Economists have for a decade or so theorized that moving productive inputs like labor and capital into the firms that make the best use of them is a prime engine of economic growth. But measuring how well this allocation is taking place across economies is a daunting empirical task. Nonetheless, a new Stigler Center working paper by Lenzu and Manaresi tries to do precisely that.
Economies are characterized by a large degree of firm heterogeneity. Firms differ in terms of size (capital and labor endowment) and in terms of productivity (their ability to efficiently combine inputs in production). A well-functioning market should be able to allocate capital and labor to the most productive firms that provide the highest value use; otherwise inputs would be misallocated.
Over the last decade, a growing number of studies have argued that input misallocation may explain a large share of cross-country differences in aggregate productivity (Banerjee and Duflo 2005; Hsieh and Klenow 2009), prompting interest from policymakers on this topic. In fact, policy-relevant frictions—such as taxes, regulations, lack of competition, financial markets imperfections, and institutions—are all believed to be drivers of misallocation, as they prevent resources from freely flowing across firms in the economy. As a result of these frictions, some firms might be smaller and others larger than their “socially efficient” size.
Despite the theoretical appeal of the concept of resource misallocation, empirically identifying which firms are inefficiently utilizing inputs, how large the aggregate cost of misallocation is, and which frictions are most relevant for it is a daunting task. Most previous literature addressed these issues by imposing strong theoretical assumptions on the data. These assumptions are not necessarily met in reality and have considerable impact on the empirical results (Haltiwanger et al. 2017). The availability of detailed information on firms’ production decisions and on their user cost of capital and labor, and the implementation of appropriate econometric techniques, allows us to develop a new metric that, overcoming some of the limitations of previous literature, lets us estimate the efficiency of capital and labor accumulation of each firm. We show that this metric can be readily used by researchers and policymakers, as it can be directly related to specific frictions in credit and labor markets, and easily aggregated across firms to quantify the impact of such frictions on aggregate productivity and output.
We conduct our study in Italy, analyzing approximately 4,000 non-financial businesses (mostly privately held) that operated across several industries between 1997 and 2013. Aside from data availability, the Italian economy is an interesting case study per se, as it has been facing two “lost decades” of economic growth and there is currently a lively debate about whether input misallocation may play a role in explaining this unsatisfactory performance. Using our measure, we calculate that aggregate output of the Italian corporate sector could be 3 to 4 percent higher were resources reallocated from firms that over-utilize them to the most productive producers that lack them. Credit and labor market frictions appear to be two important drivers of resource misallocation.
Measuring Efficiency of Input Utilization: The Distribution of MRP-Cost Gaps
In a frictionless economy, firms should equate the marginal revenue product of each input (henceforth, MRP) to its user cost. The difference between these two quantities, which we call the MRP-cost gap, can be considered a useful metric to estimate whether resources are over- or under-accumulated at the firm level. A positive MRP-cost gap for a given input (capital or labor) suggests that it may be efficient to increase its utilization, lowering the MRP of that input down to its user cost. Conversely, a negative MRP-cost gap signals over-utilization of the input, as the user costs exceeds its marginal return.
We assemble a comprehensive bank-firm-employee matched database that contains information on firm-specific wages, borrowing costs, balance sheets, and bank credit for all non-financial corporations active in Italy between 1997 and 2013. This data allows us to estimate MRP-cost gaps of both capital and labor inputs for every firm-year observation.
In Italy, the distributions of MRP-cost gaps turn out to be dispersed and highly right-skewed, which is indicative of suboptimal investment and employment policies. For example, the average capital gap is 37 percent; the average labor gap amounts to 9,000 euros. The 90-10 percentile differences are almost 3 times larger. These differences suggest that significant output gains could be attained through resource reallocation.
Several phenomena contribute to the size and dispersion of realized MRP-cost gaps. Economic uncertainty and real adjustment costs naturally drive a wedge between realized marginal revenue products and user costs. But policy-relevant frictions can also rationalize the above-mentioned distribution of MRP-cost gaps. These include credit market frictions, market power, heavy taxation, bureaucratic costs, tariffs and subsidies, and frictions in the market of corporate ownership and control.((Pellegrino and Zingales (2017) trace sluggish Italian productivity growth to the lack of meritocracy in the selection and rewarding of managers of Italian firms that prevented them to take advantage of the ITC revolution. See Bugamelli and Lotti (2018) for a critical review of the existing explanations of the “lost two decades.”)) In this column, we describe the role of credit and labor market frictions in explaining the variation in MRP-cost gaps observed in the data.
MRP-Cost Gaps of Capital and Credit Market Frictions
On the capital side, we study the response of the MRP-cost of capital to idiosyncratic shocks to credit supply. Consistent with the theoretical prediction that variation in gaps across firms captures heterogeneous shadow costs of capital, we find that, all else equal, a firm’s MRP-cost shrinks following an episode of credit expansion, whereas we observe MRP-cost gaps widening in response to a negative credit-supply shock (figure 1). The response is stronger for firms that had a positive MRP-cost gap before the shock (under-capitalized firms), especially in response to credit expansions, and the sensitivity of gaps to credit shocks is stronger for more productive firms. On the contrary, the MRP-cost gaps of firms with zero or negative MRP-cost gaps (i.e., those that operate with a capital endowment close to or above target) show a small or no response to credit-supply shocks, either positive or negative. This analysis suggests that MRP-cost gaps are indeed capable of identifying in the data those firms that are more likely to be capital constrained, and that credit constraints are a significant factor preventing under-capitalized firms from undertaking profitable investments.
MRP-Cost Gaps of Labor and Labor Market Frictions
On the labor side, we analyze the impact of a labor market regulation that, until 2015, imposed firing costs that varied as a function of firm size: large severance payments were required of employers with more than 15 employees and significantly smaller payments of firms with 15 employees or less (Garibaldi and Violante 2005, Sestito and Viviano 2016). Size-dependent firing costs are an adjustment cost that generates variation in MRPs and, if not undone by properly designed wage contracts (Lazear 1990), generate misallocation. We find that as the 15-employee threshold is approached to the left, the average gap between the marginal revenue product of labor and wages increases, indicating an inefficiently low labor demand by the set of firms that choose to operate below the regulatory threshold. The presence of government-mandated severance payments also affects firms above this threshold, inducing them to operate with a smaller labor force than the one they might have chosen in the absence of the size-dependent regulation.
The impact of the size-dependent regulation can also be appreciated in a dynamic context: when facing a productivity shock, firms that are just below the 15-employee threshold are less likely to hire workers to avoid incurring the increase in firing costs. These results are consistent with the hypothesis that the government-mandated severance payments curb economic growth by discouraging firms from increasing their size despite the growth opportunities that might be available.
Boom-and-bust cycles in credit markets can affect aggregate TFP due to a deterioration in the efficiency of capital allocation.
Finally, our paper casts light on the aggregate implications of resource misallocation in Italy. We use the estimated MRP-cost gaps to assess how micro-level distortions in investment and employment choices by firms translate into aggregate output and TFP losses. We calculate that, in any given year, aggregate TFP and output of the Italian corporate sector could be 3 to 4 percent higher following a reallocation of production factors, by taking resources away from firms that over-utilize them, and redistributing these resources to the most productive producers who are lacking them.
Substantial empirical evidence documents a significant decline in aggregate TFP and output during economic downturns, particularly following episodes of financial instability (Jermann and Quadrini 2012). An open question is whether a change in the scope of resource misallocation, on top of (or instead of) technology shocks, contributes to explaining the co-integration of business-cycle fluctuations and aggregate TFP. Our results speak to this question by showing that, indeed, boom-and-bust cycles in credit markets can affect aggregate TFP due to a deterioration in the efficiency of capital allocation. We find that gains from reallocation are one-third higher during periods characterized by financial instability—the financial crisis (2008–9) and following the outbreak of the sovereign debt crisis (2010–13)—compared to those estimated during the 1997–2004 period (figure 2). The scope of misallocation is found to be more severe in services and construction than in manufacturing, and more severe in the southern regions when compared to northern and central regions of the country (figure 3), mimicking the differences in the quality of institutions and markets (Putnam 1994).
Francesco Manaresi is a researcher at the Bank of Italy, Structural Economic Analysis and Labour Market Division.
Simone Lenzu is a PhD Student at the Department of Economics at the University of Chicago and Student Fellow of the Stigler Center.
Banerjee, A. V. and E. Duflo (2005). Growth theory through the lens of development economics. Handbook of economic growth 1, 473–552.
Bugamelli, M. and F. Lotti, eds. (2018). Productivity growth in Italy: a tale of a slow-motion change. Bank of Italy Occasional Papers, no. 422.
Garibaldi, P. and G. L. Violante (2005). The employment effects of severance payments with wage rigidities. The Economic Journal 115(506), 799–832.
Haltiwanger, J., R. Kulick, and C. Syverson (2017). Misallocation measures: The distortion that ate the residual. Working Paper.
Hsieh, C.-T. and P. J. Klenow (2009). Misallocation and manufacturing tfp in china and india. The Quarterly Journal of Economics 124(4), 1403–1448.
Jermann, U. and V. Quadrini (2012). Macroeconomic effects of financial shocks. The American Economic Review 102(1), 238–271.
Lazear, E. P. (1990). Job security provisions and employment. The Quarterly Journal of Economics 105 (3), 699–726.
Sestito, Paolo, and Eliana Viviano (2016). Hiring incentives and/or firing cost reduction? Evaluating the impact of the 2015 policies on the Italian labour market. Working paper.
Putnam, R. D., R. Leonardi, and R. Y. Nanetti (1994). Making democracy work: Civic traditions in modern Italy. Princeton university press.
Sestito, P. and E. Viviano (2016), Hiring incentives and/or firing cost reduction? Evaluating the impact of the 2015 policies on the Italian labour market. Bank of Italy Occasional Papers, no. 325.
Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.