Tyranny of numbers

Iran’s economic recovery: the view from employment data

Posted in Employment, General, Sanctions, Unemployment by Djavad on October 20, 2014

In my last post I examined if the quarterly growth in the GDP using data released by the Central Bank of Iran (CBI) that indicated a robust economic recovery during spring 2014.  As promised, I will now review the most recent employment figures released by the Statistical Center of Iran (SCI).

The GDP data showed the economy had grown by 4.6%, spring on spring (but not during the quarter). After two years of negative economic growth and the positive shocks of Rouhani’s election and the easing of sanctions, one quarter of positive growth should not have surprised anyone, but skeptics were unconvinced.

Then a week later came the news of unemployment having dipped below 10% last summer (2014), a first in a long time. Clearly, the fall in unemployment adds to the credibility of the GDP numbers, especially since imports data already indicate that  strong demand for intermediate industrial goods continued into the summer. Still, the same issues that raise doubts about the  for the GDP are relevant for employment — seasonality and source of the improvement.

There is no question that the easing of sanctions had something to do with falling unemployment.  Industry, which led the growth of output also led the growth of employment.  Compared to spring 2014, industrial employment grew twice as fast as total employment, up by 3.4%, followed by services (up by 2.0%) and agriculture, which was down by 1.3%.

So, the first question is what part of these changes can be attributed with seasonal factors and what part is genuine growth in employment.  A problem that makes comparison of employment data over time challenging is variation in the participation rate.  Often times, discouraged workers drop out of the labor force and answer “no” to the question have you been actively seeking employment in the previous week or month.  Many labor economists therefore prefer to use “the jobless rate” instead of the rate of unemployment.

The jobless rate gives you the same picture as unemployment rate if you assume that all variation in the participation rate during a short period are the result of the variation in the number of discouraged workers. So, I fix the participation rate at 37%, which is about the average over the last few years and measure unemployment as the difference between the size of the labor force thus estimated and the number reported as employed.  This exercise returns an unemployment rate of 9.3% for summer 2014, down from 10.4% the preceding quarter and 10.7% in summer 2013.  In general, this series tracks the SCI numbers pretty well.

In other words, the unemployment rate is actually down and it is not because of lower participation due to increase in the number of discouraged workers.  But the improvement in the economy it indicates may not be part of a cyclical rebound. It may instead be the results of positive exogenous shocks — namely, Rouhani’s election and the temporary abatement of sanctions following the November 2013 agreement with Western powers — the effect of which may not last into next year.

One should keep in mind that unemployment data are better descriptions of change over time than the actual employment conditions at any particular time.  The reason is that underemployment is prevalent in Iran and, as many observers never fail to complain, the SCI data consider anyone with at least one hour of work during the week the before interview as employed.  This is not a criticism of the SCI because it follows international guidelines set by the International Labor Office.  All complaining economists and journalists should should try to get over this and stop complaining.

What seems interesting in the SCI published data is that unemployment alone does not offer a good description of the change in the labor market conditions in Iran.  As the figure below shows, at times the labor force participation rate better reflects them.  For example, after the shocks of subsidy reform, sanctions, and devaluation, it is the participation rate that, with some delay, goes down, indicating lax demand for labor.  The unemployment rate also goes down, but that is most likely because many unemployed workers are not being counted because they have dropped out of the labor force.

Figure.  The rates of labor force participation and unemployment for the population 10+

lfp_unemployment

Source: SCI quarterly reports. If you cannot view the figure click here.

Advertisements

2 Responses

Subscribe to comments with RSS.

  1. herethereevery said, on November 3, 2014 at 8:39 am

    In the complete absence of a free press, free civil society, etc, why do you take it as a matter of course that the regime’s claims about unemployment (10%) should be taken at face value?

    If the regime were to falsify its stats, who in the country is free to publicize their doubts without fear of the authorities?

    Are there ANY claims by the clerical regime that should be doubted or all its claims to be believed?

    Are we to believe that the IRI has about the same unemployment rate as the US a few years ago and under half that of Spain? Really?! That doesn’t pass the “sniff test”.

    • Djavad said, on November 5, 2014 at 4:48 am

      I agree that the unemployment rate published by SCI is not a good measure of the conditions of the labor market. However, for purposes of comparison across countries and over time, we need to adopt some definition, and the ILO’s, which is followed by many countries, is as good as any.

      Having said that, in order to understand the plight of the unemployed, one must do more detailed work. This is something that is lacking in Iran because there is not enough research being done at the moment, at least research that finds its way into English language publications. Fortunately, much of the survey data is made available to researchers by SCI, so if people want to work with the data they can. Complaining about the data needs to be backed up with demonstration of its problems. I should add that in general it is extremely hard to falsify micro data. Such data can be of poor quality but it is hard to falsify them systematically without it showing up in research. My work with the income and expenditure surveys convinces me that the data is of comparable quality to similar surveys in neighboring countries, and with careful work one can learn a lot from them. The comparison with US and Spain need this type of research, not simply comparing the numbers. What do people do who work more than one hour but less than, say, 20 hours? They probably lead very different lives in Iran than in US.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: