Manufacturing Export Performance, Public Capital and Proximity to the Technological Frontier of Countries*

Abstract

The paper seeks to understand the pattern of manufacturing exports in 35 developed and developing countries and 99 industrial sectors over the period 1999-2013 using a methodological approach that relies on propositions of the neoclassical trade theory. Next to the impact of traditional factors such as the endowment of public capital, we look at the importance of accounting for public capital heterogeneity, by distinguishing countries based on their proximity to the technology frontier. We consider public capital as a non-rival factor, which has a property of “creation of atmosphere” in a model that treats factor intensity as a mechanism of industrial development. We use the system GMM technique to estimate the elasticities of public capital. Our results show that public capital plays a differentiated role in the capacity to export industrial goods depending on whether countries are far or close to the technological frontier.

Keywords:

• Manufacturing exports
• Rybczynski’s theorem
• public capital
• proximity to the technological frontier
• GMM

JEL Classification:

• F11
• H41
• O47
• C33

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The definition of public capital used in our data allows it to be equated with public infrastructure and thus used as a proxy. According to this definition, public capital includes public infrastructure such as economic infrastructure (highways, airports, roads, railroads, water and sewage systems, electricity and gas utilities, pipelines, and telecommunications), social infrastructure (public schools, hospitals, and prisons), and non-produced assets (sub-soil assets). See IMF database.

2 According to our definition of close to or far from the technology frontier. See Section 3 for more details.

3 According to the concept of proximity to the technology frontier, there are, on the one hand, the countries close to the frontier and, on the other, the countries far from this frontier. But in countries far from the border, in addition to African countries (Benin, Central African Republic, Ivory Coast, Gabon, Morocco, Mauritania, Niger, Senegal, Sierra Leone, Tunisia and Togo), we have emerging countries (China, Greece, India, Indonesia, South Korea, Malaysia, Philippines and Thailand), and even newly industrialized countries such as South Korea. Thus, we separate the sample of countries and prove that public capital accumulation is different; then, we test and show that public capital plays a different role for countries far from the technology frontier.

4 The notion of less advanced countries is what we call countries far from the technology frontier. This notion includes both African and emerging countries. The concept of emerging countries refers to all the countries that have, in recent decades, emerged from underdevelopment, and which are of growing importance in the world economy. By convention, these countries are said to have an average per capita income of between 10% and 75% of that of developed countries, which leaves a rather disparate group. See Table A7 for a list of the countries far from the border.

5 For a better exploitation of the space, we aggregate the 99 sectors into 16 sectors according to the international classification HS89/1992.

6 See Romalis (2004) for more details on the use of factor share, defined as the ratio of the use of a given factor to the use of all existing factors, as a proxy for factor intensity.

7 See Ledezma et al. (2016) for more details.

8 see Feenstra et al. (2015) for more details on the construction of the TFP indicator

9 Author’ calculations based on the IMF data

10 See previous footnote.

11 Insofar as the findings show some disparities in the sense of what we call proximity to the technology frontier since there are African countries and emerging countries, this distinction is necessary and indispensable.

12 China, Greece, India, Indonesia, South Korea, Malaysia, Philippines and Thailand.

13 Benin, Central African Republic, Ivory Coast, Morocco, Mauritania, Niger, Senegal, Sierra Leone and Togo.

14 See Figure A4 in the appendix for more details on FDI flows in the different country categories.

15 This approach permits us to transform the instruments to make them orthogonal to fixed effects

16 This dynamic representation implies three non-linear restrictions on the unrestricted parameters that generally hold in the baseline regressions. To save espace the lagged terms are compressed in the presentation of the results. This form incorporates the traditional dynamic panel endogeneity bias that we address with system GGM internal instruments. We assume that all endowment factors (RHCI, RCI, HCE, KPRL, KPUL, inward FDI) are potentially contemporaneously correlated with the country specific effects $({\eta }_i),$ sectors specific effects $({\eta }_j)$ as well as idiosyncratic shocks $({\upsilon }_{\mathit{ijt}}).$

17 When transforming the regressors into first difference, potential biases caused by omitted variables and the fixed country-specific effect (which may be correlated with the explanatory variables) are removed, as these do not vary with time.

18 Countries close to and far from the technology frontier are defined according to whether the countries are respectively above or below the 50th percentile of our measure of proximity to the technology frontier (total factor productivity measured relative to that of the US). See Section 1 for more details.

19 In this table, we have decided to not include the results for all sample and the two subperiods.

20 We present here the results only for the period 1999-2009.

21 Colums 1, 3 and 4 of Table 2.