Infrastructure investment and Spanish economic growth, 1850–1935

https://doi.org/10.1016/j.eeh.2006.06.002Get rights and content

Abstract

This paper analyzes the impact of infrastructure investment on Spanish economic growth between 1850 and 1935. Using new infrastructure data and VAR techniques, this paper shows that the growth impact of local-scope infrastructure investment was positive, but returns to investment in large nation-wide networks were not significantly different from zero. Two complementary explanations are suggested for the last result. On the one hand, public intervention and the application of non-efficiency investment criteria were very intense in large network construction. On the other hand, returns to new investment in large networks might have decreased dramatically once the basic links were constructed.

Introduction

This paper explores the economic effects of infrastructure investment during the first stages of Spanish industrialization. Despite abundant research on the topic, there is no agreement among historians on the impact that increases in the stock of infrastructure had on late nineteenth and early twentieth century Spanish economic growth. Research on the subject has focused to a large extent on the railroad system, which accounted for the highest share of infrastructure investment at the time, and which has been seen, alternatively, as either one of the “heroes” of Spanish industrialization or one of the explanatory factors for the country’s economic underperformance.

Tortella (1973) set the basis of the pessimistic view. He pointed out that Spanish railroads were constructed ahead of demand and, as a consequence, railroad companies faced a situation of excess capacity that made them unprofitable during most of their lives. To make things worse, the construction of the Spanish railroads had very small “backward linkages”, since imports of the necessary materials were given generous tariff exemptions that deprived the Spanish iron industry of a crucial source of demand. Later on, however, Gómez Mendoza (1983) suggested a much more positive interpretation on the subject. He stressed the importance of the transport cost-saving effect of the Spanish railroad system. According to this author, the social savings of the Spanish railroads and their effects on economic growth were much larger than in other European countries due, among other reasons, to the lack of opportunities for the development of inland waterways in Spain.

Subsequently, historians have raised several caveats about Gómez Mendoza’s optimistic interpretation. On the one hand, Tortella, 1999, Comín et al., 1998 have indicated that, despite the social saving evidence, the low density of use and the lack of profitability of the Spanish railroads constitute powerful proofs of their economic failure. On the other hand, further research has provided substantially lower estimates of the social savings of the Spanish railroads in the late nineteenth century (Barquín, 1999, Herranz-Loncán, 2002). However, at the same time, Herranz-Loncán (2005a) has pointed out that the social savings do not capture all the impact of the railroads, due to the difficulty to measure TFP spillovers. These might have been quite relevant in Spain, where nineteenth century market integration provoked a substantial change in the geography of economic activity (Tirado et al., 2002, Rosés, 2003, Pons et al., 2007), as well as the gradual convergence in living standards across Spain (Rosés and Sánchez-Alonso, 2004). In summary, the economic impact of the Spanish railroads is still the object of an open debate, in which no agreement has been reached among researchers so far.

This paper re-addresses the question of the growth impact of Spanish infrastructure investment during the late nineteenth and early twentieth century from a different viewpoint. Instead of focusing on railroads, as most previous research, it adopts an aggregate perspective that takes into account the whole infrastructure sector. Using a new data set of Spanish infrastructure for the period 1845–1935 and the most recent estimates of Spanish output and capital formation, I estimate a vector autoregressive (VAR) system that allows analyzing the relationship between infrastructure investment and economic growth. The VAR framework has often been used to measure the growth impact of infrastructure in present economies (Clarida, 1993, McMillin and Smyth, 1994, Otto and Voss, 1996, Pereira and Andraz, 2005). Among other advantages, this technique allows accounting for the non-stationarity of time-series data and for the presence of mutual causation and indirect relationships among different variables. In addition, VAR methods are especially appropriate in the case of historical research, in which the paucity of quantitative information prevents from using more sophisticated techniques (e.g., Groote et al., 1999).

This paper describes how, in a preliminary analysis, the VAR approach seems to indicate that Spanish infrastructure investment did not have any significant long or short-term impact on productivity between 1850 and 1935. Previous papers that found infrastructure increases not to have any positive effects on productivity growth are, among others, Tatom, 1991, Holtz-Eakin, 1994, McMillin and Smyth, 1994, Baltagi and Pinnoi, 1995, Balmaseda, 1996, Garcia-Milà et al., 1996. However, all these works analyze mature high-income economies, for which the available evidence on potential infrastructure shortages is inconclusive (Gramlich, 1994, Fernald, 1999). The absence of productivity effects of infrastructure is more difficult to find and interpret in industrializing countries in which the most basic infrastructure is being built. Groote et al. (1999), for instance, have observed a clear positive short-to-medium term reaction of Dutch GDP to infrastructure increases between 1853 and 1913.

The outcomes of the VAR analysis may be better understood when Spanish infrastructure investment is disaggregated into two categories: local-scope infrastructure and large nation-wide transport and communication networks. In this case, the estimation results show that the former had an unmistakable positive impact on productivity and the apparent lack of growth effects of infrastructure in the aggregate analysis was actually driven by the latter. These results may be explained on the basis of the differences between both infrastructure categories. Unlike local-scope infrastructure, nation-wide transport and communication systems were characterized by two distinct features. Firstly, government intervention and non-efficiency allocation criteria were very influential in the investment decisions, and reduced the returns to some of the resources invested. And, secondly, the expansion of large networks was characterized by decreasing returns. As a consequence, although the construction of the basic network links probably provoked substantial changes in the economy, returns to additional investment were rather low. This paper provides therefore some possible explanations for other historical cases or for present LDC in which the effect of infrastructure on output may appear to be muted in econometric analyses.

The paper is organized as follows. Section 2 describes the infrastructure dataset that has been used in the estimation and offers a brief summary of the process of infrastructure construction in Spain between the central years of the nineteenth century and the eve of the Civil War of 1936–1939. Section 3 presents the model that has been used to analyze the relationship between infrastructure and economic growth, and the main estimation results. In Section 4, I discuss some possible explanations for the lack of response of Spanish productivity to investment in transport and communication networks. Section 5 concludes.

Section snippets

Infrastructure investment in Spain, 1845–1935

Most econometric analyses of the growth impact of infrastructure investment use public capital figures as a proxy for infrastructure. This procedure may be justified to some extent in the case of present economies (Gramlich, 1994, p. 1177), but it would have been misleading for late nineteenth and early twentieth century Spain, where a large number of assets, such as railroads, tramways, the telephone system, power distribution networks and some hydraulic works (i.e., more than 50 per cent of

Econometric model and results

To approach the growth impact of infrastructure investment, I use a standard production function with one aggregate output (y) and three inputs: infrastructure capital (g), machinery and equipment capital (k) and labor (n). All series are taken in logarithms:

State intervention and decreasing returns to network expansion in Spanish infrastructure investment

Late nineteenth and early twentieth century Spain is a good example of Millward’s (2004) considerations about the different intensity that state intervention had on large transport and communication networks and on local-scope infrastructure. Industries of the second group, such as urban transport and power distribution, largely consisted of small scale undertakings, which stemmed from the initiative of private firms or, in a few cases, from municipal governments with short financial capacity.

Conclusions

Using cointegration and VAR techniques, this paper shows that investment in local-scope infrastructure exerted a clearly positive impact on Spanish economic growth between 1850 and 1935. By contrast, returns to investment in large nation-wide networks were not significantly different from zero. The second finding may appear, either as a confirmation of the views of the most pessimistic Spanish railroad historians, or as a surprising result, given the significant transport cost-saving effects of

References (51)

  • Cadot, O., Röller, L.-H., Stephan, A., 1999. A Political Economy Model of Infrastructure Allocation: An Empirical...
  • Á. Calvo

    Los inicios de las telecomunicaciones en España: el telégrafo

    Revista de Historia Económica

    (2001)
  • Clarida, R. H., 1993. International Capital Mobility, Public Investment and Economic Growth. NBER Working Paper No....
  • F. Comín et al.

    150 años de historia de los ferrocarriles españoles

    (1998)
  • R. Cordero et al.

    El sistema ferroviario español

  • A. De la Fuente et al.

    Infrastructure and education as instruments of regional policy: evidence from Spain

    Economic Policy

    (1995)
  • A.W. Evans

    A theoretical model for local roads grants

    Urban Studies

    (1991)
  • J.G. Fernald

    Roads to prosperity? Assessing the link between public capital and productivity

    American Economic Review

    (1999)
  • E. Fernández Clemente

    De la utopía de Costa a la intervención del Estado: un siglo de obras hidráulicas en España

    (2000)
  • T. Garcia-Milà et al.

    The effect of public capital in state-level production functions reconsidered

    Review of Economics and Statistics

    (1996)
  • P. García Ortega

    Historia de la legislación española de caminos y carreteras

    (1982)
  • D.R. Glover et al.

    The effects of population density on infrastructure: the case of road building

    Economic Development and Cultural Change

    (1975)
  • A. Gómez Mendoza

    Spain

  • E.M. Gramlich

    Infrastructure investment: a review essay

    Journal of Economic Literature

    (1994)
  • P. Groote et al.

    Infrastructure and economic development in the Netherlands, 1853–1913

    European Review of Economic History

    (1999)
  • Cited by (0)

    I thank the financial support provided by the Spanish Ministry of Education Grant SEJ2005-02498 and the Bank of Spain. I gratefully acknowledge comments by participants in seminars at the London School of Economics and the Universities of Barcelona, Valencia and Zaragoza. I am particularly indebted to Dudley Baines, Nicholas Crafts, Jordi Pons, Leandro Prados de la Escosura, Ramon Ramon, Max Schulze, Carles Sudrià, Daniel A Tirado, Hans-Joachim Voth and two anonymous referees. This implies no responsibility on their part for the shortcomings that may remain in the paper.

    View full text