Journal of Urban Economics 79, 2014, 20-38 | With Moretti
We evaluate the effects of state-provided financial incentives for biotech companies, which are part of a growing trend of placed-based policies designed to spur innovation clusters. We estimate that the adoption of subsidies for biotech employers by a state raises the number of star biotech scientists in that state by about 15 percent over a three year period. A 10% decline in the user cost of capital induced by an increase in R&D tax incentives raises the number of stars by 22%. Most of the gains are due to the relocation of star scientist to adopting states, with limited effect on the productivity of incumbent scientists already in the state. The gains are concentrated among private sector inventors. We uncover little effect of subsidies on academic researchers, consistent with the fact that their incentives are unaffected. Our estimates indicate that the effect on overall employment in the biotech sector is of comparable magnitude to that on star scientists. Consistent with a model where workers are fairly mobile across states, we find limited effects on salaries in the industry. We uncover large effects on employment in the non-traded sector due to a sizable multiplier effect, with the largest impact on employment in construction and retail. Finally, we find limited evidence of a displacement effect on states that are geographically close, or states that economically close as measured by migration flows.
Infrastructure Spending as Fiscal Stimulus: Assessing the Evidence
Review of Economics and Institutions 5, Winter 2014, 1-24
Transportation spending often plays a prominent role in government efforts to stimulate the economy during downturns. Yet, despite the frequent use of transportation spending as a form of fiscal stimulus, there is little known about its short- or medium-run effectiveness. Does it translate quickly into higher emploiyment and economic activity or does it impact the economy only slowly over time? This paper reviews the empirical finding in the literature for the United States and other developed economies and compares the effects of transportation spending to those of other types of government spending.
Review of Economics and Statistics 95(5), December 2013, 1480-1500 | With Daly and Johnson
We assess the importance of interpersonal income comparisons using data on suicide deaths. We examine whether suicide risk is related to others’ income, holding own income and other individual and environmental factors fixed. We estimate models of the suicide hazard using two independent data sets: (1) the National Longitudinal Mortality Study and (2) the National Center for Health Statistics’ Multiple Cause of Death Files combined with the 5 percent Public Use Micro Sample of the 1990 decennial census. Results from both data sources show that, controlling for own income and individual characteristics, individual suicide risk rises with others’ income.
In NBER Macroeconomic Annual 2012, 27, ed. by Jonathan Parker and Michael Woodford | University of Chicago Press, 2013. 89-142 | With Leduc
We examine the dynamic macroeconomic effects of public infrastructure investment both theoretically and empirically, using a novel data set we compiled on various highway spending measures. Relying on the institutional design of federal grant distributions among states, we construct a measure of government highway spending shocks that captures revisions in expectations about future government investment. We find that shocks to federal highway funding has a positive effect on local GDP both on impact and after 6 to 8 years, with the impact effect coming from shocks during (local) recessions. However, we find no permanent effect (as of 10 years after the shock). Similar impulse responses are found in a number of other macroeconomic variables. The transmission channel for these responses appears to be through initial funding leading to building, over several years, of public highway capital which then temporarily boosts private sector productivity and local demand. To help interpret these findings, we develop an open economy New Keynesian model with productive public capital in which regions are part of a monetary and fiscal union. We show that the presence of productive public capital in this model can yield impulse responses with the same qualitative pattern that we find empirically.
American Economic Journal: Economic Policy 4(3), August 2012, 251-282
This paper estimates the “jobs multiplier” of fiscal spending using the state-level allocations of federal stimulus funds from the American Recovery and Reinvestment Act (ARRA) of 2009. Because the level and timing of stimulus funds that a state receives are potentially endogenous, I exploit the fact that most of these funds were allocated according to exogenous formula factors such as the number of federal highway miles in a state or its youth share of population. The estimates imply that each million dollars of announced stimulus in a state was associated with approximately eight jobs created or saved in that state as of one year after the ARRA was enacted. The implied cost per job is about $125,000.
Journal of Economic Behavior and Organization 80(3), 2011, 435-442 | With Daly, Oswald, and Wu
Suicide kills more Americans than die in motor accidents. Its causes remain poorly understood. We suggest in this paper that the level of others’ happiness may be a risk factor for suicide (although one’s own happiness likely protects one from suicide). Using U.S. and international data, the paper provides evidence for a paradox: the happiest places tend to have the highest suicide rates. The analysis appears to be the first published study to be able to combine rich individual-level data sets–one on life satisfaction in a newly available random sample of 1.3 million Americans and another on suicide decisions among an independent random sample of about 1 million Americans–to establish this dark-contrasts paradox in a consistent way across U.S. states. The study also replicates the finding for the Western industrialized nations. The paradox, which holds individual characteristics constant, is not an artifact of population composition or confounding factors (or of the ecological fallacy). We conclude with a discussion of the possible role of relative comparisons of utility.
National Tax Journal 63(4), 2010, 967-994 | With Chirinko
The standard model of strategic tax competition–the noncooperative tax-setting behavior of jurisdictions competing for a mobile capital tax base–assumes that government policymakers are perfectly benevolent, acting solely to maximize the utility of the representative resident in their jurisdiction. We depart from this assumption by allowing for the possibility that policymakers also may be influenced by the rent-seeking (lobbying) behavior of businesses. Businesses recognize the factors affecting policymakers’ welfare and may make campaign contributions to influence tax policy. This extension to the standard strategic tax competition model implies that business contributions may affect not only the levels of equilibrium tax rates but also the slope of the tax reaction function between jurisdictions. Thus, business campaign contributions may directly influence business tax rates, as well as indirectly shape tax competition, and enhance or retard the mobility of capital across jurisdictions.
Based on a panel of 48 U.S. states and unique data on business campaign contributions, our empirical work uncovers four key results. First, we document a significant direct effect of business contributions on tax policy. Second, the economic value of a $1 business campaign contribution in terms of lower state corporate taxes is approximately $6.65. Third, the slope of the reaction function between tax policy in a given state and the tax policies of its competitive states is negative, and this slope is robust to business campaign contributions. Fourth, we document the sensitivity of the empirical results to state effects.
Review of Economics and Statistics 91(2), 2009, 431-436
The proliferation of R&D tax incentives among U.S. states in recent decades raises two important questions: (1) Are these tax incentives effective in achieving their stated objective, to increase R&D spending within the state? (2) To the extent the incentives do increase R&D within the state, how much of this increase is due to drawing R&D away from other states? In short, this paper answers (1) “yes” and (2) “nearly all,” with the implication that the net national effect of R&D tax incentives on R&D spending is near zero. The paper addresses these questions by exploiting the cross-sectional and time-series variation in R&D tax credits, and in turn the user cost of R&D, among U.S. states from 1981-2004 to estimate an augmented version of the standard R&D factor demand model. I estimate an in-state user cost elasticity (UCE) around -2.5 (in the long-run), consistent with previous studies of the R&D cost elasticity. However, the R&D elasticity with respect to costs in neighboring states, which has not previously been investigated, is estimated to be around +2.5, suggesting a zero-sum game among states and raising concerns about the efficiency of state R&D credits from the standpoint of national social welfare.
Journal of European Economic Association 7, 2009, 539-549 | With Daly
The use of subjective well-being (SWB) data for investigating the nature of individual preferences has increased tremendously in recent years. There has been much debate about the cross-sectional and time-series patterns found in these data, particularly with respect to the relationship between SWB and relative status. Part of this debate concerns how well SWB data measures true utility or preferences. In a recent paper, Daly, Wilson, and Johnson (2007) propose using data on suicide as a revealed preference (outcome-based) measure of well-being and find strong evidence that reference-group income negatively affects suicide risk. In this paper, we compare and contrast the empirical patterns of SWB and suicide data. We find that the two have very little in common in aggregate data (time series and cross-sectional), but have a strikingly strong relationship in terms of their determinants in individual-level, multivariate regressions. This latter result cross-validates suicide and SWB micro data as useful and complementary indicators of latent utility.
Journal of Business and Economic Statistics 27(1), January 2009, 52-70
This article explores the relationship between capital composition and productivity using a unique,
detailed dataset on firm investment in the United States in the late 1990s. I develop a methodology for
estimating the separate effects of multiple capital types in a production function framework. I back out
the implied marginal products of each capital type and compare these with rental price data. I find that
although most capital types earned normal returns, information and communications technology capital
goods had marginal products substantially above their rental prices. The article also provides evidence of
complementarities and substitutabilities among capital types and between capital types and labor.
Journal of Public Economics 92(12), December 2008, 2,362-2,384 | With Chirinko
Over the past four decades, state investment tax incentives have proliferated. This emergence of state investment tax credits (ITC) and other investment tax incentives raises two important questions: (1) Are these tax incentives effective in achieving their stated objective, to increase investment within the state?; (2) To the extent these incentives raise investment within the state, how much of this increase is due to investment drawn away from other states? To begin to answer these questions, we construct a detailed panel data set for 48 states for 20-plus years. The dataset contains series on output and capital, their relative prices, and establishment counts. The effects of tax variables on capital formation and establishments are measured by the Jorgensonian user cost of capital that depends in a nonlinear manner on federal and state tax variables. Cross-jurisdictional differences in state investment tax credits and state corporate tax rates entering the user cost, combined with a panel that is long in the time dimension, are key to identifying the effectiveness of state investment incentives. Two models are estimated. The Capital Demand Model is motivated by the first-order condition for a profit-maximizing firm and relates at the state level the capital/output ratio to the relative user cost of capital. The Twin-Counties Model exploits both the spatial breaks (“discontinuities”) in tax policy at state borders and our panel data set to relate at the county level the relative user cost to the location of manufacturing establishments. Using the Capital Demand Model, we find that own-state capital formation is substantially increased by tax-induced reductions in the own-state price of capital and, more interestingly, substantially decreased by tax-induced reductions in the price of capital in competitive-states. Similarly, using our Twin-Counties Model, we find that county manufacturing establishment counts around state borders are higher on the side of the border with the lower price of capital, but the difference is economically small, suggesting that establishments are much less mobile than overall capital. Extensions of the Capital Demand Model also reveal that state capital tax policy appears to be a zero-sum game among the states in that an equiproportionate increase in own-state and competitive-states user costs tends to have no effect on own-state capital formation.
Investment Behavior of U.S. Firms over Heterogeneous Capital Goods: A Snapshot
Review of Income and Wealth 54(2) , June 2008, 269-278
Recent research has indicated that investment in certain capital types, such as computers, has fostered accelerated productivity growth and enabled a fundamental reorganization of the workplace. However, remarkably little is known about the composition of investment at the micro level. This short paper takes an important first step in filling this knowledge gap by looking at the newly available micro data from the 1998 Annual Capital Expenditure Survey (ACES), a sample of roughly 30,000 firms drawn from the private, nonfarm economy. The paper establishes a number of stylized facts.
Among other things, I find that in contrast to aggregate data the typical firm tends to concentrate its capital expenditures in a very limited number of capital types, though which types are chosen varies greatly from firm to firm. In addition, computers account for a significantly larger share of firms’ incremental investment than they do of lumpy investment.
Micro and Macro Data Integration: The Case of Capital
In A New Architecture for The U.S. National Accounts, NBER/CRIW Volume, ed. by Jorgenson, Landefeld, and Nordhaus | Chicago: University of Chicago, 2006. 541-609 | With Becker, Jarmin, Klimek, and Haltiwanger
Micro and macro data integration should be an objective of economic measurement as it is
clearly advantageous to have internally consistent measurement at all levels of aggregation–firm, industry and aggregate. In spite of the apparently compelling arguments, there are few
measures of business activity that achieve anything close to micro/macro data internal
consistency. The measures of business activity that are arguably the worst on this dimension are
capital stocks and flows. In this paper, we document, quantify and analyze the widely different
approaches to the measurement of capital from the aggregate (top down) and micro (bottom up)
perspectives. We find that recent developments in data collection permit improved integration of
the top down and bottom up approaches. We develop a prototype hybrid method that exploits
these data to improve micro/macro data internal consistency in a manner that could potentially
lead to substantially improved measures of capital stocks and flows at the industry level. We
also explore the properties of the micro distribution of investment. In spite of substantial data
and associated measurement limitations, we show that the micro distributions of investment
exhibit properties that are of interest to both micro and macro analysts of investment behavior.
These findings help highlight some of the potential benefits of micro/macro data integration.
Review of Economic Dynamics 7(1), January 2004, 1-26 | With Sakellaris
We estimate the rate of embodied technological change directly from plant-level manufacturing data on current output and input choices along with histories on their vintages of equipment investment. Our estimates range between 8 percent and 17 percent for the typical U.S. manufacturing plant during the years 1972-1996. Any number in this range is substantially larger than is conventionally accepted with some important implications. First, the role of investment-specific technological change as an engine of growth is even larger than previously estimated. Second, existing producer durable price indexes do not adequately account for quality change. As a result, measured capital stock growth is biased. Third, if accurate, the Hulten and Wykoff (1981) economic depreciation rates may primarily reflect obsolescence.
Journal of Monetary Economics 51(1), January 2004, 1-32 | With Caselli
We look at disaggregated imports of various types of equipment to make inferences on cross-country differences in the composition of equipment investment. We make three contributions. First, we document strikingly large differences in investment composition. Second, we explain the differences as being based on each equipment type’s degree of complementarity with other factors whose abundance differs across countries. Third, we show that the composition of capital has the potential to account for some of the large observed differences in total factor productivity across countries.
Economic Systems Research 15(3), September 2003, 371-398
In this paper, I develop a regression-based system of labor productivity
equations that account for capital-embodied technological change, and I incorporate this system into IDLIFT, a structural, macroeconomic input-output model of the U.S. economy. Builders of regression-based forecasting models have long had difficulty finding labor productivity equations that exhibit the “Solowian” property that movements in investment should cause accompanying movements in labor productivity. The production theory developed by Solow and others dictates that this causation is driven by the effect of traditional capital deepening as well as technological change embodied in capital. Lack of measurement of the latter has hampered the ability of researchers to estimate properly the productivity-investment relationship. Recent research by Wilson (2001) has alleviated this difficulty by estimating industry-level embodied technological change. In this paper, I utilize those estimates to construct capital stocks adjusted for technological change and then use these adjusted stocks to estimate Solow-type labor productivity equations. It is shown that replacing IDLIFT’s former productivity equations, based on changes in output and time trends, with the new equations, results in a convergence between the dynamic behavior of the
model and that predicted by traditional (Solowian) production theory.
Review of Economic Dynamics 5(2), April 2002, 285-317
This paper provides an exploratory analysis of whether data on the research and development (R&D) spending directed at particular technological/product fields can be used to measure industry-level capital-embodied technological change. Evidence from the patent literature suggests that the R&D directed at a product, as the main input into the “innovation” production function, is proportional to the value of the innovations in that product. I confirm this hypothesis by showing that the decline in the relative price of a good is positively correlated with the R&D directed at that product. The hypothesis implies that the technological change, or innovation, embodied in an industry’s capital is proportional to the R&D that is done (“upstream”) by the economy as a whole on each of the capital goods that a (“downstream”) industry purchases. Using R&D data from the National Science Foundation, I construct measures of capital-embodied R&D. I find they have a strong effect on conventionally measured total-factor productivity growth, a phenomenon that seems to be due partly to the mismeasurement of quality change in the capital stock and partly to a positive correlation between embodied and disembodied technological change. Finally, I find the cross-industry variation in empirical estimates of embodied technological change accord with the cross-industry variation in embodied R&D.
Estimating Returns to Scale: Lo, Still No Balance
Journal of Macroeconomics 22(2), Spring 2000, 285-314
Using detailed data and a unique instrument set, estimates of returns to scale in U.S. manufacturing were obtained at various levels of aggregation. With a few key exceptions, empirical puzzles previously found are confirmed and further investigated. One implication of these findings is that there is essentially no evidence of large increasing returns necessary in many recent macro models. Also, the finding of significant heterogeneity among 4-digit sectors casts doubt on the use of the representative firm paradigm in macroeconomic modeling. These results suggest the presence of vast reallocation effects among firms within sectors, manifesting itself as decreasing returns.