Technological performance, economic performance and behaviour: a study of Argentinean firms during the 1990s.

SUMMARY

Based on empirical information for the Argentinean case, it is claimed that the biological metaphor which states that an open market is sufficient to select the best performing firms is often invalid, in a macroeconomic context characterised by high degrees of uncertainty. In such an environment, it could be rational for profit-seeking individuals to avoid any long-term tie to any particular policy. Like policy-making itself in Argentina during the 1990s, individual decisions were driven by short-term objectives, which hampered technological performance. The paper uses Survey information in econometric models to demonstrate that two distinct patterns of behaviour emerged: one corresponding to technological performance and the other to economic performance.

KEY WORDS

macro-micro economic links; technological behaviour; instability; uncertainly; firm performance; Argentina

1. INTRODUCTION

In the context of structural adjustment reforms, developing countries have been challenged to adopt a series of policies towards market liberalisation in trade, capital and labour. The prevailing orthodoxy in economic theory justified this switch in policies, claiming that it would improve 'economic health' and therefore would foster growth and efficiency. For instance, from a neoclassical point of view market liberalisation ought to provide positive incentives for technological change.

Thus, according to certain dynamic approaches, grounded in endogenous growth theory, it has been argued that intensifying international exchange would open channels of communication that facilitate the transmission of technical information (World Bank, 1992). This faster accumulation of knowledge would imply a more rapid reduction in the cost of production (Grossman and Helpman, 1991). Besides, it has also been claimed that openness would replenish the local economy given that international markets provide access to up-to-date knowledge bases (Atiyas, Dutz et al., 1992) and enable entrepreneurs to introduce new varieties at a faster pace (Grossman and Helpman, 1991). Moreover, international competition would encourage the use of new and distinctive ideas and technologies in production (Balassa, 1988) and might reduce duplication (Grossman and Helpman, 1991). Furthermore, given that firms in developing countries innovate by imitating the innovations in products and processes made by other firms (usually located in the developed world), importing capital goods that embody new technology would encourage innovation (Barro and Sala-i-Martin, 1995, Obstfeld and Rogoff, 1996). Finally, the existence of technological spillovers in knowledge (Barro and Sala-i-Martin, 1995) and in human capital (Romer, 1990) also justify openness of trade.

Similar points of view based on the presumed superiority of knowledge of international companies plus the existence of technological spillovers were commonly used as sound arguments for those advocating foreign direct investment as a source for local development (Blomstrom and Kokko, 2002; Haskel, Pereira et al., 2002), which would be motivated if the capital markets were liberalised (see Marin and Bell (2003) for a discussion).

Furthermore, capital market liberalisation would also eliminate financial repression, and as a consequence would induce a more efficient allocation of credit towards productive activities (Frenkel, 1998; Harris, Schiantarelli et al., 1996; Jaramillo, Schiantarelli et al., 1996).

These theoretical approaches, which belong to the mainstream in economics, assume that individual behaviour could be theoretically reduced to the mechanism of a stimuli-response silicon chip in the human mind which automates behaviour. As a consequence the macro environment could be analysed by adding up the actions of many 'representative' individuals, who react homogeneously. That is, both individual heterogeneity and diversity of social attitudes and values are neglected in the analysis. Some authors would compare the economic and social changes, triggered by a policy arrangement that liberalises markets, as a process similar to biological evolution and natural selection (Alchian, 1950). In such a framework, the (now freer) market should select the more efficient firms.

This paper challenges the validity of those assumptions of individual behaviour and suggests instead that at least some part of individual behaviour could be explained by an analysis of the agent's environment. As social individuals, agents will share some of the attributes that contribute to their rationality with some other agents within their environment. The environment that I will be considering is the 'national (Argentinean) system of production'; therefore in order to understand the behaviour of Argentinean firms, we will need to look first at the big picture and disentangle its outstanding features.

Argentinean economic history has been characterised by a succession of short cycles of stop and go, together with erratic macroeconomic policies (which exhibited a great variety of economic orientations that constantly alternated between orthodoxy and heterodoxy).

Consequently, a social-specific reactivity that accounted for and coped with macro structural uncertainty was developed in the Argentinean environment. The claim that guides the discussion that follows is that the persistent macroeconomic instability rewarded the short-term over the long-term, and therefore short-term strategies prevailed over technological strategies (which require long-term horizons). As a consequence economic success might not mean the same as technological success; or, in other words, patterns of successful economic performance might have diverged from patterns of technological performance, meaning that economic performance did not necessarily yield a (long-term) efficiency criterion.

Rather than focussing attention on one policy in particular, this paper emphasises the effect that macro uncertainty might create on technological behaviour. It highlights the importance of behavioural dimensions when analysing technological impacts that might be triggered by changes in macro policy, especially those changes that, in being driven by short-term macro objectives, could enhance macroeconomic uncertainty.

The empirical information used comes from the National Survey of Technological Behaviour ('The Survey'), (1) which covers the most successful period of the Convertibility decade. (2) During that decade (1991-2001), Argentina followed the recommendation of Structural Adjustment Programmes and grew at a 4% annual cumulative rate and was presented throughout the world as one of the best pupils of the International Monetary Fund (IMF). (3)

This paper is divided into four main sections, with an introduction and conclusion. The next section (Section 2) illustrates Argentinean macroeconomic volatility and stop-go cycles vis a vis those from United States during the last 70 years. The following section (Section 3) develops the theoretical hypothesis that was briefly discussed in this introduction. Section 4 explains the methodology that will be used in Section 5 to illustrate empirically the theoretical propositions. The conclusion summarises the results and formulates exploratory propositions on possible consequences of the changing macro environment after devaluation.

2. CONTEXT OF MACROECONOMIC INSTABILITY

This section attempts to illustrate the historical pattern of Argentinean macroeconomic instability using information on GDP rates of growth. As can be seen in Figure 1 that uses United States as a benchmark, the growth rates for Argentina have fluctuated widely. Except for the immediate pre and post World War Two (WW2) period, Argentinean series are much more volatile.

[FIGURE 1 OMITTED]

Table 1 confirms what was suggested in the plot. The standard deviation of the Argentinean series is double that of the American series 1950-2002. The coefficient of variation, which puts into perspective the absolute variability by comparing with the average rate of growth, shows that shocks have been much more severe for the Argentinean economy even when considering the period of WW2. In the same vein Argentina has always shown a higher probability of negative growth.

In an attempt to find evidence on economic cycles, I have constructed periodograms. (4) A periodogram graphs the spectral amplitude of a series. Peaks represent the length in time of each of the cycles that are verified. Given the technique used, once a cycle is identified it is filtered in the identification of longer cycles (that is to say that if a cycle of four years is identified, it is obvious that cycles of eight, twelve, etc. years will also exist; however, they will not inevitably appear as peaks in the graph). The y-axis shows the spectral density that points out the intensity. (5) Therefore, a series with low values of spectral density could be interpreted as a soft or acyclical series. Similarly, when a series shows peaks uniformly distributed over successive periods, it will mean that the variability is random.

The first evidence highlighted by the graphs for the period 1930-2002 (Figures 2 and 3) is that the length of Argentinean cycles are much more erratic than the length of American cycles. For the latter, cycles last seven and a half years approximately. In contrast, given the randomness attached to the periodicity of Argentinean series, it is more difficult to find systematic cycles there; however, if we were to do so, we could say that Argentinean cycles last less than four years.

[FIGURES 2-3 OMITTED]