With Japan having general elections today and Germany facing them next month, I thought now might be as good a time as any to have a look at a topic which could turn out to be very important in the months to come: the real underlying rate of unemployment in both these countries.
While the present focus of most press attention is on the fact that GDP in Germany and Japan nudged upwards between April and June (over Q1), we should never forget that this increase follows substantial falls in output. Japan’s real GDP fell at a record pace in Q4 2008 and Q1 2009 (annualized declines of 13.5% and 14.2%, respectively), and German GDP fell by a quarterly 3.5 percent in Q1 and an annual 6.7% - making for the fourth consecutive quarter of negative growth. In both cases the fall in output was accompanied by only a much more moderate decline in employment.
Part of the explanation for this recent return of both economies to growth lies in the fact that both countries have very substantial stimulus and employment protection programs in place, and these to some extent mask the extent of the output slump. At the same time both countries have been in run up periods to national elections, while both of them have rapidly aging populations, rising health and welfare costs and steadily deteriorating gross debt to GDP positions. It is therefore highly likely that the positive stimulus programs will wane somewhat after October as both governments are forced to move from very expansionary fiscal positions, to more or less "belt tightening" ones, and the big issue which lies in front is estimating just how far the respective labour markets can deteriorate in the two countries as a result. Fortunately analysts at Nomura (for Japan) and Societe Generale (for Germany) have recently produced what are very timely studies which help us get a better appreciation of the true underlying situation.
Japan's Struggles To Raise Exports In The Face Of Tepid Domestic Demand
Apart from the temporary relief which came from a headline GDP growth reading in the second quarter, the news coming out of Japan at the moment is almost uniformly bad. The unemployment rate is now at a record high, raising even more doubts about the real sustainability of the recent economic recovery. The jobless rate rose to a worse than expected 5.7 per cent in July, up from 5.4 per cent in June. By Spanish or Latvian standards this may seem very tame, but if you take into account the extent of government subsidized "hidden unemployment" the true underlying rate may be nearer 12 or 13%, or at least this is what analysts at Nomura (see below) have recently been arguing.
In addition, Japanese core consumer prices fell at the fastest annual pace on record in again in July, potentially putting pressure on a reluctant Bank of Japan to rein in deepening deflation. Core consumer prices - the bank of Japan's preferred measure - which exclude volatile fresh food prices but include oil costs, fell 2.2% in the year to July.
Average monthly Japanese household spending fell in July by a price adjusted 2.0 percent from a year earlier to 285,078 yen, down for the first time in three months, while Japan’s July exports fell 1.3 per cent on a seasonally adjusted basis from June, a real big deal this in a country whose economy is almost entirely dependent on exports. Shipments in July fell 36.5 per cent by value year on year, outpacing the 35.7 per cent decline in June.
click to enlarge
Under the Japanese employment protection scheme, the government pays two thirds of the wages of workers in certain specified situations. As of June 2009, some 2.383mn workers had applied for employment adjustment subsidies for 2009 onward (see chart). Although only 1.891mn had been approved as of June, Nomura expect this figure to eventually rise closer to the number of applicants.
In order to try to get a measure of the impact of this scheme on the unemployment level the Nomura analysts did a number of tests, and found that when indexing the number of those employed and real GDP to the output peak of Q4 2007 there was a considerable gap between the output adjustment and the employment one. Based on a simple calculation which assumed this "gap" to offer a good proxy for the amount of “hidden underemployment”, they arrived at a figure for “hidden jobless” in Q1 2009 of 4.7m. This number is far higher than the June unemployment figure of 3.48m. If the hidden jobless are included together with the registered “unemployed”, they calculate that the unemployment rate would have leapt from 5.4% to 12.2%.
This chart below shows estimates made by the Nomura analysts of the extent of hidden joblessness during previous economic downturns. In the majority of cases the number of hidden jobless did not rise at all, or employment fell by more than GDP did, suggesting that employment adjustments were quite swift. They did find, however that there was a comparatively large rise in the numbers of hidden jobless in Q4 1973, triggered by the first oil shock and in Q2 1997, following the Asian currency crisis, problems in the Japanese financial system and a consumption tax hike. The number of hidden jobless, estimated at about 4.7m in Q1 2009, seems to be well above the figures associated with these two earier downturns.
However, as the Nomura analysts point out, even if the difficult labour market conditions are not fully reflected in the unemployment figures, a deterioration in adjusted labor supply-demand could easily lead to major declines in wages even while companies keep the number of hidden jobless down by means of government support. In fact, according to Nomura it would be hard to explain the fact that the decline in wages (total cash earnings of full-time employees) in H1 2009, at 4.7% y-o-y, was far bigger than the equivalent declines in annual average wages of 2.3% in 2002 and 0.4% in 2003, when unemployment also reached new highs, if you don't take the hidden jobless factor into account (see chart). Thus the resulting pressure on wages and household income could easily become a serious impediment to any genuine fully fledged economic recovery.
As of June 2009, payments under the government's subsidy scheme for employment adjustment totaled ¥101.14bn. If the number of approvals grows to meet the number of applicants, Nomura estimate total payments will increase to ¥127.42bn, placing additional strain on fiscal finances. The Nomura analysts argue that dividing the cost of the hidden jobless and in-house unemployed between companies, households and government could well represent a major policy challenge for the new Japanese administration when it assumes office. Given the seriousness of the social and political problems that result from sharp rises in unemployment, it is quite possible the next Japanese government will strengthen the employment adjustment subsidy scheme by, for example, further relaxing its terms and conditions, and it is rather unlikely to try to limit the number of eligible cases by tightening conditions. The net upshot of this, will, of course, be a further deterioration in Japan's gross debt to GDP position.
Given this, and despite the distortions being produced by the huge number of hidden jobless, Nomura think the unemployment rate is unlikely to be allowed to rise too far beyond 6.0%. We will see.
And Germany Does The Same
Despite some small improvement in headline GDP numbers, the great German job machine effectively ran out of steam last autumn, and since that time the German economy has been adding jobs at an ever slower pace. Now the rate of job creation has turned negative, and less Germans are employed every month than they were a year earlier.
German unemployment rose again in July. The number of people out of work increased 52,000 to 3.46 million on an unadjusted basis. The seasonally adjusted total actually fell by 6,000, according to the statistics office due to statistical changes. Without the impact of the changes, the office estimates unemployment rose by 30,000. German unemployment began to increase in November after falling steadily for more than three years. The seasonally adjusted jobless rate was unchanged at 8.3 percent in July.
Analysts at Societe Generale, have examined the case of the German employment protection program, and point out that while official unemployment in Germany has in fact only risen moderately in the current recession the underlying real effective rate is much higher. The unemployment rate (using the ILO measure) has risen by just 0.6ppt from its 7.1% low in Q4 2008, while in the euro area as a whole, the rate is up by 2.2ppt to 9.4% from its March 2008 low of 7.2%. As they say, it is also quite clear that this relative stability owes much to the widely-used practice of so called short-time working (Kurzarbeit).
The Societe Generale interpretation is broadly supported by survey evidence which suggests that the rate of contraction in employment has eased, suggesting there will be an even slower increase in unemployment in coming months. For example, the employment component of the PMI survey in manufacturing industry rose to 37.9 in July from a low of 32.9 in April, and in the services sector to 49.0 from a low of 45.2 in May. Employment intentions (in the European Commission survey) have also come off the lows in all sectors, but still remain in negative territory, implying further job losses. This evidence tallies with other recent evidence from official unemployment data, which show unemployment up by an average of 8,000 per month in May-July, a big shift downward from the average monthly increases of nearly 60,000 in Q1. This level of improvement will, according to Societe Generale not be sustained, although they do not expect to see a return to the pace of unemployment gains witnessed earlier this year. Of course, as SocGen point out, company employment intentions could easily deteriorate again if growth expectations get revised down, but for the nearer term, the evidence suggests that unemployment in Germany will rise at slower rates than observed earlier this year.
On the other hand they also underline that such short-time working arrangements evidently have a "sell by" date, and can't run forever. As a result there is some concern that a major increase in unemployment in Germany is merely a matter of time.
In fact the Societe Generale analysts are not that convinced by this line of argument, since they think that German legislation has already extended the period for which companies can run short-time working from 18 to 24 months. Examining in detail the evolution of the numbers on short-time working they find that the vast majority of companies only resorted to the program in the very recent past, so that the 24 month limit will not bite until late-2010. Until the turn of the year 2008/09, the recourse to short-time working was very small indeed. Aside from the seasonal increases in the first quarters of 2007 and 2008, the numbers were small at around 50,000. To put the number in context, they point out that this represents 0.1% of the labour force and is equivalent to the monthly gains in unemployment that were recorded this year. Since then, the numbers resorting to the program have indeed exploded and by March of this year (the latest available data), there were 1.3 million workers with shortened hours, and this number has probably now risen to around 1.4 million. These are clearly big numbers, amounting to about 3% of the labour force. If they were added to unemployment figures, total unemployment would rise to the previous historic peaks of around 5 million.
However, given that this increase only began in the final two months of 2008, these schemes could easily run for another 18 months, at least as far as the administrative rules are concerned. Whether the German fiscal position will allow this once the new government is installed is another question entirely. Indeed, according to an article in the Financial Times last week, Germany faces a potential wave of corporate restructurings just after the elections, restructurings which will involve substantial job losses and which have not been announced previously due to the existence of an implicit "pact" not to announce big job cuts ahead of the September 27 ballot. We will, as they say, see the proof of the pudding here in the ultimate eating, but levels of German fiscal support at the present level cannot run for that long unchecked after the election results are announced.