Selected Macroeconomic Models

views updated


SELECTED MACROECONOMIC MODELS A logical explanation of any phenomenon, or a set of mutually interdependent phenomena, is a model. This, referred to as theory, can be purely verbal in nature or mathematical in form. A model is an abstract and simplified picture of a realistic process, given in the form of mathematical equations. It helps in forecasting and policy analysis by government and business. Economists and public policy analysts not only require the direction of the effect of one variable on the other, but also the extent of impact and trade-offs. This necessitates casting theory in the form of mathematical equations, variable measurement, and impact calibration.

Three types of macroeconomic models were developed for India since the early 1950s. They are: input-output (I-O); computable general equilibrium; and econometric models. The objective of all models is structural analysis, forecasting and policy evaluation.

Input-Output Models

With the advent of planning in the early 1950s, focus was on models of plan-variety concerning growth and investment allocation. One aspect central to planning models is the sectoral interdependence and sectoral balance for a given macroeconomic growth rate. Input-output models form a core element of planning exercises. The economy is viewed as consisting of n producing sectors, each producing ideally a single homogeneous output. The intersectoral flow matrix charts the flow of output of a sector to each of the sectors and to final demand over a specified period, generally one year. The final demand relates to consumers, government, investment, and the foreign sector. It is a two-way table (each sector appearing in a row and in a column) describing the production structure of the economy. The ratio of output of i-th sector used as input by j-th sector, divided by output of j-th sector, represents input-output coefficient. They are technological parameters. The configuration of such n-by-n coefficient matrix represents the I-O table. It is a linear production structure. Inputs bear a fixed proportion to output, constant returns to scale prevail, and no substitution possibilities exist between inputs. The model is static. The intersectoral commodity flow matrix that yields the I-O table is in value terms.

In the intersectoral flow matrix, final demands appear as a column. Imports appear with a negative sign. They balance supply and demand. Imports, which enter as inputs in the production process, appear in a separate row. Rewards for primary inputs, wages, and profits, as well as indirect taxes and subsidies can also be shown as additional rows. If the intersectoral flows are represented at producer prices, the column sums represent the total cost of producing that sector's output. This should be equal to the corresponding row sum, suitably adjusted for taxes and subsidies. The total of row sums should be equal to the total of column sums, and the grand total represents the gross value of goods and services produced in the economy. This represents not only intersectoral flows but also final demand and disposition of total output as inputs and rewards for primary factors. This can be viewed as a social accounting matrix.

The Indian Statistical Institute built an input-output table for 1950–1951 with thirty-six sectors. The Gokhale Institute of Politics and Economics published an input-output model for 1963. The Planning Commission and the Central Statistical Organization constructed the tables for 1968–1969 and subsequently published a new one every five years. The latest table refers to 1993–1994. The various tables differ in the extent of aggregation.

The analytical problem in the solution of the I-O model is one of solving linear simultaneous equations. Planning exercise involves setting an economy-wide growth target, calculating associated final demands, and solving the I-O model to yield sectoral output targets. These output targets aid in calculating sectoral investment requirements, employment, and other related dimensions. In each of the Five-Year Plan exercises in India, these calculations have been made in much detail.

The calculation of forward and backward linkages in an economy is crucial to evaluate alternative strategies of development. Backward linkages reflect the impact of a sectoral output change on all other sectors that supply inputs to it, while forward linkages measure consequent increase in demand for other sectors' output. Key sectors in development are those that have strong backward and forward linkages. These exercises have been made for India with the help of I-O tables. Another application relates to measuring the impact of price changes of primary factors or any other inputs on sectoral prices and overall price level. The estimates represent accounting exercise and do not reflect the effect of many other important factors that underlie price behavior.

Computable General Equilibrium Models

Computable general equilibrium (CGE) models have two facets. First, general equilibrium connotes viewing the economy as a complete system of interdependent economic activities by different agents, for example, producers, households, investors, the government, importers and exporters. The interrelationships are through a web of intersectoral output flows and price connections. The second term, "computable," signifies an empirical system that can be implemented. Rules for the functioning of individual markets and the behavior of agents in the markets, that is, causal links that determine equilibrium mechanisms, are specified. The underlying production structure is the I-O model. CGE models have a close interface both with the I-O and econometric models.

In the early 1980s substantive research on CGE models for India was initiated at the National Council of Applied Economic Research. The model is maintained, frequently updated, and improved over the years. Particularly notable is the recent effort to integrate segments of behavioral macroeconometirc models with the CGE model. The model originally comprised eight sectors, including infrastructure, services, three in agriculture (food, other crops, and livestock), and three in industry (consumer, intermediate, and capital goods). Three income classes (agriculture income recipients, nonagricultural wage income earners, and nonagricultural nonwage income earners) were distinguished. Public finance and money were also later brought into the system. In recent versions, sectoral decomposition has widened. Agricultural output is determined outside the CGE model through an agricultural submodel, treated as exogenous to CGE solutions.

Since the early 1980s, the model is used regularly for making annual forecasts of important macroeconomic indicators, including output, prices, trade, and fiscal and external deficits. Forecasts are made for only one year ahead, and sequential forecast over time do not generally include dynamic interlinks. Many policy simulations have also been made over the years. Policy simulations relevant to the 1990s, which witnessed far-reaching structural adjustments and macroeconomic policy reforms, to name a few, relate to reduction in tariff and nontariff barriers to trade, lower domestic indirect taxes, changes in the exchange rate, reduction in government investment expenditures, and changes in interest rates.

Macroeconometric Models

India has had a history of macroeconometric modeling dating back to the 1950s, unparalleled among developing countries. Many models were constructed, and to date about fifty models, covering different time periods and focusing on issues relevant to those times, have been constructed.

Macroeconometric models for India are based on annual time series data. They are estimated by econometric methods and are subject to statistical inference. They are also subject to in-sample validation, in terms of their ability to replicate historical series. Almost all of them have had a policy focus. Most of the models have had only a short- to medium-run character. They are dynamic in nature. Models have been concerned with the level of economic activity, aggregate and sectoral, price behavior, fiscal and monetary phenomena, intersectoral linkages, private investment and its linkages with public investment, consumption and savings, public sector resource mobilization, current and investment expenditures and their composition, budgetary deficits and pattern of financing, trade flows, balance of payments, exchange rate and nexus between the twin deficits, budgetary and external, among others.

The theoretical approach to macroeconometric modeling in India has been eclectic in character. Specification of components of final demand is Keynesian. Unlike in the simple Keynesian approach, the economy is not entirely driven by effective demand. Supply constraints are well jeweled in the models. Agriculture activity is determined by land, a limiting factor and natural resources. Harnessing these resources is facilitated by capital. In the nonagricultural sectors, output is viewed as constrained by stock of capital, the scarce factor and its utilization. Capital utilization is conditioned by effective demand, which, in turn, is influenced by level of overall economic activity in general, and agriculture in particular, and availability of critical inputs, such as agricultural raw materials, infrastructure and imported materials.

Level of activity in most of the recent models is disaggregated into agriculture, manufacturing, economic infrastructure, and services. Price determination in agriculture is governed by a flex price (supply-demand balance) mechanism, subject to public intervention through procurement and support prices. Price formation in other sectors is based on markup over costs, including wage rate, administered prices of critical intermediates, and imported inputs. Markup is viewed as being influenced by excess demand/liquidity in the economy, proxied by money stock to overall output. This variable is also factored in the agricultural price determination. Money supply is modeled as the outcome of public desire to hold money, government operations, external factors, and monetary policy. Interest rate and exchange rate are also broadly influenced by a similar set of factors. Turning to foreign trade, imports are determined by domestic activity and relative prices, exports by world economic activity, relative prices, and domestic availabilities.

In most of the models large parts of public activity is given. External economic environment is a datum. Most of the recent models are driven by public investment-quantum and composition, current expenditures and their composition, monetary-fiscal policies and external economic environment, among others. In the short run, weather (rainfall) is a decisive factor.

One disappointing aspect of macroeconometric modeling in India has been that each model has turned out to be a one-time exercise. It is only since the early 1990s that sustained ongoing work on a structural macroeconometric model began jointly at the Institute of Economic Growth and the Delhi School of Economics. That model is the single largest macroeconometric model for India. The system has 347 equations. The model has five production sectors: agriculture, manufacturing, economic infrastructure, services, and public administration and defense. Agriculture is further subdivided into food and nonfood segments. Besides output, capital formation, and price behavior relating to each of the sectors, the model includes separate subsystems, dealing with trade and balance of payments, money and banking, public finance, private consumption and savings. The model is a constituent of the Global LINK Model, being operated under the auspices of the United Nations. Biannual (September and March) forecasts for LINK are made regularly, and several policy simulations have been carried out. The forecasts incorporate the dynamics of the model and are made for more than a year ahead. The India model is constantly undergoing revisions to incorporate policy changes, taking advantage of the latest data. The project is now housed at the Centre for Development Economics at the Delhi School of Economics.

Most of the models were subject to policy simulations. They mostly relate to increase in administered prices, enhanced public investment and change in its composition, pattern of financing public deficits, fiscal—monetary policy stances, exchange rate changes, and world economic scenario—world output, trade volume, and international prices. Normal rainfall assumption is invariably a part of counter factual or "what if" simulations for almost all the models.


The trilogy of models—I-O, CGE, and econometric—are complementary to each other. Common to all the three models is the estimation of final demand. Econometric models do not emphasize intermediate demands, as they deal with a net value-added output concept. Behavioral characteristics that underlie consumption, investment, prices, public sector, money, and trade receive more emphasis in macroeconometric models. Sectoral output interdependence is at the core of I-O and CGE models. The data requirements for each of the models are demanding, particularly in the case of developing countries. Whatever model is used, it is essential to maintain it, with refinements in specification, use of the latest data, and use of improved econometric methodologies to keep it relevant for policy analysis and forecasting.

K. KrishnamurtyJ. Mahender Reddy

See alsoEconomy since the 1991 Economic Reforms


Desai, Meghnad J. "Macro Econometric Models for India: A Survey." In A Survey of Research in Economics. Vol. 7: Econometrics. Indian Council of Social Science Research. Mumbai: Allied Publishers, 1978.

IEG-DSE Research Team. "Policies for Stability and Growth with a Comprehensive Structural Model for India." Journal of Quantitative Economics 15, no. 2 (July 1999). Edited by V. Pandit and K. Krishnamurty; abridged version of the same under the title Policy Modeling for India, edited by V. Pandit and K. Krishnamurty. New Delhi: Oxford University Press, 2004.

Klein, Lawerence R. "Econometric Models, Planning and Developing Countries." In Development Perspectives (Silver Jubilee Lectures, 1984), Institute of Economic Growth. New Delhi: B. R. Publishing Corporation, 1988.

Krishna, K. L., K. Krishnamurty, V. N. Pandit, and P. D. Sharma. "Macro-Econometric Modeling in India: A Selective Review of Recent Research." In Development Papers No. 9: Econometric Modeling and Forecasting in Asia. Bangkok: Economic and Social Commission for Asia and Pacific, United Nations, 1991.

Krishnamurty, K. "Macroeconometric Models for India: Past, Present and Prospects." Economic and Political Weekly 37, no. 42 (October 2000): 19–25.

Krishnamurty, K., and V. Pandit. Macroeconometric Modelling of the Indian Economy: Studies on Inflation and Growth. New Delhi: Hindustan Publishing Corporation, 1985.

Mahalanobis, P. C. The Approach of Operational Research to Planning in India. Mumbai: Asia Publishing House, 1963.

Narasimham, N. V. A. A Short Term Planning Model for India. Amsterdam: North-Holland Publishing Company, 1956.

Pandit, V., and K. Krishnamurty, eds. Policy Modeling for India. New Delhi: Oxford University Press, 2004.

Saluja, M. R. Input-Output Tables for India: Concepts, Construction, and Applications. New Delhi: Wiley Eastern, 1980.

Sarkar, Hiren, and Manoj Panda. "A Short Term Structural Macroeconomic Model for India: Applications to Policy Analysis." In Development Papers No. 9, Econometric Modeling and Forecasting in Asia. Bangkok: Economic and Social Commission for Asia and Pacific, United Nations, 1991.

Taylor, Lance. Structural Macroeconomics: Applicable Models for the Third World. New York: Basic Books, 1983.

Taylor, Lance, ed. Socially Relevant Policy Analysis: Structuralist Computable General Equilibrium Models for the Developing World. Cambridge, Mass.: MIT Press, 1990.

Venkatramaiah, P., A. R. Kulkarni, and Latika Argade. Structural Change in the Indian Economy: An Analysis with Input-Output Tables, 19511963. Pune: Gokahale Institute of Politics and Economics, 1984.