For more than a year now, it has been evident that the ''recovery'' from the Great Recession, which has been visible if sputtering in terms of output growth in the core capitalist countries, has not delivered anything like the increases in employment that were expected. A recent report of the ILO (''Short-term employment and labour market outlook and key challenges in G20 countries'', ILO and OECD September 2011, page 1) points out: ''With economic activity slowing in several major economies and regions, earlier improvements in the labour market are now fading, hiring intentions are softening and there are greater risks that high unemployment and under-employment could become entrenched. This makes for a highly uncertain outlook as to the timing and strength of a future recovery in employment. Continued weak growth in employment in many G20 countries, in turn, will make it impossible in the near term to close the jobs gap accumulated during the crisis, which amounts to more than 20 million jobs.''

Even the IMF, notorious for giving relatively short shrift to employment and seeing it as generally strongly correlated with output, has woken up to the severity of the problem in its latest World Economic Outlook September 2011. As Chart 1 indicates, the collapse of aggregate employment in the US and the continued stagnation at low levels in the eurozone do not point to any real recovery at all in terms of employment. Now that the global economic horizon has been darkened once again by very real fears of the next dip, the concern is that employment conditions will deteriorate further.

It is true that the IMF's analysis of the problem and possible solutions remains weak, also because it continues to stress the need to encourage ''a rebalancing from public to private demand'' in these core capitalist regions, at a time when this is completely unrealistic to expect. A major cause of the crisis was the excessive build-up of private debt (taken by both households and companies) that could not be sustained. These private agents now necessarily have to go through a period of deleveraging. In that period, if aggregate demand has to grow at all, it must come from the public sector - but the combination of bond market vigilantes and fiscal hawks active politically has forced governments in both these areas to move to premature fiscal retrenchment.

In this analysis, the fact that there was not more of an employment recovery is a source of surprise. After all, the G20 countries did at first combine to provide very large fiscal stimuli in the major countries, and in the developed world monetary easing has continued, leaving the world economy awash with liquidity. The argument seems to be ''we adopted Keynesian policies, but they have not delivered employment growth''.

This is actually less than the truth. Part of the problem is that the stimulus measures adopted in most countries were not weighted in favour of employment generation: a disproportionate amount went as bailouts and support to large financial institutions that simply used the resources to clean up their balance sheets. In the US, very little of the money went into direct state spending on activities that directly increase employment or have high multiplier effects. Social spending and government employment fell as local governments were strapped for cash; small businesses have been starved of bank credit; there has been no systematic attempt to address the continuing problem of foreclosures in residential housing markets. And now, even these half-hearted and slipshod stimulus measures are to be clawed back with the new focus on fiscal austerity.

In Europe the imbalances in the eurozone are also being dealt with in a counterproductive manner - forcing regressive austerity measures on to deficit countries and sending them into a downward spiral of falling output and employment in which their fiscal and public debt measures will only get worse. It is ridiculous to expect private investment and activity to increase to fill the slack created by public cuts, in this context of continuing crisis. So it is not surprise that employment is not recovering.

The poor performance of employment in the developed countries has reinforced perceptions that the crisis has intensified and accelerated structural shifts in global employment away from the rich countries to certain emerging markets. Certainly, the data presented in Chart 2 would appear to support that view. It is evident that total employment in developed countries has still not recovered to pre-crisis levels. However, in developing countries total employment did not fall after the crisis, and since then has continued to rise.

The data in Charts 2 to 5 (all from the ILO's Short Term Indicators of the Labour Market, September 2011) should be interpreted with some caution, since they relate only to (around 54) countries that provide recent employment data of the required periodicity, and large countries such as China and India are therefore excluded. Even so, they provide a quick estimate of the ongoing trends.

Chart 3 indicates a similar picture for paid employment (or number of employees): - the slight decline followed by stagnation at below pre-crisis levels in the developed world accompanied by continuing increase in such numbers in the developing world. The level of paid employment in April 2011 was around one per cent below pre-crisis levels in the developed countries for which estimates are available, but 15 per cent higher for developing countries.

The biggest shift, and the one that has grabbed the most attention and created disquiet in rich countries, is the shift of manufacturing employment. This was a shift that was widely discussed but did not actually take place in the preceding decades: in fact aggregate manufacturing employment in the developing world did not increase despite the widespread perception of the North ''exporting jobs'' to the South. But the very recent trends after the global crisis suggest that the shift may be occurring now.

Chart 4 shows that manufacturing employment in developed countries was more than ten per cent below the pre-crisis level in April 2010, and has since recovered only slightly to be still around 9 per cent lower in April 2011. Meanwhile, after a minor blip in early 2009, manufacturing employment in developing countries continued to increase, such that in April 2011 it was nearly 9 per cent higher than its level of four years earlier.

Surely this is a clear sign that the much feared (or much anticipated) shift of manufacturing activity to the South is finally taking place and that the location of additional manufacturing employment will now be concentrated in the South? It turns out that even this is not so clear, if such employment is further disaggregated.

Chart 5 shows the indices of paid employment in manufacturing (that is, the total number of employees rather than self-employed engaged in manufacturing activity). This presents quite a different picture: one in which the level have fallen after the crisis, in both developed and developing countries! Indeed, in April 2009 the collapse in paid manufacturing employment was similar in both regions, at around 8-10 per cent lower than the April 2007 level. The recovery in developed countries thereafter was negligible. Such employment recovered more rapidly and sharply in developing countries, but in April 2011 the level was still below the pre-crisis level of paid manufacturing employment even in developing countries.

So the only real increase in manufacturing in developing countries seems to have been in self-employment. What exactly does this mean? This really points to the proliferation of petty activities at the bottom of the production chain, typically in low productivity and low paid work that usually reflects the absence of other viable income earning opportunities. The expansion of self-employment in manufacturing in the developing world, including in some of the most ''dynamic'' emerging markets, cannot be seen as a very positive sign of industrial relocation, especially in a world in which economies of scale are still rampant. It essentially indicates a growing tendency to newer forms of organisation of production in which there is international centralisation of production but decentralisation of the actual work processes, with the risks of production borne largely by the self-employed workers themselves, at the bottom of the production chain.

Recently released survey data from India (the National Sample Survey for 2009-10) suggest that even such employment, low paid and adverse as it is, can also be fragile and transient and therefore decline. If this is a more widespread tendency, when more data are eventually available for all developing countries, we may find that aggregate manufacturing employment in developing countries has barely recovered after the crisis.

* The article was originally published in the Business Line, October 3, 2011

October 4, 2011.