The Impact of Bank Credit on Labor Reallocation and Aggregate Industry Productivity

• Published date: 01 December 2018
• DOI: http://doi.org/10.1111/jofi.12726
• Author: JOHN (JIANQIU) BAI,DANIEL CARVALHO,GORDON M. PHILLIPS
• Date: 01 December 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 6 •DECEMBER 2018

The Impact of Bank Credit on Labor Reallocation

and Aggregate Industry Productivity

JOHN (JIANQIU) BAI, DANIEL CARVALHO, and GORDON M. PHILLIPS∗

ABSTRACT

We provide evidence that the deregulation of U.S. state banking markets leads to a

significant increase in the relative employment and capital growth of local firms with

higher productivity, and that this effect is concentrated among young firms. Using

financial data for a broad range of firms, our analysis suggests that this effect is driven

by a shift in the composition of local bank credit supply toward more productive firms.

Weestimate that this effect translates into economically important gains in aggregate

industry productivity and that changes in the allocation of labor play a central role

in driving these gains.

AGROWING BODY OF EVIDENCE SUGGESTS that improvements in financial markets

can significantly contribute to economic growth and that increases in aggregate

productivity play a central role in this process.1Despite the importance of this

idea, we still have limited direct evidence on the specific economic mechanisms

through which finance affects growth and productivity. In this paper, we study

the role of local U.S. banking markets in shifting the allocation of labor and

capital within local industries toward firms with higher productivity. Intu-

itively, such reallocation changes the composition of an industry and plays an

important role in increasing the industry’s aggregate productivity as it allows

a given level of aggregate resources to generate more output.2Banks can

∗John (Jianqiu) Bai is from D’Amore-McKim School of Business at Northeastern University.

Daniel Carvalho is from Kelley School of Business at Indiana University. Gordon M. Phillips is

from Tuck School of Business at Dartmouth College. We thank Manuel Adelino; Andrew Ellul;

Mark Garmaise; John Haltiwanger; Javier Miranda; Marco Navone; Vincenzo Quadrini; Geoffrey

Tate; Michael Roberts (the Editor); Jeremy Stein; Re´

ne Stulz; Greg Udell; two anonymous referees;

and seminar participants at the AFAmeetings 2016, CSEF-EIEF-SITE Conference on Finance and

Labor, FIRS conference, FMA meetings, Indiana University (Kelley), NBER Corporate Finance SI

2015, NBER CRIW SI 2015, Northeastern University, Ohio State (Fisher), UCLA, University

of Florida, University of Maryland (Smith), USC Marshall, University of Toronto, University of

Wisconsin-Madison, and Washington University at St. Louis (Olin) for helpful comments. Any

opinions and conclusions expressed herein are those of the authors and do not necessarily represent

the views of the U.S. Census Bureau. All results have been reviewed to ensure that no confidential

information is disclosed. All of the authors have read the Journal of Finance’s disclosure policy

and have no conflicts of interest to disclose.

1See Jayaratne and Strahan (1996), Beck, Levine, and Loayza (2000), and Guiso, Sapienza, and

Zingales (2004).

2Previous research emphasizes the role of this within-industry composition effect as a funda-

mental determinant of aggregate productivity (e.g., Olley and Pakes (1996) and Hsieh and Klenow

DOI: 10.1111/jofi.12726

2787

2788 The Journal of Finance R

potentially be important in shaping this resource reallocation because of their

role in determining which firms are financed. Therefore, improvements in bank-

ing markets can lead to a more positive link between firm productivity and bank

credit. We examine this idea and the importance of labor in this process. While

labor is a central factor used in production, limited attention has been paid to

the role of labor reallocation in explaining the effects of finance on aggregate

productivity.

We study the importance of bank lending for both labor and capital re-

allocation in the context of major U.S. state banking deregulations, which

reduced constraints on banks’ ability to expand across geographic markets

and represent historically important changes in U.S. banking markets. An

advantage of analyzing these reforms is that they capture focused changes in

the degree of bank competition in a local market. Broad patterns around these

reforms indicate that the previous reallocation effects appear to be important

in driving higher economic growth during these episodes (Jayaratne and Stra-

han (1996,1998)). As banking markets become more competitive, banks—new

or existing—might improve the screening and monitoring of borrowers. This

can happen, for example, because of an improved selection mechanism among

surviving banks or because a given bank faces stronger incentives for perfor-

mance. Improvements in the screening of borrowers should make it easier for

banks to detect higher productivity firms. Additionally, improvements in the

monitoring of firms should make it easier for borrowers to pledge their future

income. Firms with higher productivity can arguably benefit more from these

improvements because they have more income to pledge.

However, in theory, the sign and magnitude of this effect are unclear.

Increased bank entry may simply lead to a reduction in market power and

interest rates, increasing the availability of credit for all firms. Reductions in

bank market power can also harm relationship formation (Petersen and Rajan

(1995)) and lead banks to lend less to all small businesses. In addition, new

(larger) banks entering a market might not have a comparative advantage in

small business lending and hence may be worse in monitoring and screening

these firms.

Empirically analyzing the importance of the above reallocation effects has

been difficult due to data and identification challenges. Such analysis requires

microlevel data on small firms connecting their real decisions and sources of

external finance to their productivity and age. We analyze small firms by using

firm- and plant-level data from the U.S. Census Bureau. Two recent studies

by Kerr and Nanda (2009) and Krishnan, Nandi, and Puri (2015)alsouse

these data to analyze the effects of deregulation. However, these studies do not

examine the reallocation of production from less productive to more productive

firms, nor the impact on labor versus capital, and they do not study the bank

(2009)). Differences in firm productivity are typically the fundamental source of firm heterogeneity

emphasized in empirical studies and models of such composition effects (Bartelsman, Haltiwanger,

and Scarpeta (2013)).

Labor Reallocation and Aggregate Industry Productivity 2789

financing patterns of these small firms.3We also quantify the magnitudes of

the industry productivity gains implied by our results.4

We implement our analysis using microlevel data from the U.S. Census

Bureau on a broad sample of small U.S. manufacturing firms. These data

allow us to relate firms’ real decisions and financial policies to measures of

their multifactor productivity estimated using plant-level data. We argue that

the above reallocation effects should be most important for the youngest firms.

Because of older firms’ established lending relationships, new firms are more

likely to borrow from new banks entering a market. If new banks change the

composition of local credit supply,then this effect should concentrate among the

youngest firms. Differences in the screening and monitoring of banks should

also matter more for young firms, which have a limited track record, and these

firms should depend more on external funds to finance their higher growth.5

We start by analyzing how the relative growth of higher productivity

firms within a local industry changes after deregulation in their state. We

estimate this effect separately for different firm age groups and examine its

differential importance for young existing firms. Our main results employ a

triple-difference framework, where we compare changes in relative growth

across states experiencing deregulation at different points in time, and then

contrast this effect for different age groups. We find that deregulation is

associated with an economically large increase in the relative growth of more

productive young firms. Moreover, this relative growth increase is important

for both labor and capital, and it is not present for older firms. We further find

that this increase in growth takes place right after deregulation and is not

associated with preexisting growth trends.

We next investigate the role of changes in the composition of credit supply

in driving the above growth effect. First, we directly examine the importance

of such changes in the composition of credit supply. As before, we separately

examine different age groups and focus on differential patterns for young

firms. To analyze the role of bank credit supply, as opposed to credit demand,

we analyze how firms borrow across different sources of debt financing. Our

data allow us to distinguish between debt from commercial banks (bank debt)

and debt from other sources (nonbank debt), and deregulation only covered

commercial banks. Consistent with previous research (Petersen and Rajan

(1994)), we find that young and old small firms rely significantly on both bank

and nonbank debt. We argue that changes in credit demand should affect

3In their discussion of the mechanism through which deregulation matters, Jayaratne and

Strahan (1996, p. 664–65) explain that “To answer this question satisfactorily, we would like data

on bank borrowers such as the productivity and longevity . . . especially among bank-dependent

firms such as small business.”

4When examining implications for industry productivity, an important consideration is that

firms affected by deregulation might represent a small portion of aggregate industry output. We

explicitly incorporate this issue when quantifying the magnitudes of our results.

5As discussed in Section I.B, previous evidence based on aggregate data from banks’ balance

sheets shows that these reforms were associated with a significant increase in the market share of

better performing banks in deregulated states.