Some Thoughts on Wages and Competitiveness

There’s a lively debate going on about Philip’s earlier comments about competitiveness and recovery and I wanted to add to it but then wrote something so long I decided it would be best to exploit privilege and start a new post.

Let’s start with what I see as two correct points being made by those focusing on wage cuts.

1. Labour demand curves slope down. So an exogenous decline in wage rates will increase the demand for labour. And more people working will result in a higher level of output (in textbook microeconomics, this all happens at once, as the factor demand decisions of the firm simultaneously determine output.)

2. For an exporting producer, the wage that matters for their profits, and thus their labour demand, is the real product wage. So if a devaluation raises the real product wage, then this will raise labour demand and output. With this option gone, there is an increased focus on the first point.

That’s fine as far as it goes. However, consistent calls from economists for wage cuts to assist economic recovery should also be interpreted against the background of a few wider issues. I’ll focus here on four areas.

The Macroeconomics of Labour Markets

Despite what seems to me to be an exceptionally strong attitude in this country of calling on the government to solve every possible problem, we are largely a market economy and wage rates are set in a relatively decentralised fashion compared with other European countries. And despite the faith of many that unregulated labour markets should always clear to produce full employment, we have plenty of macroeconomic evidence that this is not the case.

The reality is that, in all economies, negative macroeconomic shocks tend to raise unemployment because wages never adjust quickly enough to get the labour market back to full employment. This has been a mainstream theme in macroeconomics since, at least, the General Theory.

In more recent decades, New Keynesian macroeconomic theorists have put forward a plethora of models to explain why the labour market does not operate in a simple market-clearing fashion (efficiency wages, implicit contract theory, bargaining models based on “holdups”). More recently, behavioural economists have documented the importance of “money illusion’’ which makes workers particularly resistant to cuts in nominal wages. The result is a significant amount of empirical evidence demonstrating the existence of nominal and real wage rigidity.

This is not to argue that wages are completely rigid or that the labour market does not have mechanisms to bring unemployment down after a negative shock. Macroeconomic data generally show good fits for Phillips Curve relationships such that wage growth is low when unemployment is high. But governments will generally not want to rely only on this mechanism to restore macroeconomic equilibrium because the pace of recovery will be too slow. Instead, they prefer, where possible, to use countercyclical fiscal and monetary policy.

In relation to Ireland today, these points suggest that, whatever the government does, wage cuts on their own are unlikely to deliver a fast economic recovery, though very high unemployment rates will restrain wage growth in the coming years.

But Aren’t Wages Far Too High?

One reaction to the points above is that we are in an exceptional situation that differs from that described by textbook macroeconomics. Whereas I have focused on the normal process of market-driven wage-setting, some would argue that what happened in recent years in Ireland represented some form of extreme departure from normal market practices and that we now need to come back to reality. And certainly, anyone who reads the newspapers or listens to the radio will regularly hear that “we’ve been paying ourselves far too much in recent years.”

But what’s the benchmark here? The various comparisons that one reads about nominal or real wage growth in recent years or wage levels relative to other countries still tell one little about what is a reasonable underlying level of wages. For instance, high real wages (W/P) can be justified by high levels of productivity (Y/L). A useful indicator that allows us to compare real wages with productivity is the labour share of income (WL/PY or real wages divided by labour productivity).

Data on the labour share of income from the European Commission (see page 94 of this PDF file) tell us that Ireland’s labour share of income declined throughout the 1980s and 1990s and bottomed out at 52.4 percent in 2002. So, throughout that period, productivity rose faster than real wages. This pattern reversed itself a bit in subsequent years, so that by 2007 the labour share had risen back to 58 percent. However, this is still a very low level by international standards. The EU-15 share for that year was 65.3 and has remained in a narrow 65-67 range over the current decade.

There are some measurement issues that affect this calculation—in particular the existence of some very high productivity, low employment plants in the Irish pharmaceutical sector—but I’m pretty confident that the conclusions about both the trends over time in Ireland and the level by international standards can hold up to appropriate adjustments.

By this reckoning, Irish wages could still have been judged as relatively low relative to justifiable productivity levels in 2007. Let me be clear, however, this is not the same as arguing that the housing bubble didn’t damage the economy. The remarkably skewed incentives that stemmed from the bubble meant that we ended up with a huge fraction of people employed in the construction sector and (since we were already at full employment) this squeezed out employment in the more cost-sensitive tradable sector. The next few years will see this process reversed, but only slowly and painfully.

What Can the Government Do?

Of course, our wages are not solely market-driven. The government can have an influence on wages through its influence as the state’s largest employer. By cutting wages for public sector employees, this can have an effect on economy-wide wages by influencing the various relativities. I’m not sure, however, exactly how strong this mechanism is. There is strong evidence that Irish public sector wages are above comparable prevailing market rates. Cutting public sector wages may end up reducing this premium without actually affecting private sector wages if public sector jobs are still seen as more attractive.

In any case, it should be kept in mind that the government has already taken some action here—the public sector “pension” levy reduced net take-home pay for public sector workers by about 7 percent. This is a large wage cut by international standards. It is likely that there are more public sector pay cuts to come but the impetus will largely be the state of the public finances and the evidence for public sector pay premia, rather than their impact on economy-wide wages.

Beyond public sector pay, the Irish government could perhaps encourage lower wages by deregulating the labour market. It is well known that average levels of unemployment across countries tend to correlate positively with labour market rigidities, so reducing these rigidities may help to limit the increase in unemployment.

Three points can be made on this. First, even highly deregulated labour markets such as the US are suffering from high levels of unemployment during the current recession, so deregulation is not a panacea for recession. Second, Ireland does not have a highly regulated labour market: The World Bank’s Doing Business survey ranks Ireland 38^th in the world in terms of the freedom of its labour markets, compared with 142^nd for Germany and 148^th for France. Third, social protection measures are popular with the electorate and, to some extent, many citizens may be willing to accept higher unemployment rates in exchange for labour protection laws and generous welfare benefits.

Within the set of available options and against the background of deflation, I think the comments from various government ministers that a cut in the minimum wage needs to be considered, and the Snip recommendations of reduced welfare rates, are reasonable. However, I think the overall effects on employment of these measures would be limited.

The Loss of the Devaluation Tool

Finally, there has been a lot of hand-wringing about the loss of the devaluation tool. I think this loss is being overstated. Yes, devaluation can have a positive short-term effect but its competitive effects tend to get offset fairly quickly by inflation. And remember that EMU monetary policy has contributed to counteracting our current recession by giving us historically low interest rates of 0.5%. The cost of the existence of the devaluation tool is the high interest rates that come from the threat that we use it occasionally. The current period is a reminder that the long-term gains that come from the commitment to EMU can come at the expense of occasionally substantial short-term losses.

Of course, being in a currency union influences how macroeconomic policy should be formulated. However, while many influential commentators have focused on allowing wages to get out of hand as the key mistake made during our period in EMU, I would argue that the mistakes relating to our decidedly pro-cyclical fiscal policy and shrinking tax base were an order of magnitude more important. It is these mistakes that have left us without a macroeconomic tool to soften the effects of the recession.

None of these points are meant to criticise those who argue for cutting public sector pay rates, minimum wages and social welfare rates, all of which can be justified in the current situation. However, without a useful context and reference to actual policies that the government could undertake, generalised grumbling about wage levels isn’t very useful.