Part 1: the declining share of a hard day’s work; a note on the dynamics neoliberal income distribution

Travis Fast

Thanks to the Lonely Economist for kicking off the discussion here at Relentlessly Progressive Political Economy. Lucky for me he served up a nice point of entry into my primary area of interest: neoliberal restructuring. One of the distinctive features of neoliberalism is the degree to which labour, in the advanced capitalist zone, has been pushed back on its heels; an extent to which has not been seen since before the Second World War. There are any number of metrics which could be marshalled to illustrate this shift in the balance of bargaining power between workers and employers.

One of the most straightforward metrics however is to track labour income as a share of GDP and contrast that with gross surplus as a share of GDP. In the graph (click on the thumbnail below) I have done just this for the years 1926 through to 2005. Besides providing a general periodisation for the shift from what political economists refer to as the Keynesian welfare state to the neoliberal consolidation, the graph nicely captures the changing fortunes of owners of capital and owners of human capital (note in the old economy we called these individuals workers whereas today economists refer to them as consumers who accumulate—or not—human capital; however regardless of what economists call them political economists and Revenue Canada still call them workers or employees).

It is true that I have gilded the Lilly somewhat by including the metric of gross surplus as a percent of GDP because by definition gross surplus is the residual after labour income is deducted from GDP. Yet, I included this measure to help illustrate the point: from WWII onwards, Canadian workers managed to capture an increasing share of what they produced. This trend peaked in the late 1970s from which it has been trending downward ever since and in earnest since the recession of the early 1990s reaching levels not seen since the early 1950s.

To provide and even more dramatic illustration of the nature of the neoliberal era I have plotted the metric of profits as a share (%) of labour income (click on the thumbnail below) and before tax profits as a share of GDP. Clearly the current profit boom is in good part being funded by repressing the share of output that labour receives.

Notice that profits as share of labour income increases faster than profits as a share of GDP from the beginning of the 1990s which among other things lends support to the observation that the macroeconomic logic of neoliberalism is based on the repression of worker’s capacity to get their “fair share” of the output they produce. Indeed, profits by both metrics and worker’s share of output have not been at such levels since the profiteering of WWII. The difference between then and now being primarily the trend in labours’ share of income: in the latter period it was trending up and in the present period it has been trending down. Moreover, another interesting observation that can be made from this graph is that from the early 1950s to the early 1970s labour’s increasing share of GDP does not appear to have been purchased at the expense of a declining rate of profit. This suggests that worker’s gains were paid for largely out of productivity increases.

However, from the late 1970s onwards it appears that increased profits have not only come from capitals ability to appropriate most of the productivity gains for itself but also by clawing back part of labours share. From the POV of labour, therefore, neoliberalism and all its rhetorical trappings –flexibility, positive sum gains, high road competitiveness and the like— is in reality a zero sum game in which workers are forced to take what is offered. This reality contrasts nicely with the official fiction derived from neoclassical models where workers sit in the catbird seat choosing between work and leisure and where employers are played-off one another in an infinite bidding war for human capital. In fact, our metric actually understates the degree of decline in worker’s share of value added because it includes all employee income including that of upper management and executives.

In the real economy what matters is leverage and in general there are only three ways worker’s can get leverage to bargain with employers. The first is to acquire a skill or technical competency in an area of high demand and low supply: indeed this is the individual strategy par excellence. The problem with this strategy is (a) there are very few jobs of this type so it is not a general strategy that can be pursued by all workers and (b), baring barriers to entry, it is hard for individual workers to exclude others from obtaining that specialized skill set and thereby undermining the “rents” they are earning on their human capital. The second way for workers to increase there leverage is to press for wage increases during times of low labour slack. The problem with this strategy is that periods of low unemployment are (a) relatively short and crucially (b) are not determined by the actions of workers but rather by the pace of accumulation (the economic cycle). The only way that workers can achieve, as a group and individually, on a consistent basis, a higher share of national income is to act in concert via some form of collective organisation whether that is through professional associations or unions (yes the Ontario Bar Association is a glorified craft union for lawyers as are faculty associations for prefessors and the Ontario Medical Association for doctors).

And all of this brings us directly to the causes of the decline of labour’s capacity to bargain for their share of output (GDP): the prolonged exposure to high levels of unemployment throughout the 80s and 90s and the relative decline in the percent of the labour force covered by collective agreements. The graph below provides a nice indication of the correlation between decreasing union density and the decrease in labours’ share of output. Indeed the apparent positive correlation between the level of union density and labours’ income share of output from the early 1950s to the present is so tight that it hardly merits further comment.

The relationship between unemployment and labours’ share of output is more complicated. Unlike density where the level of unionization translates in a straightforward manner into bargaining power via collective agreements the effect of low/high unemployment is less direct in terms of workers capacity to bargain. First off, workers tend to view the labour market in the same way as their income; i.e., past experience weighs more heavily in current decision-making than present circumstance. We can therefore expect a lag between the observed level of employment/unemployment and the effects on income shares. Second, where robust unemployment insurance programmes exist the effect of high unemployment on wage growth is more muted. Third, the degree of union density also helps condition worker’s responses to deteriorating condition in the labour market as the threat of the sack is less effective. As such it is quite remarkable the degree to which increasing unemployment visually correlates with a receding share of labour income (click on thumbnail below). If I were to lag the effects of unemployment on labour income shares the relationship would be even tighter.

Notice that for all the talk of record levels of low unemployment in the recent cycle if we cleave the graph into two periods 1946-1975 and 1976-2005 unemployment averaged 4.7% and 8.8% respectively. Further observe that as the unemployment trend curve gets steeper the labour share of income trend curve also moves into a state of decline. Moreover, and crucially, the present low level of unemployment does not appear to have reversed the declining trend in income shares which either reflects a lag in the effect of unemployment levels (probable) or significantly altered institutional environment in which decreasing levels of unemployment do not translate necessarily into greater bargaining power for workers as a whole (equally probable).