Search

My paper The delayed impacts of plant closures in a reformulated Leontief model showed that the time-varying impacts of changes to a regional economy could be approximated usefully by a simple elaboration of the Leontief inverse (Cole, 1988). I have applied the method in a variety of situations, typically to calculate the shifts in community income and its distribution following a major catastrophe. In all cases I use a single or multi-region social accounting matrix. Since Jackson et al (1997) have been unable to reproduce my results I shall begin with a step-by-step demonstration. Then I shall show that their failure lies in their treatment of the external (rest-of- world) sector. This appears to arise from their overly restrictive definition of input-output analysis in relation to other types of model, and their confusion as to the difference between simplification, approximation, and model validation.

In Closing the Case on Closure in Cole's Model Randall Jackson and Moss Madden critique my solution to the time-lagged regional input-output model. My rebuttal demonstrates that they: i) interpret my model incorrectly, ii) are inconsistent in their own specification, and iii) make inapplicable numerical comparisons between short- and long-term solutions. I show also iv) the correct solution to the totally closed model is neither trivial, nor singular, and v) this resolves the apparent paradox of infinite multipliers in a finite world. Moreover, vi) in my model, injecting exogenous demands equal to export earnings correctly reproduces the base output levels. The standard model does not meet this test.

The paper uses a reformulated input-output approach to calculate the medium-run regional economic impact of plant closures or similar changes taking nun account expenditure lags. The approach lends to a straightforward modification of the generalized Lecontief inverse method. An example is used show that the inclusion of expenditure lags m impact analysis ran he quite substantial and that the usual approach of partitioning the accounts into infinitely last and infinitely slow activities, and the Consequent assumption that equilibrium is reached, is an especially weak approximation. Revising the contribution of all activities to account for expenditure lags changes both short- and medium-term predictions by up to a factor of two. The approach has a variety of applications, but may be especially useful for calculating the distributed impact over time on household income, local taxes, employment, and other businesses.

This paper1 describes a Social Accounting Matrix (SAM)-based method for evaluating disaster preparedness and recovery strategies. The first section of the paper explains the overall approach and its extension to the evaluation of specific components of lifeline systems in a small Caribbean island. The next section explains the relationship between the physical and economic parameters of the energy-electricity-water lifelines and the tourism sector on the island and describes the construction of the corresponding accounting framework. The diachronic multipliers calculated from this extended matrix are used to determine the impact of potential hazards for tourism and other economic activities on the island. The case of a water storage tank, providing back-up supply to the major hotels, is used to illustrate the approach. In the final sections, the example is elaborated to highlight trade-offs between economic and non-economic costs for particular businesses or households. The events and strategies described then may be combined into an event-based scenario analysis.

I. INTRODUCTION The cultural accounting matrix approach (CAM) in the present study is a version of the extended input-output, or social accounting matrix approach, that analysts widely apply at the national and regional levels. Hicks (1942) introduced the term "social accounting." Subsequently, Stone and his co-workers (United Nations, 1953, 1968; Stone, 1973) extended this concept to develop the System of National Accounts (SNA) framework. The innovation of cross-tabulating the conventional double entry national accounting system into a single entry social accounting matrix created both improvements in accounting consistency and the opportunity to understand the role of particular economic flows. This technique thus extended Leontief's (1951) input-output system and offered a new tool for understanding the relationship between economic change and social distribution (Pyatt and Roe, 1977). Adapting SNA to the situation of less developed countries requires rethinking many international sectoral and social classifications in order to reflect local circumstances, available data, and policy needs (see Ward, 1977; Adelman et al., 1988). The taxonomy of the CAM and the corresponding cultural division of labor introduced here originates from anthropologists in the tradition of Furnival (1948), Herskovits (1961), and Glazier (1985). These authors write of the interrelationships and the correlations between economic, political, and social relationships as a function of cultural identity. For these authors, culture is not a primary trait but is part of an adaptive behavior by a group and by those who interact with it. This adaptive behavior often serves as a social construct to maintain existing segmentation--for example, through kinship and ethnic affiliations (see Glazier, 1985). One can compare this assumption to that of economists who argue that economic segmentation is perpetuated because it is functional in reproducing a capitalist hegemony (Reich et al., 1973) or that the very exclusiveness of groups with access to specific incentives determines their likelihood to act in a group-interested manner (Olson, 1971). In the context of this paper, the cultural division of labor is the set of economic interactions between ethnically identifiable populations (see Hechter, 1978). If these populations have markedly different and persistent economic characteristics and/or relationships with the economy, then the cultural division of labor is segmented. This segmentation may be both vertical and horizontal and may exist on both the supply- and the demand-side of an economy. Horizontal segmentation exists on the supply-side of an economy when people of particular cultures are associated with particular sectors of production--for example, when indigenous people predominate in traditional activities, such as Europeans in manufacturing and Asians in commerce. Vertical segmentation has arisen, for example, when an economy is largely owned and run by Europeans but operated by indigenous people or non-European immigrants. A related characteristic of vertical segmentation arises when people with comparable skills or education but of different ethnicity, gender, or religion receive different wages for the same work. Such a marked vertical cultural division of labor implies uneven wealth distribution and social inequalities. Orthodox theory does not explain the persistence of this segmentation (see Arrow, 1971; Atkinson, 1980). Explaining segmentation typically requires some analysis of the historical evolution of the economy (see Reich et al., 1973). A cultural division of labor is not confined to small societies. However, the divisions often are sharper and sometimes easier to measure empirically, and some of the institutional factors are more apparent (Benedict, 1978; Henry, 1989). Indeed, culture related factors permeate a small economy. In small islands, the problems of scale (see Kuznets, 1969; Jalan, 1982; Srinivasan, 1986; Rojas, 1989) are compounded pathologically by those of the cultural division of the economy (Horowitz, 1971; Clarke, 1984; Ward, 1989). …