Economic Models for Policy Making

Solomon Cohen

• Éditeur: Routledge
• Année: 2013
• ISBN:
  9781136220876
• Édition: 1
• Sujet: Business
• Sous-sujet: Business General

Over the past decades, many different kinds of models have been developed that have been of use to policy makers, but until now the different approaches have not been brought together with a view to enhancing the systematic unification and evaluation of these models. This new volume aims to fill this gap by bringing together four decades' worth of work by S. I. Cohen on economic modelling for policy making. Work on older models has been rewritten and brought fully up to date, and these older models have therefore been brought back to the fore, both to assess how they influenced more recent models and to see how they could be used today.

The focus of the book is on models for development policies in developing economies, but there are some chapters that relate to economic policies in transition and developed economies. The policy areas covered are of typical interest in developing and transition economies. They include those relating to trade liberalization reforms, sustainable development, industrial development, agrarian reform, growth and distribution, human resource development and education, public goods and income transfers. Each chapter contains a brief assessment of the empirical literature on the economic effects of the policy measures discussed in the chapter.

The book presents a platform of economic modelling that can serve as a refresher for practising professionals, as well as a reference companion for graduates engaging in economic modelling and policy preparations.

1Introduction

1 Focus on Economic Models for Exploring Policies

This book has some resemblance to renovations of recipes, which existed for years ago, to fit them to changing expectations and tastes of the day. There is the attractive result here that the renovations, recipes, and ingredients fit with each other, have together a common history, and are as topical today as they were then. While most of the content of the book comprises adaptations of research work done over some four decades, each chapter has been practically rewritten so that they reinforce each other and form a coherent whole that would be of special interest and immediate relevance to today’s economic modellers, policy analysts, development advisors, and the teaching profession in the related areas. In this first chapter, we introduce the lines along which the renovation venture is focused, formulated, and implemented, and the reasons why it is worthwhile to undertake this renewal attempt. We shall also discuss the outline of the book and the linkages between the chapters, and end with concluding remarks.

To start with, it is important to underline where our focus will be within the various categories of economic models. There are pure models for formulating theories; and a wide range of applied models for testing theories, explaining events, measuring relations, forecasting variables, planning development, policy analysis, for decision making, and for teaching purposes as well. Our focus is on applied models for policy analysis and policy making, which can simultaneously serve for teaching purposes. Most of the models are economy-wide models applied to developing economies, with a few on transiting and developed economies. The book also includes several partial models.

Analytical frameworks that qualify for what one would call today economic models date back from the nineteenth century, such as the theoretically oriented demand and supply schedules of Marshall; and even before that, in the eighteenth century, there was the empirically oriented tableau economique of Quesnay. But, the modern use of the term economic models was first introduced by econometricians of the 1930s, and became the recognised medium of economic sciences among theoretically and empirically oriented economists from about the 1950s onwards. It has become conventional to see economic models as being either theoretical or empirical: the first dealing with theory and the second dealing with applications.¹ There is also a tendency, not altogether correct, to conceive the first as being more suited for studying micro aspects of economic processes, while the second as being more suited for macro aspects.

The two categories, theoretical and empirical, have different epistemologies, roles, characteristics, and limitations. However, there are cases of overlapping, complementarity and reinforcement between the two groups. There are no also strict rules for developing and using either of them. The result is a fluent and shadowy landscape of models, thus creating significant challenges and difficulties in the systematisation, transparency, and evaluation of the developed models and their use.

Generally speaking, occurrence of these challenges and difficulties is less frequently encountered among the theoretical/mathematical models as compared to the empirical/applied models. There are various reasons for these differences. In epistemic terms, pure economic models are conceived as idealised frameworks for isolating, analysing and formulating crucial economic relationships. In contrast, applied economic models are built constructions of conceived worlds that necessarily contain additional properties and approximations that the modeller introduces to assure some degree of coherence between the presumed theory and the available data. It is often not possible simply to confront theory with data, since the data contain many more things than the theory. In applied modelling it is necessary to reformulate and reconstruct the model to suit the particular situation. For instance, Morgan and Knuuttila (2012) observe that as properties are added and attributed to the modelled economies and their behaviour, the model may start to look like an experimental construction rather than an idealised representation of the actual system. According to Lucas (1980), the fictional, artificial, and simulated elements that the modeller may introduce can allow for applied laboratories in which policies—which would be prohibitively expensive to experiment with in actual economies—can be demonstrated and tested out more quickly and effectively. Much earlier, Haavelmo (1944, 1964) and Tinbergen (1956, 1963) thought along similar lines by postulating that applied models should not be treated as matching devices, but as experimental designs. Today, many prominent economists² agree that the way applied model construction takes place is more of an intuitive and creative activity than one of abiding to strict and verifiable rules that fit theory to data.

Given the nature and purposes of applied modelling assignments, it can be expected that there are no generally agreed upon scientific rules for economic modelling that the profession abides with in executing these assignments. Each modeller, applying his or her artistic and creative skills, can excel by innovating along newer avenues; and if fortunate enough, he or she would end by successfully demonstrating and validating the construction and deriving valuable findings. These endeavours have significantly increased our understandings of policy problems and how to manage them. But there is a cost. The accumulation over the years of mixtures of applied economic policy models that treat hundreds of different problems and policies in different contexts makes the field of policy modelling almost beyond systematic comprehension. Add to this that with each slight difference or variation in assumptions, formulation, closure, measurement, and specifications of the simulated policy, there will always be differences in results and findings. These differences are read but are seldom subjected to further studies. It is as in many areas of art, when every singer comes up with their own songs. There is little or no effort made to explore common backgrounds. In such a mushrooming world of economic models for policy making, there is a genuine need for consolidation rounds: bundling related models and studying them in concert. Related models can be bundled in terms of their treatment of the same problem, or models that use similar data, or models constructed by the same author. By bringing together in this book various related economic models for policy making by the same author, integrating them in retrospect, and performing appropriate adaptations to highlight standards and commonalities among the models, it is hoped that the bundle contributes in some modest ways to enhancing the systematic unification and evaluation of policy modelling.

There is more reason for taking up this retrospective renovation. It is typical of applied policy modelling studies that their results are read and appreciated at the time of their publication and dissemination, but the technical and research details, devices, and approaches that were developed in the process are forgotten thereafter. It is usual that most of the readers’ attention goes to the policy findings, while the technical details are often lost and overshadowed by the policy topic. The technical details may be of interest to a relatively smaller number of readers, and the interest may be of a timely nature. As a result, useful modelling devices, methods, and approaches that were developed in the contexts of past policy research tend to vanish. New applications have to develop their own devices, methods, and approaches; they could have benefited from previous works, if appropriately bundled and systematised. Moreover, it is also true that many of the policy problems of the past are still with us today, albeit in modified shapes and in many more countries. It is hoped that when the modelling details and policy problems are brought together and integrated, as this book will do, there can be a fruitful use of the retrospective outcomes, as they prove to be adaptable and applicable to more policy applications.³ Some past works and lines of thought can be only effectively linked in retrospect, thus making the linking of works an essential task in scientific research.

2 Outline: Method and Content

The selected policy models in this book and their arrangement are the outcome of two considerations: the primary consideration relates to modelling methods; the secondary consideration relates to policy content: aspects of the setting, substance, and topic of the particular models. Regarding modelling methods, the next chapter will discuss technical issues in policy modelling. In this chapter we comment on the policy content.

Chapter 2, as was just mentioned, will deal at some length with methodological issues. On methods, it is sufficient to mention at this point that the book is divided into two parts: we deal in the first part with ten economy-wide policy models and in the second part with six partial or theme models. The economy-wide policy models belong to three model categories: the combined econometric multisector model, which we dub as CEM model, and which was the mainstream in the sixties and seventies; the social accounting matrix corresponding models, which followed immediately, in short SAM (social accounting matrix) models; and the extensive family of computable general equilibrium (CGE) models, which has become the mainstream for the last three decades. The ten chapters in the first part follow this evolution in the economy-wide policy modelling; there is more on this and related issues in the next chapter. The second part contains six models relating to various policy themes, they are partial models in the sense that they are not economy-wide. Methodologically, they follow different modelling approaches to be reviewed in due course.

It seems natural that the primary consideration in the organisation of the book goes to methods. That settled, we can now introduce aspects relating to content of the various models. First, there are the ten economy-wide policy models. These models can be grouped in terms of the country context to which they apply: thus, models are grouped into those for developing, transiting, and developed economies. Among the ten chapters, six are devoted to developing economies. These are followed by two chapters on transiting economies, and two chapters on developed economies. Second, within each of these country contexts the various models are arranged more or less following their periodic occurrence. This helps in appreciating the changing interest of development policy modelling over time in response to changing realities. Furthermore, a periodic ordering would give insight into the progressive refinement of modelling methods, approaches, and content over the recent past. In this sense, there is a positive correlation between method and content.

In this first chapter, we introduce the backgrounds to the various models of each chapter and their relations to each other. It is unavoidable in this introduction not to mention circumstantial remarks, since the circumstances at the different times have played their role in the particular specifications and applications of the models. To keep track of the main lines, such circumstantial remarks will be shifted as much as possible to endnotes.

3 Economy-Wide Policy Models

3.1 Development Context

Models applied to developing countries form the majority of the book. They represent alternative responses in the later sixties, seventies, and eighties to the felt disappointment with the development planning effort in raising the welfare of the populations at large in developing countries. In the opening decades of development economics, mainly in the 1950s and 1960s, most development planners, economists, and statisticians identified economic development solely with economic growth. The central development objective was the growth of the GDP, and sometimes the GDP per capita. A superficial view, but very prevalent among development economists at the time was that where domestic and foreign savings are insufficient to promote investment and growth, then infusions of foreign aid (that is capital and exchange), if large enough, are by themselves sufficient to induce economic growth and economic development. Next to macro econometric models that forecasted economic growth, and multisector models that advised on industrial development and foreign trade, it was very convenient at the time to construct and apply so-called gap models to developing countries. These gap models calculated for individual developing countries the kind and height of the gaps that constrain economic growth, and the required infusions and composition of foreign aid to reach predetermined targets of GDP growth. These and related models formed the technical framework for the advisory work of the United Nations Committee for Development Planning (UN/CDP), which acted as the platform for launching and revising various plans for the First, Second, and Third Development Decades, and made recommendations on targeted economic growth and required foreign aid for the developing world.⁴

To the surprise of many observers, as data were gathered and analysed on actual performances in developing countries a new picture emerged. It became recognised that in spite of attaining some reasonable growth rates in their GDP, most developing countries were showing increasing income inequalities, surges in the population falling behind poverty lines, accompanied with increasing underemployment and unemployment, and deteriorating food and other living conditions. With the exception of a few lucky countries in East Asia and the Far East that were successful in combining growth with redistribution, the analysed data for the developing world revealed the contrary.

The new picture brought a revision in interests and insights by economists and planners dealing with development problems. Four different types of reactions emerged as the response to disappointments in the achieved progress in social welfare. The responses were critical, to varying degrees, of the central role of the GDP in development planning, the models, policies, practices, and expectations of development planning. The four types of reactions can be characterised as follows.

1. The pessimistic perspective saw development planning and development policy as meaningless under the ‘soft state’. The postulates were that the skewed distribution of socio-political power hindered state power and economic development, and that radical reforms in the societal structure were necessary if development policy was to achieve its goal of a speedy and balanced social and economic development.
2. The social development perspective was more positive in tone. It saw the fault lying in the planning framework’s misplaced focus on GDP. The perspective opened the way for reformulating development models and plans in terms of incomes of social groups, employment and basic needs, and redirecting policy making accordingly.
3. The redistribution with growth perspective brought structure in the picture, and sounded more realistic in tone. It emphasised that there are dualities and linkages in the economy between rich and poor population groups, and between high-productivity and low-productivity activities that determine development structures, growth, distribution, trickle-down effects, and moving-up mechanisms. The advice was to focus on economy-wide modelling of dualities and linkages, and use them in designing and implementing packages of development strategies aimed at redistribution with growth. Methodologically, this perspective is associated with the popularity of the SAM serving as a helpful statistical tool in highlighting dualities and linkages in the economy-wide circular flow.
4. The free markets perspective took off and became the mainstream response for the time being. It is argued that in the longer run, market-oriented economies develop more rapidly in terms of both growth and equality than planning-oriented economies. Consequently, development policy should be more reliant on the free operation of the market forces of supply and demand and market-determined prices. Where state intervention is necessary due to market failure, state intervention should complement, and not be a substitute for market forces. Methodologically, CGE models came just in time to allow for the realisation of the free-markets perspective in policy modelling. The flexibility of CGE models allowed for extending these models to treat new policy areas such as liberalisation reforms, environmental sustainability, energy resources, fiscal policies, and so on.

The four responses are well represented by the models in Chapters 3 to 8, which were developed and applied consecutively to various developing countries. Chapter 3 represents the pessimistic perspective response. It presents a model that stylises socio-political power in a feudal-oriented system, that is rural India of the 1960s, and in a reform-inclined regime, that is Chile under Allende; and it explores policy alternatives under the two contrasting settings. Chapter 4 represents the social development perspective. It builds up a social development and a basic needs approach and incorporates supporting variables in a combined macro econometric and inter-industry model, and applies the model to Korea. Chapter 5 represents the distribution and growth perspective with its emphasis on linkages and leakages in the circular flow between social groups, firms, government, and sector activities. The SAM...

Table des matières

• Cover
• Half Title
• Title Page
• Copyright
• Dedication
• Contents
• List of figures
• List of tables
• Preface
• 1. Introduction
• 2. Some essentials in economy-wide policy models
• 3. Socio-political regimes and economic development: exploratory models on agrarian reform in India and Chile
• 4. Social economic development goals in economy-wide policy models: an application to Korea
• 5. Growth and distribution in SAM models: various applications
• 6. Simplified statics and dynamics in the CGE model: parameterisation and simulations for Indonesia
• 7. Growth with redistribution through liberalisation with restructuring: a CGE policy model of Nepal
• 8. Sustained development of land resources: a policy model for Sudan
• 9. Simulation results of SAM models for transiting economies: Russia falls and China rises
• 10. Transiting from fixed- to flexible-price regimes: SAM-CGE models of Poland and Hungary
• 11. Public spending multipliers in extended SAM models for a developed economy
• 12. Fiscal policy simulations in adapted CGE models: the Netherlands
• 13. Normed planning of human resource development: a roadmap model for Ethiopia
• 14. Labour market imbalances and adjustments: forecast model with RAS component
• 15. Privatisation decisions during transition: a CBA model applied to Poland
• 16. Economic policy solutions to social queuing problems: a random sampling model
• 17. Modelling convergence in economic growth between rich and poor countries
• 18. Modelling of distinctly behaving economic systems: theory and applications
• Notes
• References
• Index