Chapter 12 From the 3-equation of the new consensus to the post-Keynesian model
We can now make step-by-step changes to the NCM 3-equation model that we have studied in the previous chapter. Our aim is to show that relative small changes to the assumptions of the 3-equation model lead to substantial changes in the policy implications. The full combination of the changed set of assumptions will generate a simple and alternative macroeconomic model in line with post-Keynesian theory allowing for a different set of policy suggestions.
The main features of post-Keynesian macroeconomics can be summarised in the following list:44
- focus on a monetary theory of production according to which money is not neutral neither in the short nor in the long-run
- relevance of the principle of effective demand both in the short-run and in the long-run
- importance of the concept of fundamental uncertainty
- economic processes occur in historical, irreversible time and are largely path-dependent
- the distributional conflict plays a major role in determining economic outcomes
It follows that in post-Keynesian theory, the long-run equilibrium, i.e. the NAIRU and the corresponding inflation-stable employment level, is not independent from the demand side of the economy. In the following sections, we will look first at the various channels through which the NAIRU can be influenced by demand side factors and later, at how monetary, fiscal and wage/income policy, by affecting aggregate demand, can in turn influence the NAIRU.
12.1 Endogeneity of the NAIRU 1: labour market hysteresis
A first channel that makes the NAIRU dependent on actual unemployment and on macroeconomic policy is represented by the so-called hysteresis effect in the labour market. The hysteresis effect is a significant slowdown in the adjustment of the unemployment rate towards the NAIRU. High unemployment rates experienced during the transition phase towards the NAIRU after an initial demand shock may discourage parts of the workforce from actively looking for employment. These peculiar circumstances might have the effect that part of the workforce that has been long-term unemployed may end up exiting the labour market. This will improve the bargaining position of those who remain employed with the ultimate effect of increasing the NAIRU.
The idea of labour market hysteresis has been discussed by new Keynesian authors in the past (Blanchard and Summers 1987, 1988; Ball 1999) and has been the subject of some more recent studies, Ball (2014) and Stockhammer and Sturn (2011). These authors argued that high unemployment rates end up having an increasing effect on the NAIRU and that unemployment determined by effective demand has an impact on the NAIRU and on the inflation-stable employment level. This bring us to the conclusion that monetary policy aimed at bringing the economy back to potential after a shock might have an effect on the long-run equilibrium of the model contradicting the assumption that sees monetary policy being neutral in the long run. Post-Keynesian authors argue that the opposite would also be true. If unemployment remains below the NAIRU for a long time, the NAIRU may end up decreasing (see, e.g. Hein and Stockhammer 2009). In this case, it can be assumed that previously discouraged workers would now be encouraged to actively participate to the job search with the effect of increasing competition in the labour market and decreasing the wage demands of the workers already employed. As the workforce increases, the NAIRU decreases. In the post-Keynesian model framework, the effect of demand on the NAIRU applies to both negative and positive movements in the business cycle.
A simple form of labour market hysteresis can be incorporated into the NCM 3-equation model that we have been studying in the previous chapters. Figure 12.1 shows how the \(WS\) curve rotates counterclockwise due to an increase in the conflict orientation of employees, \(k\). Here it is assumed that the increase in \(k\) was caused by the unemployment rate being below the original NAIRU for some time after a demand shock has hit the economy. The counterclockwise rotation of the \(WS\) curve leads to a new distributional equilibrium at the left of the old equilibrium value corresponding to a new and now higher value of the NAIRU.
Figure 12.1: Rotation of the WS curve thanks to improved bargaining position of the still employed.
We illustrate this in the “hysteresis” scenario accessible via the link below. Let’s make the case of a positive shock to aggregate demand. At the beginning of the simulation, the economy is hit by the demand shock. The central bank will react by temporarily raising the unemployment rate above the NAIRU in order to reverse inflation expectations, just as we have learned with the previous scenarios. However, if the unemployment rate is higher than the NAIRU for more than one period, the hysteresis effect will be triggered and part of the unemployed will withdraw from the labour market. This improves the bargaining position of the labour force (remember that \(k\) increases) reducing output, the associated inflation-stable employment (\(L^N\) shifts to the left) and increasing the NAIRU. The central bank observes the change in the NAIRU and in the inflation-stable employment level and start to target the new values adjusting the optimal path to the inflation target. This new path is indicated by a shift of the \(MR\) curve. This case and the case of labour market hysteresis after a positive demand shock can be tested with the following interactive scenario.
12.2 Endogeneity of the NAIRU 2: interest rate elastic mark-up
The second endogeneity channel of the NAIRU that we want to investigate is the effect of a change in the interest rate on firms’ production costs and real wage target (see Hein (2006; 2023, chap. 5), Hein and Stockhammer (2009, 2011)). Here, interest rate changes arise as a result of the central bank’s response to demand or supply shocks.
In our price equation (9.8), the mark-up of firms must cover not only profits per unit but also all general production costs per unit that do not result from wage payments. Since interest costs are part of the production costs, in this case costs for borrowing capital, it seems plausible to assume that interest costs have an impact on the mark-up. Under this assumption, a change in the interest rate will have an impact on firms’ pricing, if not immediately, at least in the medium-run. If an increase in the interest rate is perceived by firms to be permanent, we then might observe an increase in the mark-up. Under this assumption, the mark-up would become a function of the nominal interest rate set by the central bank:
\[\begin{equation} m = m(i) \tag{12.1} \end{equation}\]
It is plausible to assume that companies will adjust the mark-up to an increase in the interest rate not immediately but after a certain time, as they first may want to see whether the increase in the interest rate is permanent or temporary. The reason for the possible delayed reaction of the firms may be that they do not want to scare off their customers by adjusting prices too quickly. Another reason could be that the increase in interest costs is perceived by companies only after a certain amount of time.
From a modeling perspective, the interest rate responsiveness of the mark-up means that the \(PS\) curve will shift downward after some periods as a consequence of a positive demand shock and the following reaction of the central bank, as figure 12.2 shows. This causes the inflation-stable employment level to fall, the NAIRU to rise, and wage-price-wage spirals to start when current employment is above the new inflation-stable employment. Because of the higher mark-up and the lower target real wage rate of firms which ultimately determines the realised real wage rate, the central bank’s interest rate policy has an indirect influence on the real wage rate and thus on the distribution of income between wages and profits, which consist of retained profits and distributed interest payments. In addition, the central bank interest rate policy has influenced the inflation-stable employment level and the corresponding NAIRU in the long-run.
Figure 12.2: Interest rate increases shift the PS curve.
In a simplified modeling implementation, in the interactive “interest-elastic mark-up” scenario we assume that the mark-up \(m\) becomes a function of the moving average value of the interest rate:
\[\begin{equation} m = m(\bar{i}) \tag{12.2} \end{equation}\]
The underlying assumption at play here is that as soon as companies perceive the change in the interest rate to be permanent, they will try to defend their (net) profit margins by adjusting the price mark-up. If we assume such behaviour, a problem arises for the central bank: its interest rate policy, which is actually aimed at short-term demand management, now has long-term non-intended effects on price setting and thus on the inflation rate and the NAIRU itself. In the case of excessively high inflation, tight monetary policy can now contribute to a higher NAIRU in the long-run. Monetary policy is much less efficient compared to the NCM version of the 3-equations model due to non-intended and long-run real economic effects. In the “interest-elastic mark-up” scenario accessible through the link below, the optimal adjustment process in response to a positive demand shock is disrupted by firms adjusting their mark-up to reflect increased interest costs. The change in the price mark-up leads to a reduction in inflation-stabilising employment and an to an increase in the NAIRU, affecting directly the price level and leading to a second spike in inflation. The central bank must then raise the interest rate again. By taking the cost effect of interest rate changes into account, the central bank must pursue a more restrictive interest rate policy which has macroeconomic effects beyond the short-run. Stable inflation is possible only with a higher long-run unemployment rate.
In summary, this scenario shows that once the interest rate sensitivity of the mark-up is included in the model, a restrictive monetary policy has the side effect of lowering the inflation-stable employment level and raising the NAIRU. Compared to the standard version of the NCM model, monetary policy is no longer neutral in the long-run. Relying solely on interest rate policy to fine-tune the economy leads to long-run employment losses and, if the impact on the mark-up is strong, may not even lead to a stable long-run equilibrium.45
12.3 The NAIRU as a corridor: a horizontal element in the Phillips curve
For the linear wage-setting curve discussed so far and the associated linear short-term Phillips curve, we have assumed that the rising (or falling) inflation rates caused in the distributional disequilibrium do not influence the target real wage rate of employees. This assumption is plausible if employees assume that their individual or company-specific nominal wage increases have no influence on the overall economic inflation rate. This assumption is plausible if employees assume that their individual or firm-specific nominal wage increases do not affect the general inflation rate at the macro level. This will be particularly the case in highly decentralised and uncoordinated wage bargaining, where wage bargaining takes place individually or, at best, at the firm level. However, when wage negotiations and wage agreements are conducted at industry level and, if necessary, coordinated between industries (e.g. through the wage leadership of an industry that is particularly well organised in terms of trade unions, such as the metal industry with IG Metall in Germany), the trade unions can take into account the effects of the nominal wage increases they have achieved on the overall inflation rate. If they now want to avoid rising inflation rates, for example not to lose price competitiveness against their foreign competitors or to prevent interest rate hikes by the central bank, they must assume joint responsibility for inflation. This means that wage negotiators must adjust nominal wage growth - at constant labour productivity - to the target inflation rate. This is tantamount to adjusting their target real wage rate to the level made possible by firms’ price setting, i.e. to the target real wage rate of firms.
Within a certain “normal” employment range, the real wage targets of employees coincide with the real wage targets of firms derived from price setting, as can be seen in figure 12.3. Within the horizontal element of the \(WS\) curve, there is neither inflationary nor disinflationary pressure (see Hein 2006; 2023, chap. 6). Macroeconomic coordination of wage increases will work only within this “normal corridor” of employment levels and unemployment rates assuming that the necessary conditions exist (strong trade unions, strong business associations, negotiations at the industry or macroeconomic level with effective coordination).
Figure 12.3: Horizontal element in the wage-setting and Phillips curves.
If the employment level rises above the normal corridor and the unemployment rate falls below the corresponding value, firms will start to compete for scarce labour, nominal wage growth will rise above the target inflation rate and the process of accelerating inflation will begin, as shown in figure 12.4. In the opposite situation with employment above the normal corridor, employees are willing to make nominal wage concessions in order to preserve their jobs. Nominal wage will fall below the target inflation rate and the well-known disinflation and deflation processes will start.
Figure 12.4: Employment beyond the inflation-stable employment corridor.
If target real wage rates of firms and trade unions agree for a given employment corridor, there is an horizontal element also in the short-run Phillips curve. A constant inflation rate is compatible with all employment levels within this horizontal interval and the NAIRU becomes a corridor. The Phillips curve has now turned into a function with three elements:
\[\begin{equation} \pi = \begin{cases} \pi_{-1}+k \left( L^* - L^{N_{\text {u}}} \right) & L^*<L^{N_{\text {u}}} \\ \pi_{-1} & L^{N_{\text{u}}} < L^* < L^{N_{\text{o}}} \\ \pi_{-1}+k\left(L^* - L^{N_{\text {o}}}\right) & L^* > L^{N_{\text {o}}} \end{cases} \tag{12.3} \end{equation}\]
The upper part of the figure 12.5 shows how, thanks to the introduction of the horizontal element in the short-run Phillips curve, a constant inflation rate is compatible with different employment levels and different unemployment rates. With this modification, the NAIRU is now a range of values rather than a single definite value. As soon as employment falls above (below) the bounds of this employment corridor, inflation starts to rise (fall).
Figure 12.5: Operation of the employment corridor.
In addition to the process of macroeconomic coordination in the wage bargaining process described above, there is another reason that justify the presence of the horizontal element in the Phillips curve. After deep recessions with high unemployment, employees and unions may have little or no conflict orientation during the following recovery phase when employment is on the rise. Temporary tamed conflict may well prevent wage acceleration in the upswing of the business cycle. The economy may then have persistently low inflation under adaptive expectations even when employment rises after a recession.
If we now include the Phillips curve with the horizontal element in the 3-equation model, we can show that the political assignment of monetary, fiscal and wage/income policy can be no longer separated, as the NCM approach suggests. The result is a change in the policy mix that is still valid in macroeconomic terms. With a horizontal element in the Phillips curve, there is the possibility for the central bank to tolerate lower unemployment rates than in the NCM. Active fiscal policy can now help to ensure that an inflation-stable level of employment increased by wage coordination is achieved on a long-run basis through appropriate demand management.
This change has profound implications for the proposed macroeconomic policy mix. While the NCM approach focuses on reducing labour market rigidities, fragmenting and decentralising wage bargaining, reducing the rights of employees and dismantling the social security systems, the post-Keynesian approach proposes alternative measures. Within the post-Keynesian policy framework, wage coordination should aim at achieving the inflation target. Instead of basing their nominal wage decisions on expected (i.e. lagged) inflation, workers and firms should raise nominal wages in a coordinated manner according to the rate of medium-term productivity growth plus the inflation target. This would prevent the Phillips curve from remaining at a low (high) level after an economic downturn (upturn). The post-Keynesian Phillips curve extended to include wage policy becomes:
\[\begin{equation} \pi = \begin{cases} \pi^{T}+k \left( L - L^{N_{\text {u}}} \right) & L<L^{N_{\text {u}}} \\ \pi^{T} & L^{N_{\text{u}}} < L < L^{N_{\text{o}}} \\ \pi^{T}+k\left(L-L^{N_{\text {o}}}\right) & L>L^{N_{\text {o}}} \end{cases} \tag{12.4} \end{equation}\]
In this model, effective coordination of wage bargaining appears to be a superior policy tool when it comes to dampening inflation during the upswing of the business cycle. Moreover, effective coordination of wage bargaining facilitates the stabilisation of the economy during a downturn and prevents harmful disinflation and deflation processes. This provides the basis for actively managing aggregate demand through fiscal and monetary policy to achieve high levels of non-inflationary employment.
In the interactive scenario with post-Keynesian Phillips curve accessible via the link below, a simplified version of this set of policy recommendations can be simulated. In this scenario, the economy is again hit by a permanent demand shock. In the first round, employees follow adaptive inflation expectations. If the demand shock is large enough to bring employment above (below) the upper bound (the lower bound) of the horizontal element, accelerated (decelerated) inflation is observed. Wage, fiscal, and/or monetary policy can now be used to bring the economy back to the inflation target and to a high level of employment. In choosing wage policy within the horizontal corridor of the Phillips curve, bargaining agents will follow a post-Keynesian wage rule, i.e. unit labour costs will rise with target inflation. Since we assume productivity growth to be zero, this implies that nominal wages rise with target inflation. Outside the horizontal corridor, however, wage coordination is ineffective leading to a persistent deviation of inflation from target. In this case, fiscal and/or monetary policy must intervene to influence aggregate demand and to bring demand and employment back to the horizontal element of the Phillips curve.
This scenario shows the crucial role of policy coordination for the macroeconomic adjustment after the shock. If the shock raises (lowers) inflation and demand management is combined with post-Keynesian wage policy in the horizontal segment of the Phillips curve, it will not be necessary to bring employment below (above) the lowest (highest) inflation-stabilising level to achieve the inflation target. It is sufficient to bring employment back to the inflation-stabilising corridor through monetary or fiscal policy.
Moreover, if the assumption of a horizontal element in the Phillips curve were combined with the assumption of an interest-elastic mark-up, the idea of the NCM model on monetary policy to take on the role of demand management would have to be abandoned. Demand management with fiscal policy becomes necessary to avoid negative effects of interest rate hikes in the upswing. Monetary policy committed to low and stable real interest rates (e.g. at the rate of long-term productivity growth) can also help reduce the NAIRU and thus create more room for expansionary fiscal policy and higher employment. Therefore, in the post-Keynesian framework, wage and income policy would be responsible for nominal stabilisation, while fiscal policy should be aiming at a high level of employment with the goal of real stabilisation. Monetary policy should support this policy mix by pursuing a policy of low interest rates. In stark contrast to NCM policy, the post-Keynesian policy framework requires a high degree of wage coordination, strong business organisations and strong unions operating in organised labour markets where actors recognise and exercise the responsibility of wage bargaining for nominal stabilisation.
12.4 Economic policy in the post-Keynesian model
We can conclude this chapter by briefly summarising the policy mix (or the policy assignment) that emerges from a post-Keynesian macroeconomic model Hein and Stockhammer (2009):
Monetary policy of the central bank should aim for low interest rates. Interest rates influence income distribution and through this affect the macroeconomy. Real interest rates should not exceed the growth of real output or, in the long-run, productivity growth, because only then will the distribution between wages, retained profits and interest income remain constant. Similarly, nominal interest rates should not exceed nominal output growth.46
Labour market, wage, and incomes policies affect the price level/inflation and distribution. Flexible nominal wages tend to destabilise the economy, as they cumulatively reinforce disinflation and deflation processes, as well as inflation processes. Therefore, wage policy should aim at stable growth of unit labour cost in line with the inflation target. Nominal wages should therefore grow at the same rate as the sum of productivity growth and the inflation target. At the macro level, this requires coordinated wage negotiations with strong unions and strong business associations.
Fiscal policy is given an active role in the post-Keynesian policy mix and is responsible for real stabilisation in the short and long-run. For this purpose, politically set fiscal deficit targets and debt targets are not suited. Instead, the government’s budget deficits (or surpluses) should be designed to allow for the highest possible inflation-stable employment on the demand side. To ensure that the government remains capable of acting even in the face of high deficits and high public debt, it needs the support of the central bank, which should maintain low interest rates on public debt by intervening in the financial market.47
Policy coordination: since the economic policy actors, the central bank, the government through its fiscal policy, the collective bargaining actors with the use of their instruments have an influence on both employment and inflation both in the short and in the long-run, economic policy must be agreed and coordinated between them. In the post-Keynesian approach, a clear assignment of instrument and target to the different economic policy actors, as in the NCM approach in the long-run, is not possible.
Literature
For an introduction to post-Keynesian theory, see e.g. Hein (2023, chaps. 2–3), Hein and Lavoie (2019), King (2015), Lavoie (2006), and the contributions in Hein and Stockhammer (eds) (2011). For an overview of the foundations of post-Keynesianism and developments in post-Keynesian macroeconomics, see Hein (2017). Hein (2023) and Lavoie (2014) are detailed advanced textbooks. Post-Keynesian macroeconomic models with the corresponding policy recommendations have been presented, among others, by Arestis (2013), Hein (2023, chaps. 6–7) and by Hein and Stockhammer (2009), Hein and Stockhammer (2011).↩︎
See e.g. Hein (2006; 2023, chap. 5), Hein and Stockhammer (2009).↩︎
In addition, the central bank should stabilise the financial sector by using other instruments (lender of last resort, credit controls, guarantees of sovereign debt to keep the government able to perform its role).↩︎
Through its tax and social policy, the government also influences the distribution of disposable income. In this context, progressive taxation improves automatic stabilisers in upswings and downturns.↩︎