VoxEU Column International trade

Why trade and finance increase inequality across firms and workers

Inequality, both in firm revenues and wages, varies greatly across sectors, has increased over time and is positively correlated to export opportunities. To explain these observations, this column propose a new theory in which firms’ investment at the entry stage affects the variance of the possible realisations of their productivity. It suggests that export opportunities and competition, besides reallocating resources across existing firms, increase the value of technological heterogeneity. This hints to a new powerful channel through which globalisation is making firms and wages more unequal.   

Current research in international trade and macroeconomics puts firm-level heterogeneity at a centre stage. As documented by a large literature, firms differ in size and productivity even within narrowly defined industries (e.g. Syverson 2004), and the size distribution of firms is highly skewed (e.g. Axtell 2001). Moreover, firm heterogeneity has been shown to be important for understanding trade, productivity and wage inequality. Yet, besides some well-known aggregate statistics, there is scant systematic evidence on how firm heterogeneity varies across industries, countries and time, and even scarcer theoretical explanations. In new research (Bonfiglioli et al. 2015a,b) we take a first step towards filling these major gaps.

Firm heterogeneity across US industries and over time

In our first paper (Bonfiglioli et al. 2015a) we document some little-known facts regarding how a synthetic and scale-independent measure of firm heterogeneity, the standard deviation of the log of sales, varies across industries and time in the US economy. Using establishment-level data from the US Census Bureau, we show that the dispersion of sales varies by a factor of 10 across 453 6-digit NAICS industries, and that it has increased on average by 11.8 percent between 1997 and 2007. We then explore how firm heterogeneity depends on a number of industry characteristics. Regressions provide robust evidence that heterogeneity in sales and labour productivity is positively correlated with average sales per establishment at the industry level, and that it increases following an exogenous rise in industry-level export intensity. Back of the envelope calculations suggest that the observed increase in export intensity over the sample period (21 percent) explains between 13 and 52 percent of the increase in sales dispersion over 1997-2007. These results are robust to excluding small and large firms and controlling for the number of firms, and hence do not seem driven by the granularity of the data and selection effects.

Betting on exports: Why trade can increase productivity differences

What explains firm heterogeneity? Unfortunately, leading models generate productivity differences from exogenous processes and hence cannot answer properly this question. We therefore propose a new theory in which observed heterogeneity stems from technological choices made when new products are introduced. To do so, we develop a multi-industry model à la Melitz (2003) in which endogenous investment decisions at the entry stage affect the variance of the possible realisations of productivity. The model formalises the idea that firms face a trade-off between investing in smaller innovation projects with less variable outcomes and more ambitious projects with higher variance. Such a trade-off seems to well describe the different innovation strategies pursued, for instance, by Apple and Dell. While the former invested large amounts in very innovative projects that ultimately led to the development of highly successful products such as the iPhone and the iPad, Dell kept pursuing marginal innovations which lad to steady, yet smaller sales.

A key feature of the model is that the possibility to exit insures firms from bad realisations and increases the value of drawing productivity from a more dispersed distribution. This generates two main predictions. First, export opportunities, by shifting expected profits to the tail and raising the exit cutoff, increase the value of drawing productivity from a more dispersed distribution. Hence, the chance of winning the extra prize of exporting induces firms to bet on bigger innovation projects with more spread out outcomes, thereby generating more heterogeneity in equilibrium. Second, higher entry costs soften competition, lower the exit cutoff and hence lower the value of technological heterogeneity.1

We then extend the model to show how firm heterogeneity can map into wage inequality, as in Helpman et al. (2010). When more productive firms pay higher wages, we obtain another novel result: besides introducing an exporter wage premium, trade amplifies wage dispersion also among employees of non-exporters by inducing firms to invest in technologies with a higher variance and hence making them more unequal ex-post.

US evidence: Exports, wage dispersion and innovation

Going back to the data, we use individual-level US wages over 1997-2007 to show that measures of wage dispersion vary with industry characteristics in a way that mirrors remarkably well the pattern found for the dispersion of sales. In particular, export opportunities increase significantly wage inequality at the industry level. We then provide a first attempt at testing a specific mechanism of our model, namely, that export opportunities increase firm heterogeneity by fostering investment in innovation. To do so, we follow Aghion et al. (2015) in switching to geographic data and use patent counts to build a measure of innovation intensity for a panel of US states over 1989-2007. With this data, we document two sets of results. First, consistently with the model, innovation intensity is strongly correlated with export opportunities, now measured at the state-level as in Autor, Dorn and Hanson (2013). Second, the dispersion of firms' sales at the state level is strongly correlated with innovation intensity.

Cross-country evidence: Exports, financial development and firm heterogeneity

In our second paper (Bonfiglioli et al. 2015b) we extend the model to study how financial frictions affect firm-level heterogeneity and trade. By softening competition, financial frictions lower the value of investing in bigger projects with more dispersed outcomes and hence heterogeneity, especially in financially vulnerable industries. We also show that, through this channel, credit frictions hinder the volume of exports both along the intensive and the extensive margin. We then empirically assess the predictions of the model using extremely detailed data on US imports of roughly 15,000 (10-digit) products from 119 countries and 365 manufacturing industries over 1989-2006. Starting from almost 4 million observations at the country-product-year level, we build measures of sales dispersion for each country, industry and year.

With this uniquely rich data set, we study how sales dispersion depends on financial frictions, accounted for by indicators of country credit supply and sector financial vulnerability à la Rajan and Zingales (1998), and on export opportunities, captured by measures of country-sector comparative advantage à la Romalis (2004). We find two main results. First, consistent with our model, financial development increases sales dispersion, especially in more financially vulnerable industries; export opportunities also make the distribution of sales more spread out. Second, our mechanism is quantitatively important for explaining the effect of financial development and factor endowments on export sales. In specifications for total exports, number of exported products, and exports per product, sales dispersion is highly significant and reduces the coefficients on financial variables and factor endowments by 20-60%. These findings shed new light on why credit frictions seem to be a major impediment to trade (Manova 2013).

Conclusions

Why firms differ so much in sales and productivity is one of the main open questions in the fields of international trade, macroeconomics and economic development. Existing empirical studies are limited to few countries or sectors, and theoretical models often derive productivity differences from exogenous processes.

Our research overcomes both limitations. Our results suggest that export opportunities and competition, besides reallocating resources across existing firms, also increase the value of technological heterogeneity. This hints to a new powerful channel through which globalisation is making firms and wages more unequal.

References

Aghion, Philippe, Ufuk Akcigit, Antonin Bergeaud, Richard Blundell, David Hémous (2015). "Innovation and Top Income Inequality." NBER Working Paper No. 21247.

Autor, David H., David Dorn and Gordon H. Hanson (2013). "The China Syndrome: Local Labor Market Effects of Import Competition in the US." American Economic Review, 103: 2121-2168.

Axtell, Robert L. (2001). "Zipf Distribution of U. S. Firm Sizes." Science, 293: 1818-1820.

Bonfiglioli, Alessandra, Rosario Crinò and Gino Gancia (2015a). "Betting on Exports: Trade and Endogenous Heterogeneity," CEPR DP 10938.

Bonfiglioli, Alessandra, Rosario Crinò and Gino Gancia (2015b). "Trade, Finance and Endogenous Heterogeneity" Working Paper.

Caggese, Andrea (2015). "Financing Constraints, Radical versus Incremental Innovation, and Aggregate Productivity," Working Paper.

Helpman, Elhanan, Oleg Itskhoki and Stephen J. Redding (2010). "Inequality and nemployment in a Global Economy." Econometrica, 78: 1239-1283.

Manova, Kalina (2013). "Credit Constraints, Heterogeneous Firms, and International Trade," Review of Economic Studies 80, 711-744.

Melitz, Marc J. (2003). "The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity." Econometrica, 71: 1695-1725.

Syverson, Chad (2004). "Market Structure and Productivity: A Concrete Example." Journal of Political Economy, 112: 1181-1222.

Footnote

1 A similar mechanism is at work in Caggese (2015) who finds that financial frictions lower entry and discourage radical, high-risk, innovation in a model of firm dynamics.

2,415 Reads