Diagnostic Testing in Applied Econometrics*

Economic policy formation and the development of economic theory both rely ultimately upon empirical analysis for their direction and sustenance. Recent contributions from econometric theory suggest that the quality of the advice obtained from applied econometric research can be considerably improved by more systematic testing of empirical models. These new tests help identify both strengths and weaknesses of the models. Though fundamental economic theory and good knowledge of data are the primary ingredients of applied economic research, these new diagnostic tests are an important adjunct to existing methodology. This survey introduces the non-specialist econometrician to diagnostic model testing, and thoughout emphasizes the heuristic rather than the mathematical underpinnings of the testing strategy.     

The implicit objective function of Italian macroeconomic policy: 1963:1–1980:4

Econometric estimates of policy makers' preferences can be an important source of information for ex post macroeconomic policy analysis. This paper first discusses the econometric methodology to obtain consistent estimates of the policy maker's objective function. Properties of the estimation process are derived by appropriately solving an inverse control problem. Then, Italian macroeconomic data are used to estimate the policy maker's preferences under different specifications of his objective function. The major finding is that the balance of payments equilibrium was the main target of Italian macroeconomic policy.     

A modified procedure for simulation of macroeconometric models

The rational expectations literature suggests that even the effects of unanticipated changes in policy variables may not be accurately described by conventionalsimulations of macroeconometric models. Beginning with Zellner and Palm's time-series approach to structural econometric modelling, a modified simulation procedure is developed to examine the effects of changes in policy variables which are consistent with historic time-series processes. A state-space representation and the Akaike Information Criterion are used to model exogenous variable processes, and results from modified and conventional simulations are compared.     

Protection and commercial policy: Experiments with a general equilibrium econometric model

This paper analyzes the incidence of commercial policy by simulating the effects of an import tariff in the context of a general equilibrium econometric model of the U.S. It is found that in the long run the tariff protection given to importables is mainly at the expense of exportables, rather than the nontraded goods sector. The simulations also provide the general equilibrium responses of a number of other variables to changes in commercial policy, which may be of interest from a policy perspective.     

Consumer demand for meat and the evaluation of agricultural policy

This study utilizes an econometric model of equilibrium in the U.S. livestock and feedgrain markets to investigate a number of questions of economic methodology and policy. Both nonlinear consumer demand equations that obey the constraints of neoclassical demand theory and a model of supply are used. This approach allows for the measurement of the effects on consumer welfare of actual government policies. In particular, the model is employed to analyze the welfare effects of an actual policy situation — the sale of U.S. grain to the Soviet Union in the third quarter of 1972.     

The industrial implications of an optimal reflationary mix for the British economy

This paper derives an optimal reflationary mix for the next decade and comments on the industrial implications of this policy. The discussion is based on simulations of the multisectoral dynamic model of the UK economy developed by members of the Cambridge Growth Project. The results suggest that the concentration on macroeconomic policy alone may be misguided and that instruments which attack the roots of Britain's industrial problems should receive more attention. The paper also demonstrates that the optimal control procedure is a reliable tool for steering an economy towards a desired target trajectory and that this is feasible for very large econometric models.     

CONTAGIOUS POLICIES: AN ANALYSIS OF SPATIAL INTERACTIONS AMONG COUNTRIES' CAPITAL ACCOUNT POLICIES

Countries' capital account policies might be contagious in the sense that domestic policies are driven by other countries' policies. A model of strategic interactions is developed to show that countries' best response to policy changes elsewhere consists in imitating this policy. Using a spatial econometric model, the hypothesis of policy interactions is tested in a large panel data set. The evidence shows that capital account policies are contemporaneously correlated across countries. Concerning fundamentals, the move to a fixed exchange rate regime and an increase in real world interest rates are correlated with the imposition of capital account restrictions.     

Modelling policy options in the Indonesian agricultural sector

This paper persents a simple model framework for quantitative multi-market agricultural sector policy analysis. Although the example refers to the Indonesian case, the model structure is also applicable for other country studies. Since either prices or trade requirements can be set exogenously, the model is particularly useful for policy analysis in countries using trade controls to pursue food self-sufficiency targets. Data requirements are limited to supply-and-demand elasticities, some technical coefficients on the relationship between food and non-food prices, non-price induced trends in single commodity markets and in the non-food sector as an aggregate and a base-year data set. The model was developed to be relatively transparent to policy-makers and to allow for an easy revision of the magnitude of policy parameters or trend estimates if so needed. For the case study, econometric estimates, which are consistent with the requirements of neoclassical theory, have been used to parameterize the model. Model simulations have been run for the price run-up period of 1986 to 1988 and also for the Fifth Plan period 1988 to 1993. The results show that the model is a useful instrument to investigate in multi-market and multi-sector interrelationships of a country focussing on agriculture. For the Indonesian setting moel results suggest that during the Fifth Plan period rice self-sufficiency targets are easily met without the need of a particular investment or subsidy scheme. However, rice self-sufficiency and on-target performance of other important agricultural commodity markets in Indonesia will not come along automatically with the creation of jobs up to an amount needed to fight increasing rural underemployment.     

THE TIME‐SERIES PROPERTIES OF UK INFLATION: EVIDENCE FROM AGGREGATE AND DISAGGREGATE DATA

This paper contrasts the time‐series properties of aggregate and disaggregate UK inflation. While aggregate inflation is found to be non‐stationary, unit root rejection frequencies are increasing when we use more disaggregate data. Structural break analysis suggests that structural shifts in monetary policy could alter inflation persistence. Additionally, panel evidence indicates that the unit root hypothesis can be rejected for sectoral inflation rates. Finally, we compare the persistence properties of UK inflation, finding statistically significant differences between aggregate and disaggregate series. Our analysis suggests that aggregation matters, which has important implications for econometric analysis and the conduct of monetary policy.     

Time series explanations of merger activity: some econometric results

Existing papers which have attempted to test explanations of time series patterns in merger behaviour suffer from the defects that, first, they usually consider only one hypothesis and, secondly, none use a moderm econometric methodology. Consequently, their results may be subject to the spurious correlation problem. In this paper we argue that four well-known approaches to explaining time series data in acquisitions (Gort's disturbance theory, King's Trapped Equity model, disequilibrium hypotheses and ad hoc approaches) can all be nested within a capital budgeting decision-making framework. Using a co-integration methodology, a long-run relationship was found between the quarterly number of acquisitions and the growth rate of GDP, Tobin's Q, the balance of payments, the unemployment rate and share prices. Some of the previously proposed theories were found to explain the short-run dynamic variation in acquisition activity. No support was found for Gort's disturbance theory or for King's Trapped Equity model. Conclusions for antitrust policy are suggested.     

An econometric annual model of the real sector in Israel

This paper presents a small annual econometric model of the real sector in Israel. The model was built in the Israeli Economic Planning Authority and serves for both forecasting key real macrovariables and policy analysis. It is a first step towards the construction of a comprehensive econometric model of the Israeli economy. Unlike previous models, it is based on recent data.     

Targets and lags in a two-equation model of US stabilization

A simple model of activist macroeconomic policy derives a reaction function by assuming that rational governments have performance objectives, but are constrained by the Phillips curve. Although not formally modeled, governments apply a variety of instruments to influence inflation and output, in addition to monetary policy these include fiscal policy, bailouts and exchange rates. Our econometric results are generally consistent with US economic history. One qualification is that governments appear more likely to target growth rates than output gaps. Another inference is that inflation expectations are more likely to be backward than forward looking; a variety of rational expectation models fit the data less well than do simple inertial expectations. We also find that annual data series are more appropriate than quarterly ones for studying these issues.     

Carbon tax simulations using a household demand model

The main objective of this paper is to analyse consumer response and welfare effects due to changes in energy or environmental policy. To achieve this objective we formulate and estimate an econometric model for non-durable consumer demand in Sweden that utilises micro- and macro-data. In the simulations, we consider two revenue neutral scenarios that both imply a doubling of the CO₂ tax; one that returns the revenues in the form of a lower VAT and one that subsidise public transport. One conclusion from the simulations is that the CO₂ tax has regional distribution effects, in the sense that household living in sparsely populated areas carry a larger share of the tax burden.     

Monetary policy through the ‘credit-cost channel’: Italy and Germany pre- and post-EMU

We present an empirical analysis of the ‘Credit-Cost Channel’ (CCC) of monetary policy transmission. This channel combines bank credit supply and interest rates on loans as a cost to firms. The thrust of the CCC is that it makes both aggregate demand and aggregate supply dependent on monetary policy. As a consequence (1) credit market conditions (e.g. risk spreads) are important sources and indicators of macroeconomic shocks, (2) the real effects of monetary policy are larger and persistent. We have applied the Cointegrated Vector Autoregression (CVAR) econometric methodology to Italy and Germany in the ‘hard’ EMS period and in the European Monetary Union (EMU) period. The short-run and long-run effects of the CCC are detectable for both countries in both periods. Simulation of the estimated model also confirms that inflation-targeting by way of inter-bank rate control stabilizes inflation through structural shifts of the stochastic equilibrium paths of both inflation and the output.     

中国经济周期的非对称性和相关性研究

本文利用时间序列模型等计量方法 ,对一些主要宏观经济变量序列进行随机分解和相关性分析 ,对经济周期的非对称性和长尾性质进行大量实证检验 ,对一些主要宏观经济变量序列之间的扰动和关联进行了计量分析 ,从中不仅识别了经济波动当中的各种非对称类型 ,而且分析了产生非对称的原因。本文分析得到的一些重要检验结果 ,可以作为描述中国经济波动的重要典型化事实 ,可以用于进一步分析和判断中国经济的运行趋势 ,检验和校正相应的经济理论。本文认为中国经济周期的非对称 ,主要是由固定资产投资、财政政策和货币政策的非对称性造成的 ,而价格水平和总需求等因素却保持了比较明显的稳定性。通过经济周期的非对称分析和经济变量周期成分之间的关联性分析 ,我们也具体地分析了宏观经济政策的有效性、方向性和时滞性     

New empirical evidence from assessing financial market integration,with application to Saudi Arabia

We examine whether data frequency, day of the week and econometric methodology matter in analyzing financial market integration. As case study, we investigate equity market comovements between Saudi Arabia and a set of international economies. Our findings take the literature forward and indicate that cross-market linkages are weak and subsample-dependent regardless of whether data are daily, weekly (whatever the weekday) or monthly and whatever the econometric approach. The results are relevant for investors who want to be more informed of promising investment opportunities, and for financial makers to take necessary policies to hedge against the effects of shocks.