Can Innovation Help U.S. Manufacturing Firms Escape Import Competition from China?

• DOI: http://doi.org/10.1111/jofi.12691
• Date: 01 October 2018
• Author: JOHAN HOMBERT,ADRIEN MATRAY
• Published date: 01 October 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 5 •OCTOBER 2018

Can Innovation Help U.S. Manufacturing Firms

Escape Import Competition from China?

JOHAN HOMBERT and ADRIEN MATRAY∗

ABSTRACT

We study whether R&D-intensive firms are more resilient to trade shocks. Wecorrect

for the endogeneity of R&D using tax-induced changes to R&D costs. While rising

imports from China lead to slower sales growth and lower profitability, these effects

are significantly smaller for firms with a larger stock of R&D (about half when moving

from the bottom quartile to the top quartile of R&D). We provide evidence that this

effect is explained by R&D allowing firms to increase product differentiation. As

a result, while firms in import-competing industries cut capital expenditures and

employment, R&D-intensive firms downsize considerably less.

THE RISE OF CHINA,TRIGGERED BY ITS transition to a market-oriented economy

and rapid integration into world trade, has been identified as a major source

of disruption for high-income economies, igniting the long-standing debate re-

garding the effect of trade with low-wage countries on firms and workers in the

United States and Europe and regarding which firms are better able to absorb

these stocks. In this context, innovation is often viewed as an effective shield

against low-cost foreign competition by allowing firms to climb the quality lad-

der and differentiate their products from low-wage countries’ exports. Because

wage differences are so large, so the argument goes, competing on costs is bound

to fail. Only firms that have invested in R&D and upgraded product quality

are able to compete successfully against low-cost imports (e.g., Leamer (2007)).

∗Johan Hombert is with HEC Paris and CEPR. Adrien Matray is with Princeton University.We

thank Nick Bloom; Gerard Hoberg; Rich Mathews; Stephen Redding; Amit Seru; Stefan Zeume;

Michael Roberts; and two anonymous referees as well as seminar participants at INSEAD, Copen-

hagen Business School, Stanford University, CSEF–University of Naples, Imperial College, Mon-

treal University, Erasmus University, Ghent University, the Harvard Business School; and con-

ference participants at the Labex Ecodec Workshop at HEC Paris, the London Business School

2015 Summer Finance Symposium, the 2015 Western Finance Association meeting, the 2015

Workshop on Entrepreneurial Finance and Innovation Around the World, the 2015 Workshop on

the Economics of Corporate Ownership, the European Summer Symposium in Financial Markets

(Gerzensee) 2015, the 2015 Econometric Society World Meeting, and the 2016 NBER Productivity,

Innovation, and Entrepreneurship. Zhen Ye provided excellent research assistance. Hombert ac-

knowledges financial support from the Investissements d’AvenirLabex (ANR-11-IDEX-0003/Labex

Ecodec/ANR-11-LABX-0047). Matray acknowledges financial support from the Julis-Rabinowitz

Center for Public Policy and Finance. All of the authors have read the Journal of Finance’s disclo-

sure policy and have no conflict of interest to disclose.

DOI: 10.1111/jofi.12691

2003

2004 The Journal of Finance R

This view has had a large influence on public policies. In particular, it has

provided further justification for R&D subsidies.1

There is surprisingly little evidence, however, about whether firms that have

invested in R&D are indeed shielded from trade shocks. In this paper, we

provide direct evidence of this relation. In particular, we show that firm perfor-

mance (sales growth and profitability) is less adversely affected by an increase

in import competition when the firm has ex ante invested more in R&D. The

magnitude is economically large. Moreover, we also show that the effect oper-

ates through product differentiation.

The innovation literature analyzes a related but different question, namely,

how firms endogenously adjust their R&D investment ex post, that is, after

an increase in import competition. This approach can be interpreted through

the lens of a revealed preference argument, whereby firms’ innovation choice

after trade shocks “reveals” their expectations as to whether R&D is an effec-

tive shield against import competition. Recent evidence using this approach is

mixed. Bloom, Draca, and Van Reenen (2016) find a positive relation between

import competition and innovation in Europe, whereas Autor et al. (2016) find

a negative relation in the United States.

The revealed preference argument has two important limitations, however.

First, the sensitivity of R&D investment to import competition may be infor-

mative about the sign of the relationship between returns to R&D and import

competition but not about its magnitude, as the elasticity also depends on the

R&D cost structure. Second, even the estimated sign of this relationship may

be wrong even if trade shocks are instrumented because the decision to inno-

vate in reaction to a shock depends on factors over and above the mere returns

to R&D, such as credit constraints, managers’ expectations regarding both the

effect of the shock and the gain from innovating, and agency issues that may in-

duce manager short-termism.2Because import penetration shocks affect cash

flows, studying firms’ endogenous R&D response to these shocks may yield an

incorrect answer to the question of whether R&D mitigates the negative effect

of trade shocks. For instance, it may be the case that R&D is an effective shield

against import competition but firms suboptimally cut R&D expenditures after

trade shocks due to financing or agency frictions. This would explain why the

evidence of the effect of import competition on ex post R&D choices is mixed.

In this paper, we adopt a direct approach to estimating the effect of R&D

on firms’ resilience to import competition. We test whether firm performance

1In 2013, 27 of the 34 OECD countries and a number of non-OECD economies provided fiscal

incentives for R&D (OECD (2014)). The European Union’s Lisbon Strategy envisioned making

Europe “the most competitive and dynamic knowledge-based economy in the world (...) to increase

its productivity and competitiveness in the face of ever fiercer global competition” (European

Commission (2010)).

2These frictions are particularly relevant for R&D expenditures because they are typically cut

in priority upon negative cash flow shocks. For instance, Aghion et al. (2012) show that credit

constraints force firms to cut R&D more than other expenditures during downturns, lowering

productivity growth. Bhojraj et al. (2009) show that R&D is a key strategic variable used by

managers to manipulate short-term earnings to the detriment of long-term profitability.

Can U.S. Manufacturing Firms Escape Import Competition from China 2005

as measured by sales growth or profitability is less adversely affected by an

increase in import competition when the firm has invested more in R&D before

the increase in import penetration. We conduct this test in the context of China’s

export boom and its effect on U.S. manufacturing firms. This allows us to

estimate the effect of R&D as an ex ante moderating variable on the disruptive

effects of trade with China.

To inform the identification strategy, we first develop a model of the interplay

between R&D and product market competition. The model shows that when

we make rigorous the two limitations of the revealed preference argument

discussed above and shows that regressing R&D on import penetration does

not identify the effect of R&D on firm resilience to import competition (even if

import penetration is correctly instrumented). The model further shows that

regressing firm performance on import penetration interacted with R&D yields

an unbiased estimate of this effect only if both import penetration and R&D

are instrumented.

First, China’s import penetration in the United States may be endogenous

to the performance of U.S. firms as lower productivity in the United States

may lead to higher imports to the United States. To isolate the component of

China’s rising exports that stems from internal supply shocks in China, we

borrow a classic identification strategy from international trade economics.

Specifically, we instrument China’s import penetration in the United States

at the industry level using China’s import penetration in other high-income

countries (e.g., Autor, Dorn, and Hanson (2013), Hummels et al. (2014)). There

has been tremendous import growth from China in some industries (e.g., tex-

tiles, electronics, furniture, and industrial equipment) but not in others (e.g.,

tobacco, printing, food, and petroleum). This cross-industry heterogeneity is

similar in the United States and in the other high-income economies, which

suggests that it is driven by supply shocks in China.

Second, the accumulation of R&D capital is potentially endogenous to firms’

productivity, management quality, and product demand. We thus instrument

for R&D at the firm level using tax-induced changes to the user cost of R&D

capital. After the introduction of the U.S. federal R&D tax credit in 1981,

U.S. states started to introduce R&D tax credits as well. In 2006, 32 states

offered tax credits, in some cases considerably more generous than the federal

credit (Wilson (2009)). The staggered implementation of these R&D policies

generates variations across states and over time of the price of R&D, which in

turn generates exogenous variation in firms’ R&D stock.

With these two instruments in hand, we estimate how firms are affected by

(exogenous) import competition, conditional on their (exogenous) R&D stock

before the competition shock. Our preferred specification includes firm fixed

effects to absorb time-invariant firm characteristics and industry-by-year fixed

effects to account for industry-specific productivity shocks and changes in con-

sumer demand.

We show that China’s import penetration has a sizable adverse effect on

the unconditional (i.e., independent from the R&D level) performance of

U.S. manufacturing firms. On average across U.S. manufacturing firms, a