پویایی دستمزد - قیمت و توزیع درآمد در یک مدل نیمه ساختاری کینز-گودوین

The paper studies the role of income distribution within a medium-scale macrodynamic model built in a Keynesian and Goodwinian tradition. Combining a wage and price Phillips curve, adjustments of an inflation climate, an IS relationship determining output, Okun’s law for employment and the Taylor rule for monetary policy, a semi-structural model is obtained that incorporates the most important macroeconomic channels in a closed economy. After assessing the reasonable time series variabilities in stochastic simulations, a deeper analysis is concerned with the stabilizing and destabilizing effects of the model’s parameters, and with a structural shift in income distribution. In many details of these investigations, the distinction between a wage-led and profit-led regime becomes important.

While of secondary, if at all, importance in the current mainstream theoretical literature, the role of income distribution was a key issue in the work by Richard Goodwin and is still an important issue in times when advocates of higher downward wage flexibility (and sometimes also price flexibility) have gained ground, and when in many industrial countries the distribution of income has become increasingly unequal. The model to be discussed in this paper follows Goodwin by taking the wage and price dynamics and the resulting variations in income distribution into account. In other respects, it is related to current work on modern macroeconomic modelling in a, broadly speaking, Keynesian (but not New-Keynesian) tradition. It particularly draws on the so-called KMGT disequilibrium approach advanced in Chiarella et al. (2005), which refers to Keynes, Metzler, Goodwin and Taylor as its patron saints. ‘Goodwin’ indicates that income distribution plays a crucial role in the dynamics of nominal and real variables. It is determined by the interplay of a wage as well as a price Phillips curve, and in turn impacts positively on aggregate demand via workers’ consumption and negatively via profit-oriented investment. The Metzlerian part is a consequence of goods market disequilibrium which is absorbed by inventories, whose evolution in turn feeds back on planned inventory investment and thus aggregate supply. ‘Taylor’ takes account of monetary policy and complies with the general consensus reached over the last decade that the central bank adopts an interest rate rule, most often specified as a variant of the Taylor rule. In order to simplify and to improve the scope for reliable econometric parameter estimates, the present model drops the Metzlerian feedbacks and the explicit distinction between workers’ and rentiers’ income. Besides a persistence effect, output is thus supposed to be directly dependent on the real interest rate and the wage share. For the changes of employment, a slightly augmented version of Okun’s law is employed. While these parts of the model are more akin to reduced-form relationships, the other equations have a more satisfactory structural underpinning. They comprise: (1) a wage and price Phillips curve, both enriched by the influence of an inflation climate and the wage share; (2) dynamic adjustments of the inflation climate; (3) the endogenous determination of labour productivity growth; (4) the Taylor rule for the nominal interest rate. The paper consists of two main blocks: Section 2 introduces the single components of the model, and Section 3 considers them in their entirety by means of numerical simulations. The first part of Section 3 discusses the benchmark scenario and the dynamic properties to which it gives in the stochastic setting. The second part studies the impact of ceteris paribus parameter changes on the variabilities of the main state variables. The third part examines the deterministic adjustment paths in the wake of a stylized structural shift that amounts to a deterioration of the position of workers in the wage bargaining process. In the latter two parts, we emphasize as a special feature the distinction between a so-called wage-led and profit-led regime in the determination of output. Section 4 concludes.1 It should be mentioned that different from most heterodox Keynesian modelling, the first two parts of Section 3 employ the Frisch paradigm, which views the economy as basically shock-driven. Nevertheless, the paper demonstrates that also in this framework, despite the assumed (deterministic) stability of the steady state, meaningful questions for stabilizing and destabilizing parameter effects can be asked and answered. Hence, the ‘un-Keynesian’ paradigm does not necessarily preclude the investigation of such Keynesian themes as the impact of greater wage flexibility on economic ‘stability’.

The paper has started out with the specification of a theoretical framework in the tradition of Keynes and Goodwin, where the adjustments of wages and prices are sluggish in an (augmented) Phillips curve manner, aggregate demand drives output and income distribution, in form of the wage share, feeds back on the economy through a variety of channels. Some of the building blocks had a richer structure than other, so we arrived at what can be called a semi-structural Keynes–Goodwin model. Borrowing parameter values from previous work and following the Frisch paradigm as the ruling paradigm in macroeconomics, we simulated this numerical economy (with its deterministically stable equilibrium) in a stochastic environment. Our criterion for evaluating the performance of the model were the variabilities, i.e., the standard deviations, of the model-generated time series. It turned out that (except for the employment rate) these summary statistics were quite in line with their empirical counterparts. We may claim that these results set a standard for the dynamic behaviour of macrodynamic systems that goes beyond the mere goodness-of-fit of the econometric estimations. It was therefore justified to study some of the model’s properties in greater depth. Thus, in terms of the variabilities of the main variables, we could characterize the reaction coefficients as stabilizing or destabilizing. In particular, a higher responsiveness of wages to fluctuations in employment turned out to be unambiguously destabilizing. For the effects of the other parameters, it was in several cases important to know whether the economy is in a profit-led regime (as estimated) or in a wage-led regime. Even if its semi-structural nature is accepted, the model has one weak point which in any case urges for improvement. This is the specification of Okun’s law that basically links the change of the employment rate e to the changes of utilization y. Besides other shortcomings (discussed in Franke et al., 2006), the assumption implies that the variability of e is much too low in relation to that of y. Another feature that should be mentioned in this context is our problematic identification of the number of jobs (in the concept of the employment rate) and the number of hours worked (in the concept of labour productivity). Instead of a slight modification in an equation for the composed variable “labour employed/labour supplied”, one can think of two separate adjustment equations for jobs and hours. In this way, a further utilization variable generally varying over the business cycle would enter the stage: actual hours per worker employed (i.e., an insider employment rate). Estimations of these adjustments have been recently carried out by Franke (2006, Section 6). Regarding the supply side it was there additionally found that over the last 15 or 20 years the labour force responded much more directly and vigorously to the variations of economic activity. It should not be very difficult to augment the present models with these salient features.