Tyranny of numbers

The most recent labor force survey (LFS) results for spring, 2014, that Iran’s Statistical Center released last week must be disappointing for Rouhani’s economics team.  This survey, which  is collected quarterly and is put out with remarkable speed, is the only official data that give us a sense of how the economy has been doing most recently.  The short report shows job gains in agriculture (by 25%), which may be mostly seasonal, and services (by 2%), while industry lost jobs (by 1.3%).  The fewer number of jobs in industry is disappointing because the main benefits of the agreement between Iran and the 5+1 signed last November were expected to come in industry.  The employment picture that LFS paints for industry is not very complimentary for Rouhani’s first year in office.  Industrial employment, according to LFS, has been declining since he took office a year ago.  Last summer, more than 7.5 million workers were employed in industry, 9 months later fewer than 7 million are working there.   Agriculture and services have also lost since then, though by less, for a total of one million jobs lost since he took office.

LFS has a large sample, about half a million adults, so it tends to be fairly reliable in its quarterly trend.  I do not trust its unemployment numbers because of the large fluctuations in participation rates that it records, something that I have noted previously.  Judging by the census data for 2011, the LSF probably undercounts the number of the unemployed, but not the employed.  The census and LSF of 2011 (1390) both report 20.5 million employed, but the LSF counts nearly 0.7 million fewer people as unemployed.

Last week another set (link in Persian) of employment numbers came to light, from the Ministry of Cooperatives, Labor, and Social Welfare, showing 0.7 million more jobs in 1392 (2013) compared to 1391 (2012).  The LSF data show only 185,000 more jobs added in 1392, which is more in line with my expectations of 1392 being a worse year, as sanctions tightened and the rial collapsed.  The LSF numbers are also more in line with the lackluster performance of Iranian labor markets during 1385-1392 (2006-2011).   As I reported in an earlier post, census data reveals that only 14,000 jobs were created per year during that period.

The important question is: can industrial output increase without more employment?  The Minister for Industry, Mines, and Commerce, Mohammad Reza Nematzadeh,  tossed off (link in Persian) in a speech some positive statistics of industrial expansion during the last quarter: 90% increase in light auto production, 30% in home durables, 9% in steel, and 6% in petrochemicals.  The numbers for auto and petrochemicals make sense since these two industries were specifically mentioned in the Joint Plan of Action of November 24, 2013 between the 5+1 and Iran.  One plausible way to reconcile the increases in output announced by Nematzadeh with the more systematically reported decline in the number of industrial jobs is the following:  formal industrial employment in Iran is very rigid and does not follow fluctuations in output closely.  During recession, employers hold on to their workers, perhaps with reduces hours, and use them to produce more output when the time comes to do so.

Postscript: Here is a graph of the data that I found on the SCI website (except for spring 1392). Industrial employment has been declining continuously for the last three quarters while agriculture and services show some recovery in the most recent three-month period.