Ebook Exporting And Economic Performance: Firm-Level Evidence For Spanish Manufacturing
One of the factors that are thought to be important to make some firms more productive than others is exporting. Bartelsman and Doms (2000) survey of the literature on productivity that uses longitudinal micro-level data sets points out to the link between productivity and exporting as one of the factors this literature has focused on (the rest of factors are regulation, management/ownership, technology and human capital). Studies by Aw and Hwang (1995) on Taiwan; Bernard and Jensen (1995) (1999) on the US; Bernard and Wanger (1997) on Germany; Clerides, Lach and Tybout (1998) on Colombia, Mexico and Marocco; Aw, Chung and Roberts (2000) on Taiwan and South Korea; Girma, Greenaway and Kneller (2002) on the UK, provide evidence on the fact that export oriented firms are more productive than non-exporters.
Sunk costs are the main argument outlined to explain why exporters are more efficient than non-exporters, in particular the existence of higher sunk entry costs for exporters with respect to non-exporters. The argument comes from models of industry dynamics Jovanovic (1982) and Hopenhayn (1992)- and applies also to entry and exit to export markets as suggested by Aw, Chen and Roberts (1997). According to this argument, differences in sunk entry costs can explain productivity differences between exporters and domestic-oriented firms. Building on these ideas Roberts and Tybout (1997), Clerides, Lach and Tybout (1998) and Bernard and Jensen (2001) have developed models of the decision to export. The result that firm’s previous export status is a determinant of the decision to export is interpreted, in term of these models, as a favorable evidence to the existence of sunk entry cost in the export market.
In a previous paper, Delgado, Fariñas and Ruano (2002) measure total factor productivity differences between exporters and non-exporters on the basis of a sample of Spanish manufacturing firms. The empirical analysis confirms higher levels of productivity for exporting firms relative to non-exporting firms. With respect to the relative merits of the selection and the learning hypotheses proposed to explain the greater productivity of exporters, the paper finds evidence favorable to the self-selection of more productive firms into the export market. It was much harder to find evidence in favor of learning effects in the data set. For the whole sample of manufacturing firms we do not find any systematic evidence consistent to learning-by-exporting. However, restricting the sample to the group of younger firms we observe that post-entry productivity growth is greater for young entering exporters than for young entering domestic firms without contact to the export market. With the exception of the latter result, our empirical findings are very much in line with those reported in the literature -Bernard and Jensen (1999).
The first purpose of this paper is to investigate further economic performance differences between exporters and non-exporters. We explore systematically the magnitude of these differences for various performance measures such as labor productivity, investment, wages, the composition of labor force, R&D activities, etc. Estimates of export premia are reported after controlling for time, industry, size, age and other characteristics of exporting and non-exporting firms. Furthermore, we explore if firms with different trajectories between the export and the domestic market i.e. entering, exiting firms, show systematic ex-ante differences in the level of performance and ex-post differences in their evolution.
The second purpose of this paper is to measure total factor productivity differences between exporters and non-exporters, measuring these productivity differences by the estimation of production functions. Productivity shocks are assumed to be an unobserved firm-specific effect that can be recovered as the difference between actual and predicted output and that are allowed to take a very general form. Two advantages can be identified from using this approach. The first one refers to the set of assumptions that is required to get unbiased estimates of total factor productivity when using index numbers. Some of them, as the assumption of constant returns to scale, may be relevant for measuring correctly productivity differences between exporters and non-exporters.
The estimation of production functions do no require some of these assumptions. The second advantage refers to the benefits that can be derived from the application of GMM estimators that we use. In particular, these estimators permit to control two likely sources of bias in the OLS results: 1) the elimination of unobserved firm heterogeneity that is time invariant and 2) the use of lagged instruments to correct for simultaneity bias produced by the effect of productivity of the firm itself on the input decisions. The paper estimates productivity differences for exporters and non-exporters using an unbalanced panel of Spanish manufacturing firms over the period 1990-1999. We apply estimators developed in Arellano and Bond (1991), Arellano and Bover (1995) and Blundell and Bond (1998). An illustration of these procedures in the context of the estimation of production functions can be found in Blundell and Bond (1999) and Bond (2002). Griffith (1999) contains a similar application to the analysis of productivity differences between foreign and domestic owned establishments.
The rest of the paper is organized as follows. Section 2 describes some characteristics of the data set used in the analysis and presents some basic evidence on the magnitude of performance differences between exporters and non-exporters. Section 3 discusses the method used and presents the econometric estimates of production functions for the whole sample of firms and three industries; textiles and clothing; wooden products and furniture; and food industry. Conclusions are placed in section 4.
Download
PDF Ebook Exporting And Economic Performance: Firm-Level Evidence For Spanish Manufacturing
Posted in :
