People's Democracy

(Weekly Organ of the Communist Party of India (Marxist)


No. 29

July 22, 2012


The Poverty of Anti-Unionism


Prabhat Patnaik


BOURGEOIS economics, from very early on, has sought to argue that trade unions constitute an undesirable phenomenon. This is true not just of what Marx had called “vulgar economics” which succeeded classical political economy, but even of the later exponents of classical political economy itself, like John Stuart Mill, who had propounded a “Wages Fund” theory that went as follows: at any time in society there was a fund out of which wages were advanced; if there were some unionised workers who could raise their wages by virtue of being unionised, then this meant less being left over for the non-unionised workers. Trade unions therefore could not benefit all workers; they could only benefit some workers at the expense of the others. They therefore necessarily exacerbated inequalities among workers.




This view which was articulated by a follower of Mill, Citizen Weston, at a meeting of the International Workingmen’s Association (the First International), was critiqued by Marx in a speech that took the form of a well-known pamphlet, Wages, Prices and Profits, which outlined the basics of Marx’s as yet unpublished opus, Capital, and which argued that workers as a whole could benefit through union action, in the form of higher wages, at the expense of profits. A part of the revenue that would have accrued to the capitalists would accrue instead to the workers as higher wages; and correspondingly, in lieu of the products that would have been produced to meet the capitalists’ demand arising from this revenue, different products would be produced to meet the workers’ additional demand.


After 1870 when “neo-classical economics” originated, whose exponents were authors like Marshall, Walras, Menger and Jevons, a different theoretical argument was advanced against trade unionism. This went as follows. In a capitalist economy, price adjustments occurred to “clear” all markets; i.e. through adjustments, a set of prices got established at which whatever was supplied was exactly equal to what was demanded. The labour market too “cleared” at a particular real wage rate, so that if this real wage rate came to be established, which it automatically would if the market was allowed to function freely, there would be full employment in the economy. It followed from this theory that if we actually have unemployment at the going real wage, then the reason for it lies in the fact that the market was not allowed to function freely, i.e. the real wage rate was not allowed to drop to the level required for achieving full employment; it was artificially kept up.


Trade union action according to this view therefore necessarily led to unemployment. Such action was obviously superfluous if it only succeeded in achieving for workers a real wage rate that was the “market-clearing” real wage rate, since that rate would obtain anyway through the spontaneous functioning of markets, even without unions. Trade union action became meaningful only if it got the workers a wage rate in excess of what the free and spontaneous functioning of the labour market would have fetched them. But if such a wage got established, then there would necessarily be unemployment. Hence if trade union action is effective, then it must produce unemployment. Some workers’ gain in short is other workers’ loss.


So influential was this view, namely that unemployment under capitalism was the result of “excessively high” real wages, that in the midst of the Great Depression of the 1930s, there were frantic efforts to curtail wages for reducing unemployment, until Roosevelt’s New Deal made a break with this policy.




A basic fallacy (among others) of this “neo-classical” argument was the following: if “excessively high” real wages, enforced by trade unions, was the cause of unemployment, then we should notice only unemployment, while there should be full capacity use of the capital stock (in the sense that no piece of equipment, which is not loss-making at the going real wage rate, should be lying idle). We should not in other words be witnessing both unemployed labour and idle capital occurring together. But this is precisely what we noticed in abundance during the Great Depression of the 1930s, and precisely what has characterised capitalism throughout its entire history, though of course to a lesser extent. As Michael Kalecki, the renowned Marxist economist, has said: “..…chronic underutilisation of equipment….. was a frequent phenomenon in developed capitalist economies”.


Such chronic underutilisation of equipment coexisting with chronic unemployment suggests that the constraint on the level of activity in a capitalist economy comes from the side of demand and not from “excessively high” wages. Capitalism in short is essentially a “demand-constrained” system. But then trade union action, by raising real wages, increases demand and hence employment. Far from reducing employment, trade union action therefore results in higher employment than would have prevailed otherwise.


The matter can be explained as follows: a rise in real wages, given the level of technology, and hence the level of labour productivity in any period, must mean an increase in the share of wages in output and a decline in the share of profits. Now the capitalists have a lower consumption ratio out of profits than the workers have out of their wages. Hence a shift in income distribution from the capitalists to the workers, tends to raise the overall consumption ratio in the economy. For any given level of investment then, which is not much affected by such relative share movements since it depends much more on expectations about the growth of the economy, the level of output and employment increases if the share of wages increases. Trade union action, by raising real wages, for any given level of labour productivity, at the expense of the profit-share, succeeds therefore in achieving both higher wages and higher employment, and shifts the economy to a higher level of output.




The “neo-classical” argument has been sought to be revived in a slightly different form in the recent period. This talks not of employment and unemployment, but assumes that full employment gets achieved, taking the unionised and the non-unionised workers together. If there were no unions at all, then all workers would get the same wage rate and the level of that would be one where all of them were employed. But unionisation means that some workers get a higher wage than this “equal” “market-clearing” wage; this reduces employment in the unionised sector, forcing more workers into non-unionised sectors, where overcrowding reduces the wage rate. If we compare a situation where trade unions exist, with one where they do not, then, according to this argument, employment remains the same in both cases, i.e. at full employment, but the unionised workers get higher wages, which cause a reduction in the wages of non-unionised workers through overcrowding. Trade unions in short improve the lot of some workers while worsening that of others.


This argument has been used, among others, to press for “labour market flexibility” in the Indian economy. And the fallacy of this argument, again, lies in not recognising the essential nature of capitalism as a “demand-constrained” system. It is invariably characterised by unutilised capacity, even in sectors of the economy where the workforce is unionised. And what appears as “employment” outside these sectors is often “disguised unemployment” (e.g. shoe-shine boys), whose reduction does not reduce output. The existence of substantial unutilised capacity can be established very simply in the case of the Indian economy.


Let us not even consider what “full capacity” output in the economy may be; let us just take the peak output already achieved in any given month of the year as full capacity output (even though that may itself entail significant unutilised capacity). By taking the actual output in any month relative to this peak, we get a measure of the degree of capacity utilisation for any given month, on the basis of which we can measure “unutilised capacity” for the year as a whole (or for quarters within the year). Even on this measure, the degree of capacity utilisation in the Indian manufacturing sector (which is typically where “labour market flexibility” is supposed to be introduced), has been lower in every subsequent quarter compared to the third quarter of 2009-10. In other words, the Indian manufacturing sector has been the site of significant unutilised capacity in the recent period, and the reason obviously is inadequate demand for domestic goods.




In this situation, if real wages of unionised workers are reduced, by, say, removing trade unions from the scene altogether, then the demand for domestic goods will fall even more (for reasons discussed above), which would worsen the conditions of even the workers who were hitherto non-unionised. Conversely, trade unions, by raising the wages of unionised workers, increase the level of demand for domestically-produced goods, both by raising aggregate consumption demand and also by reducing the share of imports within this aggregate because their demand is less import-intensive than capitalists’ demand; and this contributes to higher demand for labour in the economy and hence higher incomes even of non-unionised workers. It follows that trade union action benefits not only the unionised workers but also, indirectly, the non-unionised ones. It is not a case of gains for some workers and losses for others, but gains for all workers because of trade union action.


These gains are invariably associated with a decline in the relative share of profits, though not in the absolute amount of profits. Since with the rise in workers’ incomes output also expands because of the relaxation of the demand constraint, the rise in such incomes does not entail a fall in the income of the capitalists. But the relative share of profits does decline, which in turn entails a reduction in income inequalities in society as a whole.


There are some striking statistics from the United States, put together by Colin Gordon of the Economic Policy Institute. Taking a long historical period, between 1918 and 2008, it turns out that there is a remarkable negative correlation between the share of income going to the top 10 per cent of the population, and the proportion of workers who are members of trade unions. In periods when union membership increases the share of income going to the top 10 per cent of the population declines; and likewise when union membership declines the share going to the top 10 per cent increases, as has been happening in the last 30 years in that country.


Trade unions, in short, do not just raise incomes for their members. They play a larger social role, of raising aggregate demand and employment in the economy, even to the benefit of the non-unionised workers, and keeping down income inequalities. Since democracy gets seriously attenuated by large income inequalities, trade unions are an essential instrument for the maintenance of democracy. Those who argue in favour of “labour market flexibility,” which amounts in effect to doing away with trade unions, are not just propounding a wrong theory; they are undermining the foundations of democracy.