Dynamic Demand Model Research Articles

Is Imperfect Competition in the Product Market Relevant for Labour Markets?

Recent research in macroeconomics emphasises the importance of imperfect competition in the product market for labour market outcomes. We investigate one aspect of this issue by specifying a dynamic labour demand model where firms face different degrees of competition in the product market and test its predictions for 299 US manufacturing industries. We find that the long‐run equilibrium level of industry employment as well as the speed of labour demand adjustment decreases with market power. Our results imply that imperfect competition in the product market explains part of the observed labour market rigidities and also sheds new light on two “stylized facts” in industrial organisation, the observation of procyclical movements in productivity and price–cost margins in concentrated industries.

Full maximum likelihood estimation of dynamic demand models

Abstract The maximum likelihood estimation of dynamic demand models has usually been based on the likelihood function conditional on the first observations of the dependent variables. However, this neglects information which may be necessary for identifying the long-run structure. We formulate the full likelihood of a general dynamic demand model involving arbitrary lag orders, express its analytical derivatives in a relatively simple form, and propose a reparameterization which is always well-defined. The methodology is illustrated with a small empirical application, using the levels version of the CBS model proposed by Barten (1989) and annual British data on four commodities.

Sources of asymmetry in production factor dynamics

Abstract In this paper we investigate possible sources of aggregate cyclical asymmetry of production factor dynamics using a trivariate structural dynamic model of capital and labor demand and output. The two sources are: internal or behavioral asymmetry resulting from asymmetry in costs of adjusting factor inputs, and external non-linearity present in the process of real factor prices, being the model's forcing variables together with productivity shocks. In the empirical analysis behavioral asymmetry and external non-linearity are disentangled by estimation (GMM) and by simulation techniques. Simulated solutions of the model's nonlinear first order necessary conditions are obtained using an extended version of the parameterized expectations algorithm (PEA). Behavioral asymmetry accounts for about 50 percent of the curvature of adjustment costs and therefore contributes in an important way to the dynamics of production factors; external non-linearity on the contrary plays only a moderate role.

Adjustment costs and dynamic factor demand models: A presentation of two approaches applied to the U.S. cigarette manufacturing

This paper makes use of the adjustment cost hypothesis to develop and compare the results of two dynamic input demand models applied to the U.S. cigarette manufacturing. One of the models presented in this paper is the flexible accelerator model and the other is the rational expectations model. In both models, capital and tobacco stocks are treated as potential quasi-fixed inputs and test are performed investigating if these inputs are subject to adjustment costs. The results obtained from the estimation of both models support the hypothesis that tobacco stock (by itself) is a quasi-fixed input but the hypothesis that capital stock (by itself) is a quasi-fixed input is not supported. The results of the rational expectations model support the hypothesis that capital and tobacco stocks together are subject to adjustment costs. Note that this hypothesis cannot be tested by the flexible accelerator model. The short-, intermediate-, and long-run elasticities of input substitution can be calculated from the statistical results. The own price elasticities produced by both models have the correct signs except for the long-run own price elasticity for materials generated by the rational expectations model which is positive. Both models generate similar in sign and magnitude short- and intermediate-run elasticities. Domestic and imported tobaccos appear to be short- and intermediate-run complements, in both models. Most of the long-run elasticities generated by both models differ in sign and magnitude. It is important to note that domestic and imported tobaccos appear to be long-run substitutes according to the flexible accelerator model but long-run complements based on the rational expectations model. A possible explanation of the production process provided by both models is that during the short- and intermediate-run, materials can substitute for domestic and imported tobaccos. In addition based only on the rational expectations model, capital stock may substitute for domestic tobacco in the intermediate run. In the long-run tobacco stock can substitute for domestic tobacco and it can be used together with imported tobacco in the production of cigarettes. The hypothesis that there is not any difference in marginal costs of producing cigarettes for domestic and foreign markets can be rejected in both models. Moreover, the hypothesis that the cigarette industry behaves competitively in the domestic and export markets cannot be rejected. Finally, there are statistical evidence supported by both models (and especially by the rational expectations model) that health information about the dangers of cigarette consumption and government advertising restrictions have negative effects on output demands. In addition both models agree on an inelastic domestic demand for cigarettes and on an elastic export demand.

A Comparison of Methods for Dynamic Origin-Destination Matrix Estimation

An overview of existing dynamic OD-estimation methods is given, and a statistical technique to estimate parameters in dynamic travel demand models based on Bayesian inference is proposed. Using both synthesized and empirical data the new technique is compared to traditional ones such as the Kalman filter and least squares. This shows the new technique results in more accurate estimates of OD-split proportions. Surprisingly however, this does not result by default in more accurate predictions of link flow volumes.

A Comparison of a Production Smoothing Model and a Dynamic Factor Demand Model with Inventories: Applications to French Industrial Sectors

In literature two types of models exist that aim at describing the production decision(s) of entrepreneurs, taking account of production costs and costs incurred by the existence of inventory stocks of final goods. One type is called the production smoothing models and the other type the factor demand models that include inventories. In this paper both types are discussed, compared with each other and estimated. The main results are that the factor demand model is preferred to the production smoothing model since (i) costs are more "structurally" specified by which more efficient parameter estimates are obtained and (ii) arbitrary normalisation rules are not needed. GMM estimation results obtained with data from French industrial sectors also corroborate the preference of the factor demand model.

Some questions concerning dynamic almost ideal demand systems

A paper by Xepanadeas and Habib, that estimates dynamic demand models, is shown to contain some serious errors. Careful consideration of the adding-up condition is shown to be necessary when using demand systems with autoregressive disturbances. Misspecification testing should be used in all empirical work. These points are illustrated using a dynamic almost ideal demand system applied to Swedish dairy data.

The impact of BSE on the demand for beef and other meats in Great Britain

Indices of media coverage of BSE are incorporated in a dynamic AIDS model of meat demand. The publicity which BSE has received is found to have had significant effects on the allocation of consumer expenditure among the meats. A short-run impact has been identified which accounts in large part for the discernible drop in the market share of beef in the early 1990s. And more important, there also appears to be a significant long-run impact of BSE, which by the end of 1993 has reduced the beef market share by 4.5%.

Endogenous capital utilization and productivity measurement in dynamic factor demand models Theory and an application to the U.S. electrical machinery industry

The literature on dynamic factor demand models has, until recently, largely overlooked the issue of capital utilization. In this paper we allow for variations in the rate of capital utilization within the context of a dynamic factor demand model by adopting a modeling framework within which the firm combines its beginning-of-period stocks with other inputs to produce its outputs as well as its end-of-period stocks. We also define measures of productivity and capacity utilization for the adopted framework. As a by-product, the framework also provides for a consistent decomposition of gross investment into replacement and expansion investment. As an illustration, the model is applied to U.S. electrical machinery data.

A Dynamic Equilibrium Model of Housing Demand and Mobility with Transactions Costs

This study addresses intertemporal housing demand in the presence of transactions costs. The consumer solves a problem similar to a government′s calculation of the socially optimal level of a public good. Instead of the government′s optimizing across individuals, the consumer optimizes over time periods, subject to the constraint that housing consumption remains constant. The resulting immobility cost provides a latent variable that can be modeled and compared to moving costs in the move-stay decision. Housing demand, number of moves, and lengths of stay all emerge endogenously. Comparative statics results are derived, and simulation analysis shows the dynamic properties.

A Dynamic Demand Model for Liquor: The Case for Pooling

This paper estimates a dynamic demand model for liquor in the United States using panel data from 43 states. Because of taste changes over time and between states in liquor consumption, purely time series or cross sectional studies do not elicit reliable price elasticity estimates. This study makes the case for pooling and shows how one can control for individual state effects and endogeneity of the regressors using estimators suited for a dynamic demand model. Our results indicate that the long-run price elasticity is in the -0.7 range. The findings also support strong habit persistence, a small positive income elasticity, and very weak evidence of bootlegging from adjoining states. The magnitude of the long-run price effect suggests that sin taxes can serve not only as an important income source but also as a significant deterrent effect. Copyright 1995 by MIT Press.

A dynamic model of demand for Mediterranean tourism

A dynamic econometric model was specified in order to estimate tourism consumption changes by Northern European countries and the USA for major Mediterranean destinations. The model employed a flexible framework for modelling short-term dynamics as well as the long run effects of a range of variables of specific interest to the countries considered. The estimated model provided useful information for tourism demand. The income elasticities demonstrated considerable differences in tourism demand preferences between origin countries and between traditional and newly developing destinations. The own and substitute price elasticities indicated the importance of effective prices in determining tourism receipts of the destinations.

A Dynamic Model of Industrial Energy Demand in Kenya

This paper analyses the effects of input price movements, technology changes, capacity utilization and dynamic mechanisms on energy demand structures in the Kenyan industry. This is done with the help of a variant of the second generation dynamic factor demand (econometric) model. This interrelated disequilibrium dynamic input demand econometric model is based on a long-term costfunction representing productionfunction possibilities and takes into account the asymmetry between variable inputs (electricity, other-fuels and labour) and quasi-fixed input (capital) by imposing restrictions on the adjustment process. Variations in capacity utilization and slow substitution process invoked by the relative input price movement justifies the nature of input demand disequilibrium. The model is estimated on two IS1C digit Kenyan industry time series data (1961 - 1988) using the Iterative Zellner generalized least square method.

TRADE CREATION AND DIVERSION EFFECTS OF THE U.S.‐CANADIAN FREE TRADE AGREEMENT

This study empirically estimates and evaluates the economic benefits of the U.S. and Canadian Free Trade Agreement (FTA). Most past studies rely on aggregate data. The analysis here emphasizes the trade effects of removing tariff and nontariff barriers on each commodity group classified by the Standard International Trade Classification. Estimating the amount of trade expansion under FTA for both countries involves using the import demand elasticities from a dynamic demand model. Results show that U.S. imports from Canada are more sensitive to domestic, import, and world prices than are Canadian imports from the United States. U.S. imports from Canada would increase roughly £3.257 billion compared to the £2.432 billion increase for Canadian imports from the United States.

An adjustment cost model of long‐term employment in Japan

AbstractA dynamic factor demand model is presented which pays special attention to the prevalence of a long‐term employment relationship in Japan. The model is based on the representation of technology by a variable cost function with adjustment costs for employment and capital stock, where the variable cost consists of the sum of overtime costs and materials costs. With employment being quasi‐fixed and scheduled hours institutionally regulated, short‐run adjustments are mostly made by overtime hours. Application to a time‐series data on the Japanese electrical machinery industry indicates quasi‐fixity of capital and employment and reproduces short‐run overshooting of overtime hours to compensate for the sluggish adjustment of employment.

Speculative behavior in an experimental cash/asset portfolio

We develop a simple cash and asset dynamic model of the speculative demand for cash, when the price of the asset (a bond) is a random variable. The purpose of the model is to develop an experimentally testable framework within which the individual's speculative demand for money can be tested in isolation of the many confounding factors in the real world context. The model predicts an optimum reservation price rule will govern speculative behavior and implies the existence of a Keynesian liquidity trap. Higher dividend yields on assets will induce agents to hold assets rather than speculate on them; asset price risk will reduce speculative behavior, and transaction and consumption costs will leat to a “sticky price region” on an asset within which no speculative trading will take place. Finally we summarize the results of an experiment which appears to verify some of the conclusions of our model.

R&D spillovers and technology transfer among and within vertical keiretsu groups: Evidence from the Japanese electrical machinery industry

Using a dynamic factor demand model with R&D externalities, we analyze the effects of technological diffusion among and within vertical keiretsu groups in the Japanese electrical machinery industry. We find that technology transfer from a core firm to its subcontracting firms is substantial. We also find positive R&D spillovers, which stem from the R&D activities of other keiretsu groups. Particularly remarkable are the spillovers between core firms of competing keiretsu groups.

Cigarette taxation: Raising revenues and reducing consumption

This paper estimates a dynamic demand model for cigarettes based on panel data from 46 American states over the period 1963 to 1988. The first objective is to show some of the pitfalls of studies that rely on a time series regression of a specific state, or a cross-section regression for a given year. The second objective is to update the results of Baltagi and Levin (1986) from 1980 to 1988 and to study the sensitivity of these results to various ways of modelling the bootlegging effect as well as controlling for fixed or random state effects. This study finds a significant habit persistence effect, a small but significant ‘border purchasing’ effect, a significant but inelastic own price effect and a small but significant income effect.

Dynamic factor demand equations in U.S. and Canadian agriculture

This research provides one of the first empirical estimates of a data-based dynamic factor demand model for American and Canadian agriculture. Models such as these deserve more widespread use in the empirical analysis of agriculture. These models have the advantage that they do not impose inappropriate dynamics on the data. Rather they permit the data to select the appropriate dynamics. We use a model originally developed by Anderson and Blundell. This model is a general first-order dynamic model which contains as testable hypothesis several simpler models. This model permits us to estimate the long-run agricultural production structure as a subset of the dynamic parameter estimates. We will test this long-run structure for symmetry, homotheticity and neutral technical change. The estimated models may be used to test for three alternative dynamic structures. In the limit, dynamics may not be needed and we can test for the static long-run equilibrium model. Two intermediate cases are the autoregressive and the partial adjustment models which are simpler than the general model but still include dynamics. Our results suggest that the long-run equilibrium model is unsatisfactory in both countries. A dynamic model is needed. In both countries, the two more restricted dynamic models are rejected. The general dynamic model is required. In Canada, the long-run equilibrium structure is homothetic with neutral technical change. In the United States, homotheticity is also accepted but neutral technical change is rejected.

Dynamic factor demand equations in U.S. and Canadian agriculture

AbstractThis research provides one of the first empirical estimates of a data‐based dynamic factor demand model for American and Canadian agriculture. Models such as these deserve more widespread use in the empirical analysis of agriculture. These models have the advantage that they do not impose inappropriate dynamics on the data. Rather they permit the data to select the appropriate dynamics.We use a model originally developed by Anderson and Blundell. This model is a general first‐order dynamic model which contains as testable hypothesis several simpler models. This model permits us to estimate the long‐run agricultural production structure as a subset of the dynamic parameter estimates. We will test this long‐run structure for symmetry, homotheticity and neutral technical change.The estimated models may be used to test for three alternative dynamic structures. In the limit, dynamics may not be needed and we can test for the static long‐run equilibrium model. Two intermediate cases are the autoregressive and the partial adjustment models which are simpler than the general model but still include dynamics.Our results suggest that the long‐run equilibrium model is unsatisfactory in both countries. A dynamic model is needed. In both countries, the two more restricted dynamic models are rejected. The general dynamic model is required. In Canada, the long‐run equilibrium structure is homothetic with neutral technical change. In the United States, homotheticity is also accepted but neutral technical change is rejected.