Euro Labour Costs

The previous chart is calculated from some recently published Eurostat data on labour costs in the business sector. For several euro member countries, I have compared changes in their average hourly labour costs (wages, salaries, benefits, etc) to the average for the euro area as a whole. Eurostat does not seem to publish absolute levels of these costs, and only gives an index number (2008 = 100 for all countries). It might be that one country's costs have risen faster than the average, but still remain below average, or vice versa. Also, these numbers take no account of productivity developments and just show changes in an employer's hourly cost of hiring a worker. Despite these qualifications, the picture is still striking.

The chart is indexed to 100 at the start of 2001, thus showing the relative change in each country's labour costs compared to the average since then. This is not to argue that 2001 is some kind of equilibrium year when everything was fine, it just makes the longer-term development easier to see than Eurostat's index numbers with 2008 as 100.

Greece also joined the euro in 2001, and it stands out in the chart. From 2001 to 2005, Greek labour costs rose by around 15% more than the average, then, after falling back, rose again into early 2010. Thereafter, relative costs slumped along with the Greek economy. In absolute terms, the level of Greek labour costs jumped from an index number (2008 = 100) of 76.7 in early 2001 to a peak of 113.4 in early 2010. By the end of 2014, the index number had crashed to 84.3. Nominal wage costs in Greece are back to where they were in 2002, and lower still when adjusted for inflation. On the face of it, this should be encouraging for capitalist employers, but there is still barely any sign of economic recovery. Profitable production depends upon more than cheap labour.

Spain stands out too, as the euro member country with the most sustained rise in relative labour costs. From 2001 to 2010 these rose by some 15% more than the euro average. Mass unemployment in Spain has made the gap narrower since then, but, by the end of 2014, Spain's relative costs were still 10% higher than in 2001. In absolute terms, Spain's labour costs have flatlined in the past few years, rather than having fallen drastically, as in Greece. Spain's index number (2008 = 100) rose from 72.3 in 2001 to 110 in 2012, where it has since stayed.

France and Italy's labour costs have risen only a little faster than the average, by less than 5% in the period to end-2014. Germany's labour costs, however, rose less quickly than the average for a decade, and only began to rise a little faster from 2010.

Tony Norfield, 20 April 2015

How Much Do Santa's Helpers Get Paid?

A large proportion of the world's industrial employment is in poorer countries. In 2012, the International Labour Organisation estimates that total world employment was some 715 million, but only 106 million jobs were in developed economies and the European Union. The biggest number of industrial jobs among the richer countries was in the US, with 26 million, but this figure fell from 31 million in 1991. The striking contrast is with the poorer countries. China employed 234 million in 2012, up from around 135 million in 1991. India's industrial employment more than doubled over the same period to 113 million, and similar developments have occurred elsewhere. Indonesia now employs 24 million and Brazil 21 million industrial workers. As many people are aware, the growth of the industrial labour force in poor countries and the shrinkage in rich countries has been a factor in so-called globalisation. The economics behind this shift of employment, and production, is the relative cost of hiring workers. While there may be other factors impacting business decisions, from tax breaks to sources of cheap energy, labour costs are the main influence over the location of production.

The next chart gives the latest picture for the hourly 'labour compensation' in different countries for the year 2013. Compensation refers to the sum of both wages paid and various benefits available to the worker. In general, the amount of non-wage compensation is very low in poor countries. These data are taken from the US Conference Board's publication a few days ago.

I have included both China and India in the chart, although the Conference Board lists their data separately because their labour cost numbers are not seen as being comparable with the data for other countries. In addition, the Conference Board data is for 2011 (India) and 2012 (China), so I have made adjustments to produce a figure for 2013 based on reported wage increases and on the moves of India's and China's currency exchange rates against the US dollar.

Each country in the chart is indicated by its two-letter ISO code (CH is Switzerland, CN is China, BR is Brazil, IN is India, PH is the Philippines, etc). The chart also mainly shows richer countries. Partly I have to do this because there is no comparable data available for some of the poorest countries, eg none for Bangladesh or Indonesia. Also, I do this to make a point: in the current 'festive season' for the richer countries, many of the presents Santa Claus will bring have been produced by workers enduring oppressive conditions and earning a wage (and not much else adding to 'compensation') that is a small fraction of the wages available to workers in countries receiving the presents.

Hourly Compensation Costs in Manufacturing, 2013
(Index based on US = 100, for a cost of US$ 36.34)

Notably, although the US is the base country for the compensation index shown, the US figure is well below that seen for several European countries and also Australia. Switzerland, Belgium and Sweden stand out here, with numbers more than 40% higher than in the US. South Korea (SK) is at 61% of the US figure, then there is a big drop down to the compensation number for Brazil (BR) at 29%, then Taiwan, Poland and Mexico. China, the Philippines and India come last, with the Conference Board data (and my estimates) putting China's workers at an average of just below 10% of the US number, and India at just below 5%.

A final comment on the statistics: how far do the wonders of the global market work to lessen wage inequalities over time? Surely, if workers are expensive in the richer countries, there will be a shift of production to countries where workers are cheaper, and eventually this will reduce the wage gaps. The Conference Board's data show something like this in the case of China. But the hourly compensation costs in China rose from a minuscule 2.2% of the US number in 2002 to just 8.6% in 2012. These data are for a 10-year period long after the influx of foreign capital and the expansion of Chinese production began and the result remains that Chinese workers are on less than 10% of US worker compensation. The most significant example of catching up is seen in South Korea, where the ratio rose from 40% to 60% between 1997 and 2013. Other Conference Board data comparing 1997 with 2013 show far less catching up with the US: the Philippines ratio rose from just 5% to 6%. In Brazil, the ratio fell from 31% to 29%; in Taiwan from 31% to 26%. This is a neat indication of the way in which the stratification of the imperialist world economy is not overcome by market forces.

Tony Norfield, 20 December 2014

T-Shirt Economics Update

In June 2011, I published an article on this blog: "What the 'China Price' Really Means". The article discussed international wage differentials, productivity and how low wages in poor countries translated into economic gains for rich countries. Using the data I had found, together with an investigative report from Die Zeit, I made a guess that the unit labour cost of a T-shirt produced in Bangladesh was some 10-15 euro cents (it sold for 4.95 euros in a German shop). That seemed reasonable, but a reader contacted me recently to point out some problems.

If the 10-15 cents labour cost estimate were true, he noted that it contradicted the other data I cited from Die Zeit, namely the 1.36 euro daily wage of one of the machine workers in the Bangladesh factory. Or else it implied that an implausibly large number of workers were employed, perhaps around 200 per machine. So, I examined the issue again, revised my guess and have reached a more damning conclusion about the rate of exploitation of workers in Bangladesh!

The usual caveats with data apply: do the figures really measure what they claim to measure? Furthermore, there are gaps in the data available, and I had to make some estimates. However, the main reason behind the much lower guess I would make now for the unit labour cost of a T-shirt produced in Bangladesh is the rise in productivity. These data come from the Bangladesh Statistics Office, and I had not seen these, and am not sure they were even published, when I wrote my blog article. They show a much larger rise than I had previously allowed for.

Another point is that I had used the results of a study by S C Zohir, published in 2000, that the unit labour cost in 1994 of a 'shirt' in Bangladesh was 11 cents (in US dollars). I did not then take into account that if the labour cost of a (full) shirt was 11 cents, then presumably a T-shirt would cost less. Assume 8 cents for a T-shirt (excluding working on the sleeves, buttons, etc, on a full shirt).

Starting from 8 US cents unit labour cost for a T-shirt in 1994, this can be translated into Bangladeshi currency (the taka) at that point. Then, the number can be inflated by the rise in wages for Bangladeshi cotton workers, but also deflated by the increase in productivity of cotton production workers. I have done this to estimate the unit labour cost in taka for the T-shirt (in 2011). In addition, the depreciation of the taka versus the euro and the dollar since 1994 also needs to be taken into account to work out what the T-shirt costs are for the buyers in rich countries.

The end result is that instead of the unit labour cost for producing a T-shirt being 10-15 euro cents, it is very likely to have been more like 2-3 cents. Even if that estimate were 20-30% too low, it would not really make any appreciable difference, given the minuscule starting point.

Going back to the original article, on the basis of 10-15 euro cents per T-shirt, I estimated that H&M's net profit per T-shirt in 2011 was 4-6 times higher than what was paid to the Bangladeshi producer. My apologies for underestimating the fruits of exploitation: the ratio is closer to 20-30 times higher.

The lesson to draw from this is that the closer you examine the economics of imperialism, the worse it gets!

Tony Norfield, 24 September 2014

Stubborn Facts

Lenin was fond of the English saying: ‘Facts are stubborn things’. The accuracy of many so-called facts may be disputable, but it can be instructive to report on the facts published by official institutions of imperialism, ones that nevertheless throw a not very flattering light on today’s realities. This article is a complement to the ‘Imperialism by Numbers’ article I published on this blog on 1 May. It is also an update to, and an extension of, some data I reported in ‘What the “China Price” Really Means’, published on 4 June last year.

The first set of facts is shown in Chart 1. These are data that cover average hourly compensation costs, where ‘compensation’ means not only wages paid, but also the additional employer payments for social benefits such as unemployment insurance, medical insurance, and old-age pensions. The source is the US Bureau of Labor Statistics (BLS), which carried out this analysis to calculate for US corporations the total costs of employing workers in a range of different countries. The details show that it is not only wages paid that are higher in the richer countries; employee benefit costs are much higher too. Chart 1 gives index numbers based on 100 equalling $34.74, the BLS figure for the average hourly compensation paid to US manufacturing workers in 2010. Countries’ labour costs are shown as bigger or smaller bars, with the height of each bar proportional to the 100 level compensation cost in the US.

For China, average hourly compensation costs are estimated at $1.65. This was less than 5% of the costs of US manufacturing employees in the same year! Several years earlier, China’s figure was closer to 2% of US costs, but recent sharp wage rises in China have narrowed the gap a little. India and Sri Lanka have a still smaller ratio of US compensation costs, near 4% and 2%, respectively. Labour compensation costs are higher in the Philippines and Mexico, but Poland is the first country from the low end of the chart that has compensation costs that are more than 20% of the US level.

By contrast, the US, Canada, Japan and the rich Europeans tower above all the other countries shown in the chart. This group includes the so-called G7 countries, the major powers still running the world economy. Switzerland, Belgium, Germany and France have compensation levels more than 20% higher than in the US. One factor influencing the country ranking is the value of a national currency in the international market. However, the gap between the top ranked countries and the bottom ranked ones is so large that this currency factor has little influence on the overall distribution.

Surprisingly, the BLS’s data do not include any African country. Perhaps this is a problem of getting comparable statistics. For example, this is the reason that the BLS does not include figures for China and India in its standard country comparison reports, though it gives some information separately. However, Africa has also been a less important continent for US economic expansion overseas than elsewhere, and the BLS data focus far more on Europe, Asia and Latin America.

Chart 1:          Relative International Labour Costs in Manufacturing, 2010

                        (Hourly costs, US = 100 is $34.74, including non-wage compensation)

Sources and notes: US BLS. 2010 estimates based on 2007-08 BLS data are made by the author for China, India and Sri Lanka. Note that 2-letter ISO codes are used as country identifiers, and that CH refers to Switzerland, not China (which is CN).

Even the relatively minuscule labour costs for the poorer countries exaggerate the actual earnings of millions of workers. The Indian data are boosted by including only the so-called ‘formal sector’, that is the sector made up of generally larger, more organised companies that have some form of regulation and government supervision – including being included in statistical surveys! By contrast, the ‘informal sector’ is unorganised, on a much smaller scale and may include a family ‘business’ that consists of the parents, children and dependent relatives. This sector is not included in most data surveys, but it accounts for a large share of employment at much lower wages than in the formal sector. The BLS reports that 80% of India’s manufacturing employment is in the informal sector.

For China, the BLS calculations of hourly compensation do include estimates for the ‘informal sector’. In Chinese statistics this is listed under the heading of ‘town and village enterprises’ (TVEs), whereas the larger, more regulated, sector is under the heading of ‘urban enterprises’. The TVEs accounted for 70% of the total workforce, with 79.1 million workers employed in 2006; the urban enterprises sector employed the other 30%, or 33.5 million workers. Not surprisingly, in 2008 the average hourly compensation was just 82 cents in the TVEs compared to $2.38 in the urban companies.[1]

American and other foreign corporations will tend to set up in the formal sector, and will likely be paying the ‘higher’ wages. But they will still benefit from the mass of even cheaper labour from poor families who work for them indirectly, either by providing services for the larger companies, or by being what Marx called the ‘reserve army of labour’ for the formal sector. The divergence in labour costs for countries other than China, India and Sri Lanka may be less extreme. For example, South Korean costs are just below half the US figure. But there is still a very big gap.

If we look at the broader economy, rather than just manufacturing, the same picture of relative incomes holds. In fact, there is a 95% positive correlation between the figures for manufacturing compensation and for a country’s per capita GDP.[2]

Chart 2 gives a snapshot of global income inequality, based on a rough estimate of the Lorenz curve for 183 countries comprising 6.7 billion people.[3] World Bank average GDP per capita data for each country are used as the input. This method may understate global income inequality, because it assumes that everyone in country A gets the average per capita income for country A. Nevertheless, it has the advantage for our purposes of putting the different countries in focus.

Global average GDP per capita in 2011 was $9200. Of the 183 countries included in the data, 124 countries with a population of 5.0 billion (75% of the world total) had an average income below this, while 104 countries with a population of 4.8 billion had an average income below $5000 in that year.

Chart 2:          The Global Lorenz Curve, 2011 (based on GDP per capita)

Source and notes: World Bank. Data for average GDP per capita in 2011 for 183 countries is used as the basis for calculating the cumulative income distribution curve, the Lorenz curve.

If we take a common measure of inequality, the Gini coefficient, and calculate this from the data in Chart 2, the figure shows the expected high level of inequality: close to 66%. It would be more like 70% if the inequality of component country distributions were also allowed for. In that case, this measure of income inequality on a global scale is on the same level as that in the most unequal of countries for which Gini coefficient data are available: Namibia.

To give specific examples, in 2011 the GDP per capita of Switzerland was put at just over $70,000, while the US number was around $47,000, Germany was $43,000 and the UK was $39,000. Compared to these figures, China was close to $4000 and India to $1300. The data from the World Bank, the IMF, the CIA and other organisations have some differences, and the figures get revised, but the rankings and the income gaps are very similar from all sources.

The basic, and not surprising, fact is that the world economy is very unequal. When we look at the mechanisms that underpin this fact, we find that the inequality has much less to do with differences in labour productivity than with the way that some countries get privileges in the world economy at the expense of others.

Tony Norfield, 22 May 2012

---

[1] See BLS Monthly Labor Review, April 2009. The data noted here are for 2006.

[2] Using the full set of BLS data for 34 countries’ compensation costs in 2010, I found there to be a 0.951 correlation coefficient with the respective countries’ per capita GDP in 2011 as reported by the IMF. This shows that the manufacturing wage/compensation is closely related to the broader economic income of the country. This is a sign that the richer, and usually imperialist, countries can afford to pay their production workers more. As the ‘China price’ article indicated, this has more to do with imperial power than being based on higher productivity.

[3] The Lorenz curve is closely associated with the ‘Gini coefficient’ of inequality mentioned later. It is a common, summary graphical measure of inequality. The 45-degree line indicates where 10% of the population gets 10% of the total income, 20% gets 20% of the total, etc. As such, it represents a line of equality of income in the population. The divergence of the Lorenz curve from this 45-degree line shows the extent of inequality. Wikipedia has a general explanation of this statistical measure and its relationship to the Gini coefficient.