The Politics of Employment

It has long been gospel among political commentator and political scientist alike that the rather unforeseen longevity of the Arab Gulf monarchies is tied inextricably to their abilities to buy off political opposition using revenues generated from their control of natural resources. Ordinary Gulf citizens, so the argument goes, are content to disavow politics in exchange for guaranteed state employment and freedom from taxes. For taxation without representation, we all know, is the one sure formula for democracy. How then can one expect untaxed, wealthy, contented citizens to be democrats?

Yet among the lessons we have learned from the case of Bahrain over the previous three months is that government benefits in the Arab Gulf do not translate neatly into more politically quiescent citizens, precisely because these benefits themselves are not distributed in a politically-agnostic way. In Bahrain and elsewhere in the region, government benefits do not buy government support; government support buys government benefits.

Nowhere is this more obvious than in the sphere of public-sector employment. In the past two weeks alone, hundreds if not thousands of Bahrainis have been fired from state-sector jobs after being accused of participation or complicity in anti-government protests. These include:

• a dean, 7 teachers, 25 administrators, and 62 students at the University of Bahran; and five Ph.D. students enrolled at foreign universities
• 85 workers at Batelco
• 100 from ALBA (Aluminum Bahrain)
  
• 100 from APM Terminals (which runs Bahrain's main port)
• 50 from Bahrain Airport Services
• all 15 board members of Bahrain Petroleum Company (BAPCO)'s trade union, with another 230 employees "under investigation"
• some two dozen Gulf Air employees
• 30 doctors and nurses from the Ministry of Health
• members of sports clubs
  

I'm sure there are others (and will be more to come), but you probably get the idea--which is that in Bahrain public-sector employment is not a tool with which to gain regime supporters but a tool with which to reward those who are already supportive, or at least are not known to be otherwise. This point is so clear that we need not even reference the wholesale exclusion of Bahraini Shi'a from the police and military (more on this in a future post!) in order to make it.

As a matter of fact, we need not even reference any of the sacked state-sector employees noted above. This is because my 2009 Bahrain survey demonstrates that well before the current political crisis erupted, Shi'a Bahrainis were already disproportionately excluded from, or at least under-represented in, the public sector. And this is true even after controlling for relevant individual-level factors that may influence one's sector of work such as age, gender, education level, and so on.

As we see in the graph above, around 38% of employed Shi'is interviewed in my Bahrain survey reported working for the public sector. This question was asked directly in the demographics section of the survey, i.e.: "What is your sector of work: public or private?" By contrast, a bit more than 50% of employed Sunnis report working in the government sector. (For those statistically-inclined, the Pearson’s chi-squared test statistic measuring the statistical independence of the two means allows one to reject the null hypothesis that they are equal, with an associated p-value of 0.061.)

Yet for many reasons this result is not satisfying. For one thing, the graph above does not account for other variables apart from ethnicity that are also likely to affect one's sector of employment. For another thing, our sample here is limited only to those who are employed, meaning that our sub-sample is not a random sample of the Bahraini population but a random sample of the employed Bahraini population, which is a truncated distribution. So if there are individual-level factors that are correlated with both ethnicity AND employment, our results above may be biased.

Fortunately, however, this worry turns out to be unfounded. Though I will not go into the full details here, there are well-known statistical models that help account for precisely this sort of bias (I use a Heckman selection model). When we estimate the relationship between ethnic group membership and sector of employment given that one is employed, we find that being a Shi'i is NOT related to whether one is employed per se, but has a strong, negative relationship with public-sector employment specifically. And this after controlling for education level, age, and gender.

Below we find the predicted probability that a random Bahraini is employed in the public sector given that s/he is employed, based on the results of the regression model described above. We see that these probabilities are essentially the same as the proportions reported above in the first graph, meaning that the relationship between ethnicity and employment sector is robust.

The likelihood that a Shi'i Bahraini is employed in the state sector given that s/he is employed is around 38%, a Sunni Bahraini 52%. Being a Sunni in Bahrain, then, increases one's likelihood of public-sector employment by about 37%, all else being equal. That is to say, for two individuals of the same age, gender, and education level, distinguished only by ethnic membership, the Sunni is more than one-third more likely to be employed in the public sector given that s/he is employed. No small advantage indeed.


I think the basic argument has been made: while the recent spate of politically-motivated firings from Bahrain's public employers offers the latest evidence that state-sector jobs in Bahrain are not distributed in a politically-agnostic way, in fact this phenomenon is nothing new. Even in 2009 (and one presumes long before), Bahraini Shi'a were disproportionately under-represented in the public sector.

Of course, none of this comes as news to Bahraini Shi'a, who are all too aware of the political bases of employment in Bahrain. The following cartoon I found on an opposition forum some time ago, showing a representative of the "Ministry of Sectarianism" distributing "government jobs" to "unemployed graduates" of "the Sunni sect" and "clerical positions" to those of "the Shi'i sect." My survey results, then, only give further evidence to what had already been long assumed.


For now, I will leave you with this uplifting quote (and one that bears on the argument here about reward and punishment) with corresponding video from Sh. Nasr bin Hamad Al Khalifa, a son, as his name implies, of the Bahraini king. It was delivered, if you can't tell by the 12-foot flag in the background, on Bahrain state TV via telephone:

"Bahrain is an island with no escape passage; everybody who interfered in these issues will be punished and everybody who took a stand [supporting the regime] will be awarded. The people who stood with or against the king are well-known to us."