Openness, lobbying, and provision of infrastructure.

I. Introduction

It is widely acknowledged that infrastructure is critical for a nation's growth and economic development. The Worm Development Report (The World Bank 1994) points out that "the adequacy of infrastructure helps determine one country's success and another's failure--in diversifying production, expanding trade, coping with population growth, reducing poverty, or improving environmental conditions." Many studies have found a strong correlation between infrastructure and productivity. For example, Easterly and Rebelo (1993) and Hulten (1996) find that the stock of infrastructure is strongly related to economic growth in a cross-section of countries. Munnell (1992) concludes from a survey that public infrastructure has a significant, positive effect on output and growth. The significant role of infrastructure in economic development suggests that an inquiry into the determinants of infrastructure provision may be important for development policy.

In developed and especially in developing countries, the public sector provides the bulk of infrastructure investment. In developing countries, the public sector provides about 80% of total infrastructure investment, with the private sector contributing 7% and bilateral and multilateral aid contributing 12% (The World Bank 1994). Government not only finances most infrastructure but, in developing countries, provides a significant portion of all public investment expenditures--rarely less than 30% and sometimes as high as 70% (The World Bank 1994). Therefore, any analysis of the determination of infrastructure provision needs to consider the incentives facing the government in providing this public good.

There is anecdotal evidence to suggest that infrastructure provision is significantly higher in open economies than in closed ones. Mody (1997) suggests that infrastructure investment as a share of GDP in six major East Asian economies (1) ranges from 6 to 8% and averages about 4% for the typical developing country. These six economies also have relatively high ratios of trade (exports plus imports) to GDP. Mody argues that an important reason for this commitment to infrastructure provision is "to maintain competitiveness in export markets" (p. xii).

Even within a country there may be significant differences in infrastructure provision across sectors. Sectors exposed to international trade may exhibit higher infrastructure investment than sectors that are closed to the international economy. For example, in India, government expenditure in the telecommunications sector has grown at a significantly higher rate than in other infrastructure sectors such as power, steel, and roads. Research by Pingle (1999) suggests that this is due to the export-oriented software industry, which relies heavily on telecommunications infrastructure. (2) In recent years, government expenditure in the telecom sector actually exceeded planned targets, while in other sectors expenditures trailed substantially behind target levels (Ahluwalia 1998).

The relationship between infrastructure investment and openness to trade is evident in cross-country data, as shown in Figure 1. Here, infrastructure investment is measured by the share of government expenditure, in GDP, on transport and communication. This data is obtained from IMF (1986) government finance statistics and includes construction of roads; water, air, and rail transport; postal, telephone, telegraph, cable, and wireless communication systems; and communication satellites; but it excludes radio and television broadcasting systems. The openness variable on the x axis is an exogenous measure that is a function of geographic variables alone, originally proposed by Frankel and Romer (1999). It depends on exogenous country characteristics such as distance from other countries, country size (measured by population and area), whether or not it is landlocked, and borders shared with other countries. Causality should, therefore, run from openness to infrastructure, rather than in the reverse direction. The figure shows a clear positive relationship between these two variables. The correlation coefficient is 0.41 and is significant at the 5% level.

An important incentive to provide infrastructure, as Mody suggests, may stem from a desire to maintain the country's competitiveness in export markets. In particular, infrastructure investment may reduce costs for domestic firms that compete with foreign firms in export markets. In this paper, we ask whether access to trade leads to higher infrastructure investment and whether such investment is welfare improving, under social planning as well as lobbying activity.

We develop a model of trade in which provision of infrastructure makes the domestic firm more productive and steals market share from a foreign firm in both the domestic and the export markets. (3) Both trading partners invest in infrastructure, but this market-stealing effect provides an incentive to invest more in infrastructure in open economies than in closed ones. We examine the equilibrium stock of this capital when the government invests to maximize national welfare and when its investment decision is influenced by producer lobbies. We consider two lobbying scenarios--when taxes are imposed only on the high-income lobbying group and when they are imposed on the entire population.

[FIGURE 1 OMITTED]

We show that open economies invest more in infrastructure in all settings. However, infrastructure levels under the open economy compared to those under the closed economy can be welfare improving only when income and lobbying power are highly concentrated and the tax base is limited to the high-income group. When the tax base is wide, an open economy always provides too much infrastructure relative to the optimal amount. This result is consistent with empirical findings, such as those by Young (1995), who observes that there is excessive investment of capital in Singapore, a highly open economy. Our results are also consistent with Mohtadi and Roe (1998), who show that lobbying can be welfare improving, contrary to what is generally assumed. They use an alternative framework in which lobbying can lead to overprovision or underprovision of the public good, depending on whether spillovers associated with lobbying activity are small or large, respectively.

Several studies have examined the interaction between infrastructure and trade. Bougheas, Demetriades, and Morgenroth (1999) suggest that infrastructure reduces transport costs and therefore increases the volume of trade. They test this theory with data from the European Union countries and find strong empirical support for this relationship. Bougheas, Demetriades, and Morgenroth (2003) show that strategic behavior by countries in improving transport infrastructure may lead to overinvestment. The focus of both of these papers is on transport-related infrastructure. Investment in infrastructure reduces the costs of trade and therefore benefits both foreign and domestic producers. We, on the other hand, focus on investment that reduces the costs of production for domestic producers alone. This could include transport-related infrastructure such as internal road and rail networks but may be more relevant for nontransport infrastructure such as electricity, which is likely to mostly benefit domestic firms. Moreover, the above papers focus on comparing infrastructure investment between trading countries with different endowments, determinants of infrastructure such as geography, and effects of infrastructure on the volume of trade. Our focus is on whether trade can be a determinant of infrastructure provision and how the decision-making process affects this relationship. We compare investment in open and closed economies under social planning and under lobbying activity. In this sense, the specific infrastructure issues we address are somewhat orthogonal to the above studies. (4)

Section 2 develops a simple model of infrastructure provision and compares open and closed economies under a social planner. Section 3 extends the model to a lobbying economy. Section 4 concludes the paper.

2. The Model

We examine the levels of infrastructure investment in an economy when it is open and when it is closed to international trade. Our focus is on the home country, which trades with a foreign country. In both the open and the closed economies, there are two firms producing in the home country. In the open economy, there is one domestic firm and it competes against a foreign firm in both the home and the foreign country. These firms produce only in their home country. There is no foreign direct investment. In the closed economy, both firms produce in the home country. This framework allows us to focus on the key difference between the open and the closed economies--the strategic advantage against foreign firms. (5)

We compare the outcomes in these two trade regimes in the home country under different decision-making processes--when infrastructure is provided by a social planner and when it is provided by a government influenced by producer lobbies. The economy has two goods, X and Z, the latter being the numeraire good, both of which can be traded. Good X is produced under imperfect competition. The assumption of imperfect competition may be reasonable for manufactured differentiated goods, which account for most exports, including those from developing countries (Rauch 1999). (6) The market for the numeraire good is assumed to be perfectly competitive. Infrastructure affects the productivity only of the X sector. It has public good characteristics in that it reduces the cost of all firms operating in the X sector in the country where the investment occurs.