Free trade does not lower wages or cause persistent unemployment
Gerard Jackson
There is nothing new in the current hullabaloo about free trade, jobs and trade deficits. Believe it or not, a World Bank study, as reported by Florence Chong (The Australian, 12/2/97), argued that free trade causes widespread persistent unemployment, even though the gains from trade outweigh the costs.
Not having read the study I have to rely on Chong's report. At first I found it incredible that the World Bank could have produced such a damning finding with respect to free trade, until I reminded myself that this is the same outfit that produced another studying apparently claiming that economies can grow without savings.
That it should have made a claim about free trade that violates fundamental economic laws as well as historical experience should have been cause for little surprise but a great deal of distress.
What made Chong's report important, however, is that it drew attention to the old economic fallacy that free trade lowers real wages and causes unemployment. Once again, this fallacy has been resurrected and given a new lease of life by those who simply lack the intellectual capacity to grasp even basic economic theory, and others whose only object is to profit, either financially or politically, from the fear and economic ignorance of the mass of people. (Are you listening, Senator Kerry?)
It is now time to give some basic economic facts an airing:
a) There can be no widespread persistent unemployment in a free market so long as there is sufficient land and capital available to employ people.
b) The view that free trade lowers real wages to the level of poor trading countries was scornfully dismissed by Professor Haberler who stressed that the argument that free trade "leads to an equalisation of factor prices is fundamentally false".
c) Wages can only be equalised if labour is mobile, i.e., there are no barriers to labour moving from low-wage countries to high-wage countries. In this respect, while the public's fear about mass immigration and wage rates is justified fears about free trade are not.
d) Wages are largely determined by the ratio of labour to capital: the more capital relative to labour, the higher wages will be.
e) Free trade causes countries to allocate factors of production to their most productive uses thus raising total output and welfare.
The Treasury and the Reserve Bank pointed out some years ago that Australia's unemployment problem was not caused or aggravated by free trade policies. Moreover, in a free labour market free trade brings about a more efficient allocation of labour. Only when labour markets are prevented by wage-fixing from reallocating labour will unemployment emerge.
Put another way, when the cost of hiring someone exceeds the value of that person's output, he will not be hired. For example, if the daily value of a man's output is $100 (and it is we as consumers that set the value, not capitalists) and a union demands $110 per day or nothing, then it will be nothing and the dole queue will grow longer. And this is what has happened in Australia. Unfortunately, it seems almost impossible to get this economic truth discussed in the Australian media, let alone accepted.
Even though protectionists claim that free trade causes persistent unemployment (as opposed to transitional unemployment which every economy experiences) has been shown by economic theory and empirical research to be without foundation, their accusation that it will lower real wages to the same level as our low-wage trading partners deserves a considered reply, even though it too is without foundation.
What we may call, for the sake of the argument, the 'equalisation theory' seems on the surface to be quite reasonable. After all, do not market economists tell us that there is a tendency for the market to equalise the prices of all goods? True, they do.
However, the vitally important proviso is that the good has to be mobile, ie., transportable; the easier and cheaper it is to transport the good the greater will be the tendency toward price equalisation. It is obvious, for instance, that it is the existence of immigration barriers that prevent workers from low-wage countries crossing into high-wage countries thus bringing about the international equalisation of wages rates.
There is, however, the argument that even though low-wage workers cannot migrate, exporting their products to high-wage countries will still have the effect of driving wages down. Therefore free trade policies should be resisted.
A more refined version of this fallacy states that it will only be those employed in labour-intensive industries and whose products have to compete with cheap imports who will suffer significant wage cuts.
Therefore importing cheap goods becomes a substitute for importing 'cheap labour'. This view ignores the economic fact that wages in poor countries are largely determined by the ratio of labour to capital, just as they are in rich countries. Wages are low in poor countries because productivity is low, and this is low because workers have painfully little capital to work with. Hence free trade does not drive down the wages in poor countries.
Asia provides an excellent example of the truth that domestic productivity drives wages. A few years ago a Federal Reserve Bank of San Francisco study found that low wages in poor Asian countries are directly related to low productivity. This rather obvious finding also concluded, however, that their unit labour costs were actually close to the American level.
This apparent paradox was the result of differences in productivity. So even though labour costs in the Philippines and Thailand, for example, were low relative to U.S. labour costs, they were high relative to the value of their product which, of course, brings us right back to the quantity of capital as the determining fact and not free trade policies.
Despite economic theory and empirical evidence to the contrary, there are still those who preach that free trade with poor countries is damaging the welfare of unskilled labourers in the West by depressing wages and causing unemployment to rise.
In 1992 International BusinessWeek published a lead article that claimed that the increase in foreign trade was responsible for the "unprecedented surge in income equality between the most and least educated halves of the U.S. work force".
The argument that free trade with poor countries employing labour intensive techniques "would have a dramatic effect on lower-skilled workers in the U.S." was forcefully put. The result was that poverty and unemployment would increase, the cost of welfare would rise and so would taxes. The authors also tried to use the Stolper-Samuelson factor price equalisation theorem to support their dismal and, fortunately, baseless conclusions.
Before dealing with their claims on rising unemployment and falling incomes, it would be best to dispose of authors' theoretical justification for their conclusions. The equalisation theorem was developed from the Heckscher-Ohlin principle which basically states that trading advantages between countries arose from their different factor endowments.
Samuelson's theorem extended it and stated that free trade will lead to factor prices being brought into line with each by the prices of their products being equalised. Therefore free trade becomes a substitute for factor mobility.
So even though immigration laws prevent people from poor countries moving en mass to high-wage countries thus depressing wages, importing the products of their services will indirectly achieve the same effect.
Regardless of the authors' support the Samuelson theory does not hold up in the real world.
Now the Heckscher-Ohlin principle led to the conclusion that factor equalisation between trading countries with different factor endowments would occur if they shared the same technology. But it was also assumed that the technology was almost perfectly divisible (perfect competition, homogenous capital and identical production functions were also assumed); meaning that it could be efficiently employed even on a tiny scale.
Obviously, the key is in sharing the same technology. However, technology can only be applied through heterogeneous capital goods (investment) which cannot be perfectly divisible. This brings us to the Stolper-Samuelson theorem which does not use the shared-technology restriction to support its conclusion that free trade will equalise factor incomes.
For a factors' income (wages, for example) to be equalised in a situation where its mobility is confined to given areas, not only would it be necessary for the product of the lower paid factor to a) have free access to the other areas the factor would also b) have to be specific.
The reason for 'b' should be obvious: the price of any factor cannot be driven down below the value of its alternative use. The price of labour is determined by the value of its marginal productivity; this in turn is determined by the size of the country's capital structure. The smaller the capital structure relative to the population, therefore, the lower real wages will be and vice versa.
For American wages, for example, to fall in real terms the capital structure must either shrink or the labour force must grow faster than capital accumulation. The BusinessWeek article claimed that from 1979 to 1989 real pay and benefits for factory workers fell by 6 per cent and this confirmed the Samuelson theorem.
Lestor Thurrow used similar logic to claim that the trade deficit had driven average wages down by 6 per cent by forcing a million people out of high-paying manufacturing jobs into low-paying services jobs. However, taking Thurrow's figures at face value Professor Krugman showed that this alleged movement of labour could only have cut real wages by 0.3 per cent.
I have already stated that labour would have to be specific for its wages to be forced down by low-wage imports until wage rates in both countries were equalised for the same type of labour. But a 1985 Department of Labour study found that people who lost their jobs in the apparel and textile industries tended to find other jobs that paid as much as their old jobs.
To support the authors’ case imports from low-wage countries would have had to increase significantly; but such imports were only about 3 per cent GDP at the time as opposed to 2 per cent in 1960. Moreover, if the Stolper-Samuelson theorem seemed to be operating then the prices of traded goods produced by unskilled American labour would have fallen.
This does not seem to have happened. Even if the prices of these goods have fallen, this would be in perfect keeping with free trade theory. Critics have forgotten that under comparative advantage factor combinations in each country are rearranged by price signals to increase total output.
So if free trade did not cause falling real wages and increasing inequality, what did? Maybe nothing, at least regarding wage rates. Even though average real cash wages appeared to have fallen since 1973 real gross wages have risen. Between 1973 and 1994 real hourly labour costs rose by a 0.4 per cent a year.
The same period also seemed to experience a significant fall in productivity. As already pointed out, it is productivity that drives wages. What has deceived people about wages is that only part of the wage has been measured. That a layman should do this is pardonable; that an economist should do it is a mortal sin.
Of course, the whole question of movements in real U.S. wages is a dog's dinner. Leonard Nakamura was recently asked by the New York Times to reassess past inflation figures. He calculated that the annual inflation rate in the 1970s was overstated by 1.25 per cent and that this error has steadily risen to reach 2.75 per cent today. This means that instead of real wages falling during that period they actually rose by 35 per cent!
Whichever way one looks at it, America has done a lot better than some statistics suggest. Moreover, it is crystal clear that the case for free trade still stands unsullied and undented.
Gerard Jackson is Brookes' economics editor
BrookesNews.Com
Monday 19 April 2004