CONCLUDING REMARKS

The review of the theoretical literature on investment in this chapter presented the conventional demand-side theories of investment, theories of irreversible investment under uncertainty and the supply-side theories of investment. The review of the empirical literature included individual country and cross-country comparison studies of investment at firm, industry and aggregate levels. Findings based on various investment models and econometric techniques vary depending on individual industry and country specific factors and measures of uncertainty.

Moreover, it seems that investment behavior is influenced by political environments and by the efficiency of institutional arrangements. Most empirical studies are not based on testing the theoretical models, but are motivated by them. Broadly speaking, different models serve various purposes and there is no single model even for the same problem in the same country. Agenor and Montiel (1996, p.12) state that ‘the standard analytical tools of modern macroeconomics are indeed of as much relevance to developing countries as they are to industrial countries, but that different models are needed to analyze familiar issues’.

In sum, the accelerator model of investment assumes a fixed capital-output ratio and therefore is restrictive as the substitution between factors of production is constrained to zero. Similarly, the cash flow models, according to which optimal capital stock is merely affected by internal cash flows, do not consider any roles for the substitution of production factors. The Jorgensonian

98

neoclassical investment models, however, allow for substitution of input factors of production as a crucial element of production and cost, hence these models remain a common reference in the studies of the neoclassical theory of domestic investment. The distinguishing feature of the neoclassical models is that they are based on an explicit model of optimization behavior that associates the optimal level of capital stock to capital prices, interest rates, output and tax policies. Jorgenson assumes a perfectly competitive market. Yet, in practice, investment decisions can be affected by capital market imperfections. The literature accordingly identifies financial factors in causing and intensifying financial market imperfections in the economy. To study the effects of capital market imperfections on investment, a range of variables that may represent financial constraints are incorporated to the reduced form investment models, of which the most commonly used ones are cash flows. The literature often divides firms into sub-samples for which the extent of financial constraints is expected to vary. In addition, a growing body of literature emphasizes on the effects of uncertainty on investment behavior. At large, the findings support that increased uncertainty adversely affects aggregate and disaggregate investment patterns.

Since investment models have been mainly developed for market-driven economies, they are only expected to be partially applicable for studying investment behavior in mixed-market, developing or resource-based economies. The latter are usually faced with soft budget constraints, where ambitious investment and output targets replace profit- or value-maximizing objectives. In fact, the role of governments in allocating resource rents to various economic sectors gives rise to non-market determinants of investment in these countries. Hence, investment behavior in the respective economies could be distorted, leading to under-investment in firms which could potentially be profit-making or over-investment in loss-making firms as a result of inefficient credit allocations by the authorities. In particular, in oil-rich and oil-based developing economies such as Iran, investment could be chiefly oil-driven. Therefore, in determining domestic investment in these economies, the influence of market-oriented variables as specified by the conventional investment models may be relevant, but partially.

Further, the survey of the literature on the resource curse thesis provides several explanations in addressing this paradox of plenty. At large, the literature identifies a range of arguments for the curse such as the structuralist and the Dutch Disease theories, the paradigm of the rentier state, volatility of resource prices and revenues as well as institutional causes. Nevertheless, the

99

question as to what causes the ‘curse’ rather than ‘blessing’ in resource-rich and developing economies cannot be answered by a single explanation. This is because these economies differ in their type of government, economic policies, political institutional economic system and international relations. This suggests that rather than trying to impose some sort of generalization, a case-by-case or a country-specific approach would be more appropriate in providing explanations for the resource curse. Therefore, the focus of this study is on the oil-rich and oil-based economy of Iran due to the country’s distinctive economic, political and institutional structure. As discussed in Chapter Two, the control of the state over resources coupled with different policies adopted by the government during the period under study brought about structural shifts in actual investment patterns within the Iranian economy by altering the distribution of financial resources towards services and manufacturing. This suggests that the country has been the subject of the Iranian-type Dutch Disease and, therefore, it is important to empirically evaluate this argument in the context of Iran’s actual investment behavior during the years under consideration.

In brief, the Iranian economy has become an oil-based economy since the 1950s. The country has undergone dramatic political and economic upheavals during the period under study including the big nationalization and the introduction of the Islamic financial system during the early 1980s. Throughout the study period, the state has enjoyed a high degree of control in distributing a sizable share of oil income to decide on the pattern and pace of investment, and has played an influential role in the evolution of political structure and institutions of capital accumulation in the country. Also, the extent of government development expenditure for capital accumulation has been highly influenced by the expected oil income and its availability during the years under study. However, the availability of oil income as an important source of financing investment has been subject to uncertainty and Iran’s ability to borrow from the international capital markets has been restricted. Furthermore, the country’s economy is a partial market economy and the allocation of financial resources to firms in various economic sectors is not completely driven by the market mechanism. Having reviewed the literature on investment, and given Iran’s unique political economy and institutional setup, some interesting questions arise that are in line with the objectives of this study:

1. (i) Are the modified neoclassical-accelerator type investment theories relevant and applicable at all in the context of the mixed-market, oil-rich and oil-based economy of

100

Iran? (ii) If yes, how successful are these models in describing aggregate- and sector-level economic determinants of investment in the Iranian economy?

2. To what extent can the modified neoclassical-accelerator type investment models, to which oil-driven financial constraint measures are incorporated (as specified by cash flow models), explain investment patterns in the country?

The second question further allows accommodating the principles of the theoretical framework of investment under uncertainty as financing investment in the Iranian economy are expected to be driven by the availability of oil income, and hence subject to uncertainty.⁴⁴

Consequently, the next chapter first aims at discussing the theoretical propositions behind the hypothetical relationships based on the modified neoclassical-accelerator type models of investment; the chapter also describes the methodology employed in the empirical analysis, namely the cointegrated vector autoregressive (CVAR) method, to determine the theory-consistent long-run cointegrating relationships. This is because a cointegration interpretation within the neoclassical-accelerator type framework is supported in the literature as an enriching method to model investment behavior (see among others, see Gerard and Verschueren, 2000).

Moreover, employing the CVAR methodology, the identification and validation of long-run relations are supported by the evidence rather than by imposing them as a priori and allows producing new insights by testing economic phenomena associated with more than one economic theory.

44 See Chapters Five (Sections 5.4.5) for a discussion on oil-driven financial constraint measures used in this study.

101

4. THEORETICALFOUNDATIONSANDMETHODOLOGY