might expect multinational enterprises (MNEs) to produce a different set of products than purely domestically focused firms, even in the same industry, given their need to serve both domestic and foreign customers and other factors. Recent trade literature, such as Mayer et al. (2014), formally models and finds evidence of product mix heterogeneity in related contexts. This evidence is one of the motivations behind current efforts in the international statistical community to produce extended supply and use tables (ESUTs), which can show breakdowns of economic activity and operational characteristics by firm ownership. Our study describes differences in product scope, diversification, type, and concentration between MNEs and non-MNEs in the United States. We also explore how these product choices relate to operating characteristics, such as firm size and productivity, and how firms organize multiproduct production across establishments. We use linked Census-BEA data that allow us to identify MNEs in the Census business register universe. Our analysis uses establishment-level data on sales by product type from the 2012 Economic Census for the services and distributive trade sectors, which contains sales data for over 3,000 distinct products. The rich product detail in the Census data is an integral component of the supply and use tables produced by the BEA (Young et al. 2015). Our analysis complements current efforts by BEA using the linked product data to produce U.S. economy-wide ESUTs disaggregated by MNE ownership type (e.g., Fetzer et al. 2018). Identifying firm-level product heterogeneity within industries informs the current discussion in the international statistical community over ESUTs, in terms of their importance and specific format. This research further contributes to existing research on multiproduct multinational firms, which primarily focuses on manufacturing industries (e.g., Bernard et al. 2011, Mayer et al. 2014), and on product choice heterogeneity within industries, by examining new dimensions of heterogeneity along product choice (e.g., Escaith 2015, Piacentini and Fortanier 2015).

ABSTRACT. The U.S. Department of Agriculture (USDA) has been monitoring the farm sector’s productivity performance for decades. Today, having incorporated recommendations made by an AAEA taskforce (USDA, 1980) and by a second more recent panel (see Shumway et al., 2014, 2017), the USDA bases its official productivity statistics on a sophisticated system of production accounts (USDA-ERS). In this presentation we will: 1. introduce the production accounts used to measure U.S. agricultural productivity statistics; 2. address some major changes after the second data review and with the most recent data release (2020); 3.discuss some findings and sources of growth from the U.S. productivity accounts (1948-2017); 4. Talk about our ongoing research projects that aim to improve the quality of US agricultural productivity accounts measures.

ABSTRACT. NOAA Fisheries Office of Science and Technology has initiated a national program to develop and report economic performance indicators for fishing vessels. One major product from this effort was publication of productivity change estimates. Annual Multifactor productivity (MFP) measures were produced using a Lowe index number approach for vessels operating in 20 catch share programs. This study proposes a new method to measure regional MFP based on a Törnqvist index and alternative output and input measures. We draw vessels data from nine major fisheries in the Northeast region of the United States, spanning the years 2007-2018. The proposed input measure sheds light on estimating MFP for the aggregate commercial fishery sector.

ABSTRACT. We describe differences between the commonly used version of the U.S. Census of Manufactures available at the FSRDCs and what establishments themselves report. We find substantially more extreme values of productivity (both measured TFPQ and TFPR) in the originally reported data. Furthermore, there is substantially more dispersion of commonly used measures such as the standard deviation or the interquartile range. To capture covariance, we follow the methodology of Hsieh and Klenow (2009). Measured allocative efficiency is substantially higher in the cleaned data than the raw data - 4x higher in 2002, 20x in 2007, and 80x in 2012. Many of the important editing strategies at the Census, including industry analysts’ manual edits and edits using tax records, are infeasible in non-U.S. datasets. We reanalyze cross-country comparisons starting from unprocessed U.S. and Indian data and using common data cleaning strategies: a simple trimming-outliers approach and a new Bayesian approach for editing and imputation. Under both methods there is little evidence that allocative efficiency is significantly worse for formal firms in India than in the United States. If anything, we find the opposite.

ABSTRACT. In modern trade theory, foreign firms are generally assumed to be more productive than those that only serve the domestic market. As a consequence, there has been a renewed interest in the trade literature on analyzing differences in firm productivity across countries and sectors and the role policy barriers play in accentuating these differences. A significant hurdle impeding this research endeavor, however, is finding readily accessible firm level databases that cover a large set of countries, industries, and firms.

In conjunction with firm-level productivity analysis, there has also been a greater push in the trade literature to understand trade and investment in services, which have become significant contributors to global commerce. Following these developments in the literature, this paper focuses on the question of whether foreign affiliates of multinational services firms are more productive than their domestically owned counterparts and whether trade barriers can explain such differences in productivity.

By combining a rich firm-level dataset with measures of trade barriers in services sectors at the country-and-sector level, we connect productivity of domestic and foreign services firms with the regulatory barriers faced by foreign firms to enter the market. Firm-level data is obtained from Bureau van Dijk's Orbis database, allowing us to construct a measure of total factor productivity (TFP) for services firms at the country and sector-level. Notably, the Orbis database provides information on firm ownership, and so we are able to distinguish between domestic firms and foreign-owned affiliates in a particular market. The OECD's Services Trade Restrictiveness Index (STRI) is used to determine the costs associated with setting up a foreign affiliate in a particular country and sector.

Overall, at the firm level, we find a negative and significant relationship between services trade barriers and productivity for domestic firms, while foreign firms see a positive relationship between services trade barriers and productivity. These findings indicate that policy barriers may help explain difference in the average productivity across domestic and foreign firms, since only the most productive foreign firms are able to incur the significant costs associated with regulatory barriers to serve the domestic markets through their foreign affiliates.

ABSTRACT. The Broadband Initiatives Program (BIP) is the largest USDA program to promote broadband development in rural areas. Enacted as a part of the 2009 American Recovery and Reinvestment Act, in 2010 BIP funded 300 broadband infrastructure projects serving almost every state, with a total value of $2.9 billion (net, excluding projects that were approved but not completed). Almost all of these projects were “last mile” projects, and most provided fiber-to-the-premises. This study investigates the impacts of these projects on broadband availability and house sale prices using Rural Utilities Service data on the program, FCC data on broadband availability, and CoreLogic property tax and deed data. A combination of matching and difference-in-difference (DD) regression methods is used to identify the impacts of these projects. For the analysis of impacts on house prices, prior trends in house prices in project service areas and matched non-project areas are used to test the parallel trends assumption of the DD estimation.

ABSTRACT. Tackling climate change and environment sustainability is a challenge that has emerged as an urban policy issue to be addressed across cities. The buildings sector is an important factor in environmental sustainability. In 2014, buildings accounted for 31% of total global final energy use, 8% of energy-related CO2 emissions and 54% of final electricity demand. More specifically, the United States Green Building Council (USGBC) claims that together the residential and commercial sectors account for 39% of total CO2 emissions per year and are responsible for more such emissions than any other country in the world with the exception of China. Consequently, pathways have been designed to mitigate and adapt to climate change and its impacts on cities. One such national response is the introduction of the USGBC’s Leadership in Energy and Environment Design (LEED) green building rating system in 2000.

In 2006 the Government of the District of Columbia (DC) enacted The Green Building Act, which requires that new privately-owned nonresidential buildings with 50,000 or more square feet of gross floor area meet the USGBC’s LEED certification standards. As a result, it is presumed that nearly all large private non-residential buildings built in recent years are green in certain respects. Surprisingly, residential buildings have also been adhering to LEED-certification standards, which suggests that residential developers have recognized the economic value and opportunities of going green and have decided to emulate this practice.

Despite LEED becoming a leading design standard for green buildings in the U.S., the existing literature on this topic is limited in scope. Most if not all studies focus solely on the price effects of certification for commercial office space. One of the main selling points for building green is lower expenses. However, even more understudied is the effect of green-labeling on building expenses. Using a hedonic pricing model, this study investigates the effect of voluntary LEED-certification on effective rent and affordability for large multi-family residential buildings in DC. It also examines to what extent lower operating expenses contribute to rent premiums. My research appears to be the first that separates energy costs from total operating expenses, which allows me to identify to what degree energy efficiency plays a role in rent determination.

I assemble a unique dataset consisting of micro-level building and neighborhood characteristics to provide a systematic and empirical analysis of environmentally sustainable buildings on marketplace outcomes. Data were sourced from CoStar and the DC government’s 2018 income and expense, individual income tax and real property databases.

This study finds that LEED buildings have lower operating expenses, higher rent and higher-income tenants than their conventional counterparts. More specifically, only Silver- and Gold-level buildings save on operating expenses and command higher rents. The results show that the economic benefits of environmental sustainability are not equal across different certification levels. With regard to savings on utilities expenses, the level of certification is an important determining factor for cost reductions. Lastly, this research finds that LEED buildings located in lower-income zipcodes that are predominantly black, charge lower rents.

ABSTRACT. This paper examines the causal impact of international capital flows from China on one of the least accurately measured and understood global safe haven assets: U.S. residential real estate. We demonstrate using aggregate capital flows data that the recent rise in unrecorded inflows in the U.S. balance of payments is likely attributable to inflows being used to purchase U.S. residential properties, mainly to inflows originating from China. We then exploit a novel, direct measure of Chinese demand for U.S. residential properties at the local level by using a web traffic dataset from a real estate listing website that specializes in marketing foreign residential properties to users based in Mainland China. With a difference-in-difference matching framework, we find that house prices in China-exposed areas of major U.S. cities have on average grown seven percentage points faster than in similar neighborhoods with low exposure to Chinese buyers over the period 2010-2016, when the U.S. experienced two episodes of surges in safe haven flows from China. We then show that the time variation in the average excess price growth in China exposed-areas comoves closely with macro-level measures of U.S. capital inflows from China. Following periods of economic stress in China, inflows from China to the U.S. jump and the price growth gap widens, suggesting that Chinese households view U.S. housing as a safe haven asset.

ABSTRACT. The effective cost of homeownership depends on the tax code. I derive a sufficient statistic formula for the effect of tax reform on house prices, and use it, together with nearly 4 million simulated tax returns, to measure the response of house prices to the Tax Cuts and Jobs Act of 2017. The measured price decline from the disincentives to itemize federal deductions is larger than the increase from higher disposable income. Measured house price declines in 269 metropolitan areas, are between 0 and 7 percent, 3 percent on average, and correlate with house price changes observed in 2018.