Endogenous Growth and Real Effects of Monetary Policy: R&D and Physical Capital Complementarities

• Published date: 01 August 2020
• Author: PEDRO MAZEDA GIL,GUSTAVO IGLÉSIAS
• Date: 01 August 2020
• DOI: http://doi.org/10.1111/jmcb.12632

DOI: 10.1111/jmcb.12632

PEDRO MAZEDA GIL

GUSTAVO IGL ´

ESIAS

Endogenous Growth and Real Effects of Monetary

Policy: R&D and Physical Capital

Complementarities

We study the real long-run effects of the structural stance of monetary pol-

icy and of inf‌lation, in the context of a monetary growth model where R&D

is complemented with physical capital accumulation. We look into the ef-

fects on a set of real macroeconomic variables that have been of interest

to policymakers—the economic growth rate, real interest rate, physical in-

vestment rate, capital-to-labor ratio, R&D intensity, and velocity of money.

These variables have been previously analyzed from the perspective of

different, separated, strands of the theoretical and empirical literature. Ad-

ditionally, we analyze the long-run relationship between inf‌lation and both

the effectiveness of real industrial-policy shocks and the market structure,

assessed namely by average f‌irm size. We present novelcross-country evi-

dence on the empirical relationship between the latter and long-run inf‌lation.

JEL codes:E41, O31, O41

Keywords:endogenous growth, R&D, physical capital, inf‌lation, money,

cash-in-advance, f‌irm size.

THE REAL EFFECTS OF INFLATION and of the structural stance of

monetary policy over the long run are acknowledged by policymakers,1as well as

We thank Pedro Bento for supplying the country-leveldata for total entry costs as a fraction of output

per worker. We thank the editor, Sanjay Chugh, and two anonymous referees for their most valuable

comments and suggestions. This research was f‌inanced by the European Regional Development Fund

through COMPETE 2020 Programa Operacional Competitividade e Internacionalizac¸˜

ao (POCI) and by

Portuguese public funds through FCT (Fundac¸˜

aoparaaCi

ˆ

encia e a Tecnologia) in the framework of the

project POCI-01-0145-FEDER-006890.

PEDRO MAZEDAGIL is an Assistant Professor at University of Porto, Faculty of Economics, and CEF.UP

(E-mail: pgil@fep.up.pt).GUSTAVO IGL´

ESIAS is at Banco de Portugal (E-mail: giglesias@bportugal.pt).

Received June 11, 2018; and accepted in revised form February 5, 2019.

1.For instance, the European Central Bank states that, over time, “price stability contributes to

achieving high levels of economic activity and employment” (text on the “Benef‌its of price stability,” at

https://www.ecb.europa.eu/mopo/intro/benef‌its/html/index.en.html).

Journal of Money, Credit and Banking, Vol. 52, No. 5 (August 2020)

C

2019 The Ohio State University

1148 :MONEY,CREDIT AND BANKING

conf‌irmed by the empirical evidence gathered by academic economists. This paper

looks into those real long-run effects in the framework of a monetary growth model

of R&D and physical capital accumulation. We analyze the long-run relationship

between the structural level of the monetary policy instrument, the inf‌lation rate,

and an array of real macroeconomic variables that have been of interest to monetary

policymakers (e.g., Bernanke and Mishkin 1997), and compare with the empirical

evidence. These variables are: the economic growth rate, real interest rate, physical

investment rate, capital-to-labor ratio, R&D intensity, and velocity of money. More-

over, we analyze the relationship between inf‌lation and both the effectiveness of

real industrial-policy shocks and the market structure, assessed namely by average

f‌irm size.

The setup of our model explicitly takes into account the close interrelation between

physical and technological inputs empirically observed along growth processes (e.g.,

Dowrick and Rogers 2002, Tamura et al. 2019) by allowing physical capital accu-

mulation and R&D to complement each other as engines of long-run growth (e.g.,

Romer 1990, Howitt and Aghion 1998, Howitt 2000, Sedgley and Elmslie 2013, Gil,

Almeida, and Castro 2017). This is a well-known analytical setup in the R&D-driven

endogenous growth literature, but which has not yet been used in the context of the

monetary growth models. Following a recent literature on R&D-driven growth (e.g.,

Chu and Cozzi 2014), we introduce money demand in the model by considering

cash-in-advance (CIA) constraints on R&D investment and also on manufacturing of

intermediate goods. However,since in our model the latter uses physical capital as an

input, then the respective CIA constraint also affects the mechanism of physical cap-

ital accumulation.2Given the considered complementarity between physical capital

accumulation and R&D, this will imply a novel kind of interrelation between the two

referred to CIA constraints, in comparison with the previous literature. This setup

also allows us to discuss the impact of a lower labor income share (as pointed out by

recent empirical literature; for example, Elsby, Hobijn, and Sahin 2013, Karabarbou-

nis and Neiman 2014) on the magnitude and the nonlinearities of the real long-run

effects of monetary policy and inf‌lation, as well as of real industrial-policy shocks.

The empirical evidence suggests that there tends to be, in the long run: (i) a

negative relationship between inf‌lation and both the R&D intensity and economic

growth (e.g., Evers, Niemann, and Schiffbauer 2007, Chu and Lai 2013, Chu et al.

2015); (ii) a negative relationship between inf‌lation and both the real interest rate and

physical investment rate (e.g., Gillman and Kejak 2011 and references therein); (iii)

a positive relationship between inf‌lation and the velocity of money (e.g., Palivos and

2.As Chu and Cozzi point out, empirical evidence clearly suggests that R&D investment is severely

affected by liquidity requirements, even more so than physical investment (see, e.g., Brown,Martinsson,

and Petersen 2012, Falato and Sim 2014, Brown and Petersen 2015), thus backing up a growth model that

features CIA requirements in both activities (R&D and physical investment). Other endogenous growth

models with CIA constraints on R&D activities are Chu et al. (2015), Huang, Yang, and Cheng (2017),

Chu, Ning, and Zhu (2019), and Chu et al. (2017). An alternative strand of the literature has focused on

the effects of money and inf‌lation in the context of (physical and/or human) capital-drivengrowth models

with CIA constraints (e.g., Itaya and Mino 2007 and Gillman and Kejak 2011; see Gillman and Kejak

2005b for a survey of earlier contributions on this topic).

PEDRO MAZEDAGIL AND GUSTAVOIGL ´

ESIAS :1149

Wang 1995, Dotsey and Ireland 1996, Rodr´

ıguez Mendiz´

abal 2006). Based on own

empirical evidence for OECD countries, we add: (iv) a positive relationship between

inf‌lation and both the capital-to-labor ratio and average f‌irm size, measured as the

stock of capital per f‌irm, in manufacturing.

Our model obtains, in the long-run equilibrium, a negative relationship between

inf‌lation and the economic growth rate, real interest rate, R&D intensity,and physical

investment rate, and a positive relationship between inf‌lation and the velocity of

money. Thus, in all cases, the sign of the theoretical relationships is in line with the

cited empirical evidence.

The focus of our paper on the joint effects on the economic growth rate, real

interest rate, and R&D intensity relates in particular to Chu and Cozzi (2014). This

paper develops a monetary model of R&D-driven growth with a setup that is close to

ours but with no physical capital accumulation. The long-run relationship between

inf‌lation and those real macroeconomic variables is negative in case the degree of

the CIA constraint on R&D exceeds that on manufacturing—as seems to be the case

empirically accordingly to the literature on f‌irm f‌inancing (e.g., Brown and Petersen

2015). In our paper, we obtain a negative long-run relationship between inf‌lation and

those real macroeconomic variables that is independent of the relative degree of the

CIA constraint on R&D vis-`

a-vis manufacturing. However, and differently from the

existing literature, the real long-run effects of inf‌lation in our model are sensitiveto the

size of the labor share. Interestingly, whether those effectsare dampened or enhanced

by a lower labor share does depend on the relative degree (larger or smaller) of the

CIA constraint on R&D vis-`

a-vis manufacturing. Moreover, we expand the array of

real variables that are analyzed in a common general equilibrium setup (namely the

physical investment rate, velocity of money, and market structure).

By looking into the joint effects on the economic growth rate, real interest rate,

physical capital-to-labor ratio, and physical investment rate, our contribution also re-

lates closely to Gillman and Kejak (2011). These authors study an endogenous growth

model of human and physical capital accumulation, extended with a banking sector.

They address the empirical evidence of a negative relationship between inf‌lation and

both the real interest rate and the investment rate, which, as pointed out by Gillman

and Kejak, may be viewed as, respectively, a positive and a negative (long-run) Tobin

effect. By exploring the interplay between R&D and physical capital, we uncover

a different mechanism to address the seemingly opposite Tobin effects, while our

results are not conditional on a specif‌ic interval of values for the inf‌lation rate, in con-

trast to Gillman and Kejak (2011). Furthermore, we derive our results in a framework

that encompasses a more complete set of relevant real variables, as described earlier.

Our paper also discusses the nonlinearity of the relationship between inf‌lation

and economic growth, following Gillman and Kejak (2005b). These authors look

into the marginal nonlinearity of the (negative) relationship between inf‌lation and

economic growth from the point of view of alternative theoretical growth models in

the literature, featuring physical capital accumulation, human capital accumulation,

or both, possibly extended with a credit sector. As already emphasized, we carry

out this analysis in the context of a model that features both R&D and physical