Asymmetric effects of financial development on export price and quality across countries

• Published date: 01 May 2019
• Author: ByeongHwa Choi,Volodymyr Lugovskyy
• Date: 01 May 2019
• DOI: http://doi.org/10.1111/roie.12390

594

|

wileyonlinelibrary.com/journal/roie Rev Int Econ. 2019;27:594–642.

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

A large body of literature has emphasized the importance of financial development (FD) for economic

growth and exporting.1 Nonetheless, recent empirical findings have shown that FD does not affect the

growth of low‐income countries.2 These findings raise questions about how FD affects exporting, and

whether FD benefits all exporting countries equally.

These aforementioned questions motivate us to explore the effects of FD on two essential compo-

nents of exporting: export prices and export quality.3 We find that these effects are highly asymmetric

across countries, and we provide an explanation for these asymmetries. Additionally, our findings

suggest that improving the quality of financial institutions alone is unlikely to boost quality or lower

export prices in low‐income countries.

Theoretically, we provide a unifying framework to explain why and how the effects of FD on

export quality and prices differ across countries. A key component of our model is based on the

Received: 10 May 2018

|

Revised: 18 December 2018

|

Accepted: 19 December 2018

DOI: 10.1111/roie.12390

ORIGINAL ARTICLE

Asymmetric effects of financial development on

export price and quality across countries

ByeongHwaChoi1

|

VolodymyrLugovskyy2

1Department of Economics,National Taiwan

University, Taipei, Taiwan (R.O.C.)

2Department of Economics,Indiana

University, Bloomington, Indiana

Correspondence

Volodymyr Lugovskyy, Department

of Economics, Indiana University, 105

Wylie Hall, 100 S. Woodlawn Avenue,

Bloomington, IN 47405.

Email: vlugovsk@indiana.edu

Funding information

The Ministry of Science and Technology of

Taiwan, Grant/Award Number: MOST

106-2410-H-002-001

Abstract

We derive two novel predictions: financial development

has a more pronounced effect on quality in countries with

greater labor productivity, and its effect on export prices is

U‐shaped in labor productivity. We confirm our predic-

tions empirically and show that the negative effect of fi-

nancial development on export prices is greatest in

middle‐productivity countries, while its positive effect on

quality is strongest in the most productive countries. Our

findings contribute to the literature on the poverty trap: we

argue that improving the quality of financial institutions

alone is unlikely to boost quality or lower prices of the

poorest countries.

JEL CLASSIFICATION

F10, F14, G20, J24, O14

|

595

CHOI and LUGOVSKYY

complementarities between different inputs—capital, labor productivity, and entrepreneurial talent—

in the production function of quality. Building on Verhoogen (2008), we employ a discrete choice

model, where the production function of quality contains the elements of O‐ring technology (Kremer,

1993). We extend it in two important dimensions: first, we introduce the liquidity constraints of ex-

porting firms in the model and assume that the cost of borrowing decreases in FD.4 Second, in line

with the labor search literature,5 we assume a positive correlation between labor productivity and

entrepreneurial talent across countries.

The model provides three testable predictions: (i) controlling for quality, better FD decreases ex-

port prices; (ii) FD has a positive effect on export quality, and more so in countries with higher labor

productivity, and (iii) for sufficiently large parameter regulating the share of capital in the production

function, the effect of FD on export prices is U‐shaped in labor productivity.

The first effect is the most straightforward: given the same set of inputs, a lower cost of borrowing

reduces the total cost and sequentially the price. The second effect is due to the super‐modularity

of the production function of quality in inputs. As a result, any reduction in the variable produc-

tion cost—including lower borrowing costs owing to better developed financial institutions—induces

firms to upgrade quality.6 Finally, the U‐shaped FD–price pattern emerges from the O‐ring technology

feature of the production function. Specifically, the production of quality requires a firm to pay a stan-

dard wage in a given economy plus a wage premium required to hire better workers to produce higher

quality. While the standard wage portion of the cost is independent of quality, the wage premium and

the amount of capital employed increase in quality. Thus, FD has a negative effect on the quality‐inde-

pendent portion and a positive effect on the quality‐dependent portion of the cost. The magnitude of

the latter effect is determined by the level of labor productivity, which produces a U‐shaped pattern.

Empirically, we test our predictions using product‐level U.S. Census Imports Data over the span of

17 years (1991–2007). We focus on differentiated goods, which are more likely to be subject to quality

differentiation. The FD is proxied by the private credit‐to‐GDP ratio, and its effect on export prices

tends to be negative but non‐uniform across exporting countries. Specifically, the effect is strongest

for countries where real GDP per worker is around U.S.$22,000: for these countries, the price de-

creases by 1.87% for every 10 percentage point increase in the private credit‐to‐GDP ratio. The effect

is much weaker for all other countries, whether wealthier or poorer. This confirms the predicted U‐

shaped relationship between the effect of FD on export prices and labor productivity. As predicted,

controlling for quality, FD has a negative effect on export prices.

To evaluate the effect of FD on export quality, we use average country‐product quality indices of

exporting goods estimated by Feenstra and Romalis (2014). We find only a partial support for our

second prediction: the effect of FD on quality increases in labor productivity, but it is positive only

after a threshold of U.S.$17,700 for per‐worker productivity.

We perform several robustness checks: (i) we use less parametric specifications with country group

dummies and their interaction terms with FD; (ii) we use Khandelwal’s (2010) quality measure in-

stead of that of Feenstra and Romalis (2014); (iii) we instrument FD with the La Porta, Lopez‐de

Silanes, Shleifer, and Vishny (1998) measures of legal origin; (iv) we use the total factor productivity

(TFP) measure obtained from the Penn World Table instead of real GDP per worker as the labor pro-

ductivity measure; and (v) we split the sample into subsamples based on the value of the elasticity

of substitution and based on the length of Khandelwal’s (2010) quality ladder. We confirm that our

results are robust to these checks.

To summarize, FD affects export determinants differently across countries. It: (i) increases export

quality and only marginally reduces export prices among high‐income countries7; (ii) significantly

reduces prices and has no effect on quality among middle‐income countries; and (iii) marginally re-

duces prices among low‐income countries. These results aid in interpreting our finding that FD does

596

|

CHOI and LUGOVSKYY

not increase exports of low‐income countries, and also help us to distinguish between the quality and

price effects of FD on exports for countries with different income levels.

Our results allow us to conclude that the complementarity in quality production works against

the least‐developed countries: they are unlikely to improve quality and increase exports by simply

improving their financial institutions. This finding contributes to the literature on the poverty trap8

by suggesting that improved access to credit markets alone might not be sufficient for better export

performance.

We further contribute to the literature on export quality and prices by highlighting the interac-

tion between income per worker and FD as export quality and price determinants. Schott (2004) and

Hummels and Klenow (2005), among others, have shown that wealthier countries export higher‐

priced goods.9 They attribute this pattern to quality differentiation. The positive correlation between

product quality and FD is also well‐documented (e.g., Ciani & Bartoli, 2015; Fan, Lai, & Li, 2015).

Additionally, Secchi, Tamagni, and Tomasi (2016) show that firms facing tight credit constraints

charge higher export prices in the presence of both quality and cost adjustment effects. Our novel

findings are that the effects of FD on export quality and prices are nonlinear and vary greatly across

countries. We also predict theoretically and confirm empirically that in some cases, FD decreases ex-

port prices while increasing export quality. These findings contribute to the ongoing debate regarding

the conditions under which price is positively correlated with quality.10

Finally, we extend the growing empirical literature documenting the systematic variation in the

impact of FD on export performance. For instance, Manova (2013) argues that FD increases both the

number of exported products and their export values, particularly in financially vulnerable sectors.

Similarly, Crinò and Ogliari (2017) show that FD raises the average quality and the relative exports

disproportionately more in financially vulnerable sectors. Berthou (2010) finds a hump‐shaped rela-

tionship between the marginal effect of FD on exports and the initial development of financial insti-

tutions in industries where firms have a high demand for external finance. Our paper focuses on the

asymmetry of the effect of FD on exports across countries, rather than across sectors, while all time‐

varying product‐specific factors are captured by a set of product–year fixed effects.11 Importantly,

we show that the effects of FD on export prices and quality differ across countries. Thus, the optimal

export‐promoting policies are likely to be country‐specific rather than universal.

The remainder of the paper is organized as follows: Section 2 presents the model; Section 3 outlines

the data and empirical strategy; Section 4 discusses the results and robustness checks; and Section 5

concludes.

2

|

MODEL

We build on the endogenous quality choice model of Verhoogen (2008) and adapt it to a multi‐country

setting and liquidity‐constrained firms. Our goal is to explore how FD affects the export quality and

price across countries‐exporters with different levels of labor productivity. We establish our model

from the perspective of a firm in Country d = 1, 2, 3, … ,D, and examine its export quality and price

choices to an importing country, Importer.

We deviate from Verhoogen’s (2008) original model in two ways. First, we simplify the model

by assuming away firm heterogeneity within a country and focusing only on the asymmetry of firms

across countries. Second, motivated by labor search models, we consider that wages and entrepreneur-

ial talent are not exogenously given but are increasing functions of the average labor productivity in

a country.