Examining Industry Wage Differentials in the Palestinian Territories

The difference between the measure of inter-industry wage dispersion used here and that in Krueger and Summers (1988) is that the latter measure is calculated by taking into account the size of the standard errors of the industry coefficients (for a more detailed explanation see footnote 16). I also use the measure used in Krueger and Summers (1988) in the remainder of the analysis.

The intifada (initiated in September 2000) and the Gaza blockade (officiated in June 2007) induced several changes in the Palestinian economies of the West Bank and Gaza. Table 7 in Appendix highlights the divergence in relevant labor market outcomes between the West Bank and Gaza after the intifada and the Gaza blockade by reporting descriptive statistics during the pre-intifada period (1999), shortly after the onset of the intifada (2002) and the post-Gaza Blockade period (2009) for both territories. While the West Bank and Gaza were always different, the changes between them were exaggerated by the intifada and the blockade. This issue is briefly revisited in Section 4 when considering the competitive wage explanation that industry differentials reflect changes in transitory demand shocks across industries. For more detail on changes in Gaza after the blockade, see Adnan (2012a).

Data for the US, Germany and Japan were taken from Gittleman and Wolff (1993); their datasets included a maximum of 20 industries, some of which were missing for Japan and Germany. For the West Bank and Gaza, I matched the two digit industries in the Palestinian Labor Force Survey Data (PLFS) to the 20 industries named above.

This may be attributed to the fact that in addition to being a third world country, Gaza suffers from the economic isolation imposed upon it from the blockade.

I exclude women because they have low labor force participation rates on the order of 16% in the period of study 2008Q4–2010. Furthermore, less than half of the women who were labor force participants were in the wage sector and reported wages. Hayo and Caris (2013) examine why female labor force participation rates in the MENA region are so low.

See Table 7 in the appendix and Adnan (2012a) for more details.

Note that (daily) wage rates in the Palestinian territories are high relative to other developing countries in the (MENA) region. This may be explained by the recent and ongoing flows of Palestinian labor to Israel’s labor market. Since Israeli wages are the highest in the region, upward pressure wages in the West Bank and Gaza is expected.

The informal economy in the West Bank and the Gaza Strip may explain why almost all wage earners in the services industry are employed in large firms. Unfortunately, the survey does not allow the researcher to distinguish between workers in the formal economy and those in the informal economy. Interestingly, Palestinian employees in Israel’s formal economy can be distinguished from those in Israel’s informal economy through a survey question about the worker’s ID type and permit status conditional on Israeli employment.

The adjusted standard deviation is computed as the square root of the variance of industry wage differentials minus the variance of the standard errors of the industry parameters:

Where β represents the vector of industry differentials and σˆi2 is the variance of the standard errors for each industry i. The extent to how much the adjusted standard deviation is underestimated depends on how large the standard errors of the parameters are. Note that the standard deviation is an underestimate since covariance terms are not accounted for.

All specifications in this paper are not sensitive to whether education is measured linearly (years of schooling) or non-linearly (four schooling groups).

Note that changes in the industry wage structure can occur regardless of the changes in the measure of inter-industry wage dispersion.

Although one can argue that workers who believe their jobs are not good matches may earn less money because their skills are not suitable for their jobs, I included this measure to address the number of imperfect matches that result from relatively skilled individuals settling for jobs that require a lower set of skills due to the increases in the unemployment rates in the West Bank and Gaza in the past 10–15 years (see Table 7 in the appendix). Nevertheless, when this control is removed from the regression, the results barely change for both territories.

This answer choice refers to workers who were offered part-time or seasonal work positions; the survey does not distinguish unsatisfied part-time workers from unsatisfied seasonal workers.

After the onset of the intifada (September, 2000), Palestinian residents in the West Bank and Gaza were subjected to a variety of labor mobility restrictions ranging from Israeli border closures where Israel and Israeli settlements were inaccessible to physical closure obstacles (e.g. checkpoints, roadblocks, earth mounds, etc.) sporadically placed to limit and monitor the movement of Palestinians. Furthermore, Palestinian residents in the West Bank and Gaza inherit ID cards issued to them at the age of 15 years by the Israeli Authorities that determine their degree of mobility (Tawil-Souri 2011, 2012). These cards are especially relevant during times of conflict when border closures and closure obstacles play a large role in labor mobility.

Since the intifada occurred in the fourth quarter of 2000, only the first three quarters of 2000 are included in the sample.

The full results also imply that some industries were more adversely affected by Gaza’s economic isolation than others. For example, the industry differentials of Commerce, Agriculture and Manufacturing plunged while the differential for the service sector rose. These results are expected given the losses incurred by the private sector coupled with the dramatic rise in public sector spending following the intifada.

The high correlation between industry differentials during the pre-intifada period and the post-blockade period in Gaza suggests that despite the large increase in inter-industry wage dispersion over the past 10–15 years, Gaza’s economy reached a new steady state where industry differentials are much larger but have similar rank order stability.

If these data were available, fixed effects for various types of collective agreements would be included; the extent in which industry differentials are attenuated after the inclusion of these fixed affects highlights the degree in which there is differential bargaining power across industries.

However, the fact that blue collar workers experience a considerably higher level of wage dispersion than white collar workers among nonunion workers in Gaza (last two rows of Table 4) also gives credence to the union threat model since blue collar workers tend to pose a greater union threat (Katz 1986)

For example, firms may pay their blue -collar workers a premium to avert unionization but will pay white-collar workers a premium as well to raise group effort (Akerlof 1982, 1984).

The reason for this is because according to the labor market segmentation described in Piore and Doeringer (1971), private sector workers are strictly supervised while public sector workers are loosely monitored and therefore heterogeneity in monitoring costs across industries will likely lead to higher wage dispersion for public sector workers. This is very similar to the previous analysis on firm size.

A more likely explanation for the lack of wage dispersion in the public sector suggests there is little to no heterogeneity in monitoring costs across industries in the public sector so that workers are loosely monitored and all industries pay a wage premium relative to the private sector but within the public sector, there is almost no inter-industry wage dispersion. The generous public sector wage premium is consistent with the expansion of public sector budgets in the territories (especially Gaza) following the intifada and the Gaza Blockade.