The E¤ect of Immigration on Unskilled Native Workers: Evidence from a Natural Experiment Muhammad Asali Immigration and the Unskilled Natives JEL Classi…cation: J61, J31

International School of Economics, Tbilisi State University, 16 Zandukeli Street, Tbilisi 0108, Georgia; E-mail [email protected].

1

Abstract This study exploits the natural experiment, provided by the start of the second intifada, to measure the e¤ect of immigration on the wage and employment of unskilled native workers. It …nds that immigration has no e¤ect on the wage or employment of unskilled Jewish workers. The wage and employment of the least-skilled Israeli Arab workers (with 0-5 years of schooling) are adversely a¤ected by immigration. The slightly more skilled Arab workers (6-11), in contrast, are positively a¤ected by immigration, suggesting a complementarity e¤ect with this group. Di¤erent explanations are proposed.

1

Introduction

September 30, 2000 marked the eruption of the second (aka Al-Aqsa) intifada— Arabic for the Palestinian uprising. The number of Palestinians employed in Israel, as a consequence, dropped by 84,000 workers (about 7% of the Israeli male workforce). I use this natural experiment to study the e¤ect of immigration on labor market outcomes of unskilled native workers.1 As immigration policy heavily hinges on the e¤ect immigrants may have on local labor market outcomes of natives, it is very important to study and estimate this e¤ect. Estimates 1

While the identi…cation strategy in this article is based on the intifada event rather than the frequent

border closures, it is worth emphasizing that Angrist (1996) was the …rst to use border closures as an identi…cation strategy to study the wages of Palestinian workers.

2

from empirical spatial (cross-city and cross-state) analysis, for the United States and other countries, were mostly small and insigni…cant.2 Borjas (2003) and Borjas & Katz (2007) associated the small estimated e¤ects of immigration with the spatial analysis approach taken so far, and suggested that the e¤ect in question is best estimated at the national level because internal migration of natives might di¤use this e¤ect. Using national level data they found a signi…cant negative e¤ect of immigration on the wage of unskilled natives. Ottaviano & Peri (2012), in the tradition of Borjas (2003), extended the national approach to the analyses of wage e¤ects of immigration, by using a nested-CES framework, imposing restrictions on the cross-elasticities between natives and immigrants of di¤erent skill groups. They arrived at modest negative, and sometimes positive, long-run e¤ect of immigration on the real wages of the least educated natives. Their …ndings, despite the analysis being at the national level, were in line with those from city-level (spatial) analyses. The current study contributes to the literature on the e¤ect of immigration on native wages and employment in di¤erent ways. First, it utilizes the fact that Israel is a small country, where national-level and spatial analyses are indistinguishable. Therefore, the estimated e¤ects are more likely to be unbiased, and thus not marred by the confusion such dichotomy of approaches poses. Nevertheless, the reported estimates in this study are more likely to represent the partial, direct e¤ect of immigrants on the labor market, given the special case 2

See the comprehensive review by Friedberg & Hunt (1995). More recent studies, like Card (2001), Card

(2005), and Card & Lewis (2007) among others, report similar outcomes.

3

of Palestinian workers who come to Israel as day workers rather than on a more permanent basis.3 Second, it shows how the e¤ect of immigration is experienced di¤erently by di¤erent native groups (Israeli Arabs and Jews in this study).4 Third, rather than assuming a uniform e¤ect of immigration over time, as is the case in much of the existing literature, it measures the potential e¤ect of immigration in the immediate run and in the medium run.5 Finally, by the very nature of the large and sudden shock to labor supply induced by the intifada, it shows the potential reactions of labor markets to sudden in‡uxes of immigrants and allows us to measure the elasticity of labor demand for di¤erent ethnic and skill groups. More recent studies also utilized the natural experiment provided by the second intifada to answer related interesting questions about the Israeli and Palestinian labor markets. Mansour (2010) utilizes the labor supply shock induced by the second intifada to study the labor market on the other side of the border, namely the e¤ect of the labor supply increase in the 3

Most of the existing literature analyzes the case of permanent immigrants. These, however, may have

indirect e¤ects on wages and prices through their demand for local goods, not only through labor supply shocks. See, for example, Lach (2007) who studies this other channel of the e¤ect of (permanent) immigrants on the host economy. 4

This is similar to Borjas et al. (2010) who, in their analyses of the impact of immigration on natives,

also distinguish between di¤erent native groups, namely black and white workers in the US. 5

Cohen-Goldner & Paserman (2011) also studied the e¤ect of immigration over time. They showed that

highly-skilled immigrants from the Former Soviet Union had had an adverse short-run e¤ect on wages of male and female natives, but this e¤ect vanished in the medium and long run. Employment had not been a¤ected by immigration, neither in the short-run nor in the long-run.

4

West Bank on the wage and employment of Palestinians, due to the returning Palestinian workers: those who previously worked in Israel but are now unable to do so, due to mobility restrictions following the start of the second intifada. He …nds that low-skilled wages and employment are adversely a¤ected by the increase in labor supply of either the low-skilled or the high-skilled workers. Miaari & Sauer (2011) have found the Israeli-Palestinian con‡ict and the presence of foreign workers to have a signi…cant negative e¤ect on the earnings and employment of Palestinians. Friedberg & Sauer (2003) have also found that foreign workers in Israel substantially reduce the wages of Palestinian workers. Aranki & Daoud (2010), nonetheless, have found that foreign workers in Israel have no e¤ect on Palestinian employment, only reducing the wage of Gazans working in Israel. They associate the substantial reduction in Palestinian employment observed in Israel with the Israeli closures policy rather than the presence of foreign workers in Israel. The subject of these studies was the Palestinian workers, rather than the Israeli Arab and Jewish workers within the Israeli labor market. More closely related to the current study is that of Miaari et al. (2012), in that it looks at the e¤ects of the con‡ict on job separation rates of Israeli Arabs and Jews, rather than on Palestinian workers or the relationship between foreign workers and Palestinians. They found that the outbreak of the second intifada increased the job separation rate of Israeli Arabs relative to their Jewish

5

counterparts.6 Earlier studies, like Card (1990), Hunt (1992), and Friedberg (2001) utilized similar natural-experimental settings, in which the number of workers increased immensely in a short period of time, to study the e¤ect of immigrants on labor market outcomes of natives. Card (1990) found virtually no e¤ect of the in‡ux of immigrants (which increased the Miami labor force by 7%) on the local wages or unemployment; Hunt (1992), studying an in‡ux which increased the French labor force by 1.6%, found that unemployment increased at most by 0.3% and wage decreased by at most 1.3%; also she found that local or international migrants were not crowded out or discouraged from migrating to mass-migration regions. Friedberg (2001), studying the in‡ux of Russian immigrants to Israel in the early 1990s, also found no adverse impact of immigration on labor market outcomes of natives. The main conclusion of these studies is that immigrants have small or no e¤ect on labor market outcomes. The e¤ect of Palestinians working in Israel on the native Israeli workers, however, was not studied before. While sharing the same simple analytical framework, the current study …nds that immigration does have an e¤ect on the wage and employment of native workers, and this e¤ect is experienced di¤erently by di¤erent native groups. In particular, I …nd that 6

Other studies utilized the Israeli-Palestinian con‡ict— around and after the second intifada— to study

di¤erent questions, for example Zussman et al. (2008) who studied the e¤ect of the con‡ict on asset prices in Israel and the Palestinian Authority; and Jaeger et al. (2012) who studied the e¤ect of violence during the con‡ict on the public opinion and forming political preferences, among others.

6

the second intifada, while having no e¤ect on the employment and wages of the unskilled Israeli Jewish workers (with 0-11 schooling years), had a positive e¤ect on the employment and wages of the least-skilled (0-5 schooling years) Israeli Arab workers but a negative e¤ect on the outcomes of the slightly more skilled Israeli Arabs (6-11 schooling years). The main …ndings in this study, as to the discrepancy between the intifada e¤ects experienced by Israeli Arab and Jewish workers, and the apparent complementarity between the more-skilled Israeli Arabs and Palestinians, are consistent with two, not necessarily mutually exclusive, potential interpretations: labor market discrimination— intensifying after the second intifada, and bilingualism of the more skilled Israeli Arab workers (who speak both Arabic and Hebrew). The article proceeds as follows. The following section describes the data used in this study. Section 3 outlines the empirical strategy. Section 4 reports the main results of the study, and Section 5 concludes.

2

Data

I build a panel of cross-sectional income data, which come from the Israeli income survey …les for the years 1999–2001. The income surveys are conducted by the Central Bureau of Statistics (CBS), and are based on questionnaires conducted quarterly at the household and individual level and covering information on demographic, personal and labor market characteristics. 7

The samples include Jewish and non-Jewish respondents living exclusively in (and thus are residents and citizens of) Israel— as opposed to Palestinians who live in the West Bank and the Gaza Strip and who are not citizens of Israel. The analyses in this study are restricted to 21–64 year-old workers, with eleven or less years of schooling.7 We focus on this group being a close substitute for Palestinian workers in Israel, who are uniformly unskilled.8 Table 1 reports the means of the main variables for workers with positive income, from the income surveys, over the study period. It pertains to less-skilled workers, with 11 years of schooling or less, aged 21–64. The starting age is chosen to be 21, as this is the average age at which (mainly Jewish) workers enter the labor market, after their military service. The worker’s hourly wage is calculated by dividing monthly income by the product of hours worked per week and working weeks per month. Reported income levels in the original data …les were already adjusted to the average price level of the survey year, thus making income comparable across quarters. I then adjusted these income levels to the average price level of the year 2000, using the consumer price index; thus the real income data are comparable across quarters and years. It is worth noting that, due to this procedure, no arti…cial jump in wages would be created at the years’ends. 7

Including the high-school graduates (twelve years of schooling) yields qualitatively similar results.

8

See Klinov (2010).

8

Table 1: Sample Means of Key Variables Hourly Wage Schooling Experience Married % Quarter Arab Jew Arab Jew Arab Jew Arab Jew Q1/1999 20.6 26.5 8.7 9.4 22.5 27.7 70 75 Q2/1999 21.5 26.3 8.7 9.4 20.0 28.2 63 77 Q3/1999 19.9 25.5 8.6 9.5 19.9 27.6 69 72 Q4/1999 21.5 27.1 8.7 9.2 21.7 28.7 73 74 Q1/2000 19.5 25.6 8.4 9.4 19.4 27.7 62 72 Q2/2000 20.3 27.3 8.9 9.2 20.4 29.5 65 75 Q3/2000 21.8 26.2 8.7 9.5 20.0 27.3 70 74 Q4/2000 20.0 27.3 8.7 9.2 20.1 28.8 68 76 Q1/2001 22.1 27.5 8.6 9.3 20.6 29.6 64 75 Q2/2001 21.1 27.0 8.7 9.3 21.6 28.5 75 73 Q3/2001 23.7 27.6 8.9 9.3 21.0 29.3 59 72 Q4/2001 21.2 26.9 8.7 9.5 21.2 28.4 74 75

Observations Arab Jew 157 569 191 543 206 523 180 548 182 521 173 499 185 487 168 495 143 468 150 532 171 460 178 508

NOTE.— Source: author’s calculations from the income survey …les. Samples include both males and females and are restricted to 21-64 years old workers with positive earnings, with 11 or less years of schooling. Hourly wage is expressed in New Israeli Shekels of 2000 (see text for details). Potential experience is de…ned as: Age Schooling 5. Reported are weighted means, using sample weights provided in the data.

9

It is clear from the table that the Jewish workers earn more,9 are more experienced, have more years of schooling, even within this narrowly-de…ned schooling group, and are more likely to be married than the Arab workers. Note that, since the potential experience is de…ned as age

schooling

5; the three military service years are counted as labor market

experience. This is reasonable if we believe that skills obtained during the military service are useful, and rewarded, in the civilian labor market. Nonetheless, none of the results in this study changes as a result of using a di¤erent de…nition to account for this fact: namely, to de…ne experience as age

schooling

5 for Arab workers and age

schooling

8 for

Jewish workers.

3

Empirical Strategy

The second intifada started on September 30, 2000, which is, exactly the last day of the third quarter of the year 2000. This fact is relevant here because the data used in the study are quarterly, and therefore, we have data about the labor market right before and right after the intifada. As a result of the intifada, and in less than a couple of months, the number of Palestinian workers in Israel dropped dramatically: from 121,000 workers at the end of the third quarter 9

This apparent wage gap between Israeli Arab and Jewish workers received more attention in Asali (2010),

where the evolution of the wage gap in Israel between 1990-2003 was documented, and possible explanations of its sources were discussed.

10

120 Palestinian Workers (x1000) 60 80 100 40

1998q1

1999q1

2000q1 Quarter/Year

2001q1

2002q1

Based on data from Bank of Israel .

Figure 1: Palestinian Workers Employed in Israel

of 2000 to about 37,000 workers in the next (fourth) quarter of 2000 (a drop of 84,000 workers, which is about 6.9% of the Israeli male workforce).10 Figure 1 shows the number of Palestinians employed in Israel in each quarter between 1998–2002. It is evident from the graph that Palestinian employment in Israel was on the rise before 2000; this was a trend that started after the Oslo Accords (Palestinian-Israeli peace negotiations) of 1993. The discontinuity in the number of Palestinians employed in Israel is also 10

The intifada in this paper is a “reverse boatlift,”— as opposed to the Mariel Boatlift of Card (1990)— in

that the number of immigrants dropped (rather than increased) sharply.

11

clear: it is exempli…ed by the drastic fall in their number in the fourth quarter of 2000— right after the second intifada started. Thereafter, the level of Palestinian employment in Israel never went back to its previous levels before the intifada. Analyses in this study focus on the period centered around this episode. The dramatic change in the number of Palestinian workers, a result of political tension, constitutes a natural experiment, during which the number of “immigrants”changes exogenously, and for reasons clearly outside the labor market. Furthermore, the intifada event study allows us to see the e¤ect of Palestinians not confounded by that of the foreign workers, who also constitute a source of “immigrants”in the Israeli labor market.11 To study the e¤ect of the drastic labor supply shock due to the intifada, I investigate di¤erent periods around this event. For example, the …rst period investigated is the “immediate run,”that is, comparing the third quarter of 2000 (ending September 30, which is the same exact day the intifada started) with the fourth quarter of 2000, right after the intifada started. The second period compares data from the two quarters before the intifada with that from the two quarters following it. I de…ne …ve study periods in the same way; these 11

Between the third and fourth quarter of 2000 the number of foreign workers increased only by 1,000

workers in the agriculture industry and 2,000 workers in the construction industry. Hence, the before-after comparison allows us to safely assume that the number of foreign workers was roughly constant around the intifada, and thus associate any e¤ects on the labor market solely with the lost Palestinian workers. Nonetheless, in the empirical analyses I have controlled for the number of foreign workers in each industryquarter cell.

12

September 30, 2000 Second Intifada Starts Q3/99 Q4/99 Q1/00

Q2/00 Q3/00

Q4/00 Q1/01 Q2/01 Q3/01 Q4/01 Period 1 Period 2 Period 3 Period 4 Period 5

Figure 2: Study Periods Around the Second Intifada

are shown in Figure 2. Taking di¤erent, symmetric intervals around the event serves several ends. First, wider bands can absorb other side e¤ects of the intifada which can be still at play right after it started. Second, increasing the length of the time span, especially that before the event, controls for trends prevailing in the labor market before the intifada started. Finally, looking farther into the future helps see the immediate run e¤ect of the intifada as well as that of the medium run: whether the e¤ects wane with time or carry longer into the future. I de…ne the dummy variable Ii to take on the value 1 if individual i is observed after the intifada and zero otherwise— that is, Ii = 1 for all observations coming from or after the fourth quarter of 2000. This is the “treatment” variable. I also de…ne the “treated” group: these are workers employed in industries in which Palestinian workers are overly represented. Figure 3 below shows the ratio of Palestinians to male Israeli workers in di¤erent industries. Data on the distribution of Palestinian workers across industries were obtained from the 13

Figure 3: Palestinians-to-Israelis Ratio Across Industries

Ministry of Industry, Trade, and Labor (ITL).12 The vertical line represents Q3=2000; at the end of which the second intifada started. It is clear from the graph that the “traditionally Palestinian” industries in Israel are agriculture and construction. Palestinian workers constituted the majority in these industries for a long time. Therefore, the e¤ect of the intifada— and the loss of Palestinian workers— will 12

Generally, the ITL data are based on surveys carried out by the Israeli Central Bureau of Statistics

(CBS), the Palestinian CBS, as well as their own surveys. I thank Mr. Beni Feferman, chair of the manpower planning authority at the ITL, for granting me access to these data.

14

be more likely felt in these industries. Therefore, I de…ne the agricultural and construction industries as the treated group. The variable Pi , which stands for traditionally Palestinianemploying industries, captures this de…nition: it takes on the value 1 if individual i is employed in either the agricultural or the construction industries, and zero otherwise. Figure 4 plots the average log-hourly-wage of both Israeli Arab and Jewish workers, by industry and quarter, for the working samples. For some industries, like electricity and banking, there were only a few (or no) observations for Israeli Arab workers— who have 0-11 years of schooling and yet are employed in these industries. The vertical line in these …gures signi…es the third quarter of 2000, at the end of which the second intifada started. Table 2 describes the distribution of Israeli workers with positive income over industry and quarter cells. The …rst column lists the construction and agriculture industries (that is, for which P = 1) and the second column pertains to all other industries (for which P = 0). The traditionally-Palestinian industries were more ‘popular’before the intifada, and constituted about 13-16.5% of the samples of employed workers. These portions went down to 9.8% in the …rst quarter of 2001 and further down to 5.4% at the end of the fourth quarter of 2001— although some of this within-year decline might be accounted for by seasonality. I use a di¤erence-in-di¤erences approach to estimate the e¤ect in question. I run the following regression, separately for every native group (Israeli-Arabs and Jews) by OLS ln wite =

e

+

1e Iite

+

2e Pite

+

3e

(Iite

Pite ) +

X

je Characteristics (j)ite

+ uite

(1)

j

where wite is the real hourly wage (in NIS of 2000) of individual i; observed at time t 15

Manufacturing

Electricity

Construction

Vehicles Repair

Transport

Banking

Business Activities

4.5 3.5 4 2.5 3

log hourly-wage

2.5 3

3.5 4

4.5

Agriculture

1999q1

Education

2001q1

Health

2.5 3

3.5 4

4.5

Public Administration

2000q1

1999q1

2000q1

2001q1

1999q1

2000q1

2001q1

1999q1

2000q1

2001q1

Quarter/Year (vertical line=Q3/2000) Arab

Jewish

Figure 4: The Average log-hourly-wage by Industry: Israeli Arab and Jewish Workers

16

Table 2: The Distribution of Workers by Industry within Quarters Percentage of Observations in Quarter Construction and Agriculture Other Industries Q1/1999 12.8 87.2 Q2/1999 16.5 83.5 Q3/1999 15.6 84.4 Q4/1999 13.0 87.0 Q1/2000 13.2 86.8 Q2/2000 9.4 90.6 Q3/2000 13.7 86.3 Q4/2000 12.4 87.6 Q1/2001 9.8 90.2 Q2/2001 6.0 94.0 Q3/2001 8.2 91.8 Q4/2001 5.4 94.6 NOTE.— Source: author’s calculations from the income survey …les. Samples include both male and female Israeli (Arab and Jewish) workers, and are restricted to 21-64 years old workers with positive earnings, and 11 or less years of schooling. First column pertains to the portion of observations in the construction and agriculture industries combined (for which P = 1); the second column covers all other industries (for which P = 0).

17

(t = Q1=1999 to Q4=2001), of native group e (e = Arab; Jew). As described earlier, I stands for Intifada— workers observed in Q4/2000 or after, and P stands for Palestiniansabundant industries (agriculture and construction). “Characteristics” is a vector of control variables which include marital status, type of locality (metropolitan areas like Jerusalem, Tel-Aviv, Haifa, Be’er-Shevaa’), schooling, experience, squared-experience, dummy for male, and the number of foreign (guest) workers employed in Israel (within i’s industry). I

P is the interaction term between intifada and

traditionally-Palestinian industries. Finally, uite is the error term, assumed to be a white noise, with mean zero, and independent of the other covariates.13 The coe¢ cient of I

P,

3e ;

is the coe¢ cient of interest in this study, estimated for both

Arabs (e = A) and Jews (e = J). This measures the e¤ect of the intifada (a negative supply shock of 84,000 workers) on the wage of each of the native groups. A positive coe¢ cient of group e means that workers of this group and Palestinian workers are net substitutes. 13

The mean-independence assumption is justi…ed here by the fact that the intifada is an event that is

independent of the labor market conditions— that is, wages and employment in the Israeli and the Palestinian labor markets before the intifada could not be used to predict it, neither were they believed to have caused it. This natural experiment aspect of the study has an attractive property rendering this error term independent of the other covariates, in particular I, which here stands for the “Intifada,” but otherwise is representing the drop in the number of Palestinian workers.

18

4

Results

Overall E¤ect on Unskilled Labor Table 3 presents di¤erence-in-di¤erences estimates of the e¤ect of Palestinians on the wage of Israeli Arabs and Jews. These are the results from estimating equation (1); since the Palestinian intensiveness of industries is measured only at the industry level, while the regressions are estimated at the individual level, the standard errors are clustered by industry. Column (1) of the table uses data from the third and fourth quarters of the year 2000. This is period 1 as referred to in Figure 2. Columns 2–5 use data from periods 2–5 accordingly, every time covering a longer period (longer by two quarters) centered at the intifada start date. The upper panel of the table reports regression outcomes for Arab workers, and the lower panel for Jewish workers. The coe¢ cient of I measures the e¤ect of the intifada on the control group: industries employing few or no Palestinians (P = 0). This is mostly positive in the table, but statistically signi…cant only for Jewish workers— suggesting that the intifada has had a positive e¤ect on the wage of Jewish workers in other industries. The coe¢ cient of P measures the average wage di¤erential between industries in the treatment group (Palestinian-employing industries) and industries in the control group, before the intifada started. These coe¢ cients are statistically not di¤erent from zero. The coe¢ cient of interest is that of I

P (…rst row of each panel), which measures

the net (di¤erence-in-di¤erences) e¤ect of the intifada on the wage of native workers. The 19

intifada seems to have a negative, but mostly statistically insigni…cant, e¤ect on unskilled Arab workers. For example, in the second period the (signi…cant) estimated e¤ect is

0:096;

which is equivalent to an estimated elasticity of +0:6:14 The e¤ects for the longer-term are mostly negative but all are insigni…cant. Also it is apparent from the table that unskilled Jewish workers are not a¤ected by the intifada. The di¤erence-in-di¤erences estimates have mixed signs and are not statistically di¤erent from zero at any studied period.

The Least-Skilled and Possible Explanations

Arguably, Palestinian workers in Israel, being uniformly unskilled, are more likely to be substitutes for the least-skilled workers even within the 0-11 schooling group. It is reasonable, therefore, to pay a closer attention to the schooling distribution of unskilled native workers. I decompose the sample of unskilled workers into two subsamples: the least-skilled (those with less than primary education, 0-5 years of schooling) and those with more schooling years (middle or high-school, 6-11 years of schooling). 14

The average number of all (Palestinian, foreign, and Israeli) workers in Palestinians-employing industries

changed after the intifada. The changes resulting from the intifada (i.e., associated with the Palestinian shortage of workers) are:

13:67%,

15:64%,

15:35%,

15:43%, and

15:04% for the …rst, second, third,

fourth, and …fth periods, respectively. Therefore, taking the estimate of

0:096 from the second period (…rst

row, second column in the table), we arrive at an elasticity estimate of 0:6. That is, for a 10% Palestinianinduced supply shock, the wage of Israeli Arabs increases by 6%.

20

I

P

Table 3: The E¤ect of the Intifada on the Wage of Unskilled Workers Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 Arab Workers :015 :096 :039 :040 :035 (:0510)

I

:078 (:0466)

P

:017

N R2 Adj. R2

I

P

I P N R2 Adj. R2

(:0490)

(:0585)

:025

:021

(:0384)

(:0273)

:096

:061

(:0525)

:042

(:0267)

:066

(:0495)

:034

(:0253)

:045

(:0866)

(:0823)

(:0676)

(:0560)

(:0502)

353 .199 .164

669 .160 .140

1,001 .141 .127

1,352 .146 .136

1,736 .153 .146

:004

:056

(0:0452)

(:1094)

:072

(:0276)

:083

Jewish Workers :050

:046

:083

:003

(:1021)

(:1014)

:046

:040

:041

(:0207)

(:0265)

:000

(:1240)

:087

(:0160)

:084

(:0137)

:082

(0:1252)

(:0876)

(:0568)

(:0625)

(:0601)

982 .193 .180

1,949 .156 .149

3,002 .158 .154

4,010 .161 .158

5,041 .158 .155

NOTE.— * p<10%, ** p<5%, *** p<1%. Dependent variable is the log of real hourly wage. I is a dummy variable that takes on the value 1 for intifada periods (Q4/2000 and beyond) and zero otherwise. P is a dummy variable for the treatment group: it takes on the value 1 if the worker is employed in “traditionally Palestinian”industries: in our study it is agriculture and construction; and zero otherwise. Every column refers to a di¤erent interval length around the intifada, which started in the last day of the third quarter of 2000. So, …rst column covers one quarter before and after the intifada started; second column covers two quarters before and two quarters after the intifada started, and so on. Other independent variables are: marital status, years of schooling, experience, experience squared, male dummy, the number of foreign (guest) workers in the industry, and type of locality. Samples are restricted to workers aged 21–64, with 11 or less years of schooling. Standard errors are clustered by industry, and are reported in parentheses.

21

To test the hypothesis of di¤erent e¤ects between the two groups (0-5 and 6-11), I de…ne the variable Hi which takes on the value 1 if individual i has 6 or more years of schooling (a primary or high-school education within the unskilled), and it takes on the value zero if the individual has less than six years of schooling. I then run the following regression, separately for every native group15 ln w = +

+ 7 (I

1I

+

P

2P

+

H) +

3H

X

+

4

(I

P) +

5

(I

j Characteristics (j)

H) +

6

(H

(2)

P)

+u

j

The above regression augments Equation 1 by adding the terms H; I I

P

H: It is worth noting that, in this speci…cation,

4

H; H

P; and

measures the (di¤erence-in-

di¤erences) e¤ect of the intifada on the very unskilled workers, those with less than six years of schooling. The e¤ect of the intifada on the more educated workers relative to the very unskilled ones, the di¤erence-in-di¤erence-in-di¤erences e¤ect, is measured by the absolute e¤ect of the intifada on the more educated workers is

4

+

7:

7:

Finally,

16

Table 4 reports the results of these regressions for Jewish workers, over the di¤erent study periods. The …rst row of the table reports the estimated the estimated

4:

7;

the second row reports

The third row in the table reports the absolute e¤ect on the 6-11 group:

15

The subscripts i (individual); e (e = J or A), and t (time) are omitted to simplify the notation.

16

Table 8 in the appendix reports the distribution of the working observations by education: 0-5 schooling

years versus 6-11 schooling years. As is clear from the table, the samples of the very unskilled workers (0-5) are very small for both populations, casting some di¢ culty on estimation and inference, and the results should be viewed with this consideration in mind. I thank the referee for pointing out this fact.

22

4

+

7:

The di¤erence in the e¤ect of the intifada on the two skill groups (the unskilled and the least-skilled), measured by ^7 in the …rst row, is not statistically di¤erent from zero. The more educated group seems to fare slightly better after the intifada (in periods 3-5) as it appears from the …rst row of the table, with a measured e¤ect of .08–.20, but this is statistically signi…cant only in the fourth period, ending one year after the intifada. The negative sign of the e¤ect on the least-skilled group is not meaningful, as one would expect a positive or no e¤ect on this group. The contradictory signs of the e¤ects on each skill group cannot be reconciled with a consistent economic explanation. Besides, the absolute e¤ects on each group are very small, at least as compared to the e¤ects on Arab workers (as shown in Table 5). The absolute e¤ects on the very unskilled, 0-5, Jewish workers (coe¢ cients of P; or the second row in the table) and on the 6-11 group (

I

4

+

7;

third row) are also

insigni…cant for all the study periods.17 This suggests that the intifada had no e¤ect on the least-skilled Jewish workers, and supposedly neither on the slightly more skilled group. A series of Wald tests for the joint signi…cance of

4

and

7

for all the study periods, yielded the following p-values: 0.94, 0.61,

0.85, 0.16, 0.29, respectively, con…rming this …nding. The respective results for Arab workers are reported in Table 5. As is clear from the 17

The (insigni…cant) e¤ect on the least-skilled Jewish workers in the …rst period, 0:038 from the second

row, can be translated into demand elasticity of

0:28; because the Palestinian-induced labor supply shock

in the …rst period was about 13.7%.

23

Table 4: The E¤ect of the Intifada on the Wage of Unskilled Jewish Workers Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 P H ( 7) :039 :137 :079 :195 :106

I

(:1362)

I

P ( 4) 4

I P

+

7

H H

I P H

Constant N R2 Adj. R2

:038

(:1356)

:0008 (:0467)

(:1294)

(:1369)

:076

(:1104)

(:1046)

:120

:180

:098

(:2259)

(:1471)

(:1368)

:061

:041

(:1156)

(:1227)

:015

(:1048)

:033

:045

:018

:043

(:1439)

(:0895)

(:0933)

(:1046)

:103

:146

:081

:118

(:1879)

(:1211)

(:1012)

(:0764)

:101

:088

:064

:080

(:1385)

(:0858)

(:0820)

:009

:051

(:1187)

(:0839)

(:0694)

:011

(:0659)

(:0902)

:008

(:1011)

:017

(:0694)

:194

(:0677)

:025

(:0609)

:026

:099

(:0572)

(:0376)

:196

:122

:145

:156

:158

(:1724)

(:0987)

(:0791)

(:0844)

(:0684)

2.273 982 .196 .180

2.376 1,949 .157 .149

2.453 3,002 .160 .154

2.385 4,010 .164 .160

2.455 5,041 .159 .156

NOTE.— * p<10%, ** p<5%, *** p<1%. Dependent variable is the log of real hourly wage. I is a dummy variable that takes on the value 1 for intifada periods (Q4/2000 and beyond) and zero otherwise. P is a dummy variable for the treatment group: it takes on the value 1 if the worker is employed in “traditionally Palestinian”industries: in our study it is agriculture and construction; and zero otherwise. H is a dummy variable that takes on the value 1 if the worker has primary school education or higher, that is, if he has 6 or more years of schooling (“High” education), and zero otherwise. The third row includes the sum of the coe¢ cients of I P H and of I P ( 4 + 7 ). Every column refers to a di¤erent interval length around the intifada, which started in the last day of the third quarter of 2000. So, …rst column covers one quarter before and after the intifada started; second column covers two quarters before and two quarters after the intifada started, and so on. Other independent variables are: marital status, years of schooling, experience, experience squared, male dummy, the number of foreign workers in the industry, and type of locality. Samples are restricted to workers aged 21–64, with 11 or less years of schooling. Standard errors are clustered by industry, and are reported in parentheses.

24

second row in the table, the very unskilled Arab workers bene…t from the intifada, because they substitute for their Palestinian counterparts. The e¤ect slightly wanes with time, but is statistically signi…cant only in the …rst period (and marginally so in the last period). This lends support to the hypothesis that the least-skilled Arab workers are substitutes to the Palestinian workers and they bene…t from the negative supply shock, thus con…rming economic theory which predicts an inverse relation between wages and immigration-induced supply shifts. The slightly more educated group (those with 6–11 years of schooling) are harmed by the intifada relative to the very unskilled group.18 Taking the second period (second column), for example, the wage of those with less than six years of schooling increases by 0.13 (second row, second column), but the wage of those with more than six years of schooling decreases by 0.247 (…rs row, second column) relative to the former group— thus, in absolute terms, their wage is lowered by 0.116 (:131

:247; third row). This is equivalent to a wage elasticity

of the least-skilled Arab workers of

0:84 (that is, for a 10% Palestinian-induced increase

in the labor supply the wage of the least-skilled Arab workers decreases by 8.4%); and to a wage elasticity of the more-skilled Arab workers of +0:74 (for a 10% Palestinian-induced increase in the labor supply the wage of the more-skilled Arab workers increases by 7.4%). These estimates are larger than most of these reported in the literature.19 18

This is estimated by the coe¢ cient

19

For example Borjas (2003) reports elasticity of

elasticity in the ballpark of

0:1 and

7

of I

P

H; …rst row in the table. 0:4 to

0:6; Cohen-Goldner & Paserman (2011) report

0:3: Ottaviano & Peri (2012) estimate a wage increase of natives

25

Table 5: The E¤ect of the Intifada on the Wage of Unskilled Arab Workers Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 :341 :247 :267 :215 P H :731

I

(:1589)

I

P 4

I P

+

:681

(:1683)

7

H H

I P

Constant N R2 Adj. R2

(:1676)

:131

:207

(:1189)

(:1892)

(:1153)

:152

(:1260)

(:1411)

:277

(:1677)

:050

:116

:060

:064

:065

(:0478)

(:0570)

(:0513)

(:0453)

(:0351)

:243

(:1585)

:103

(:0815)

:105

:045

(:1231)

(:1132)

:027

:068

(:0943)

(:1421)

:297

:069

:021

(:1690)

(:1042)

(:0864)

:098 (:0846)

H

(:1532)

:079

:002

(:0893)

(:1816)

:006 (:0993)

:003

(:1029)

:047

(:0805)

:061

(:1346)

:061

(:0695)

:083

(:1252)

:022 (:0765)

:035 (:1536)

:243

:130

:103

:084

:094

(:0859)

(:0646)

(:0668)

(:0572)

(:0572)

2.311 353 .217 .172

2.838 669 .164 .138

2.795 1,001 .144 .127

2.169 1,352 .151 .138

2.154 1,736 .158 .148

NOTE.— * p<10%, ** p<5%, *** p<1%. Dependent variable is the log of real hourly wage. Other independent variables are: marital status, years of schooling, experience, experience squared, male dummy, the number of foreign workers in the industry, and type of locality. Standard errors are clustered by industry, and are reported in parentheses. See notes of Table 4 for details.

26

The slightly-more-skilled Arab workers do not bene…t from the decrease in Palestinian labor supply as their least-skilled counterparts; they indeed experience a negative absolute e¤ect in all measured periods (third row)— albeit this is statistically signi…cant only in periods 2, 4, and 5— supporting the hypothesis that the 6-11 Israeli Arab workers might be net complements to Palestinian workers. The net complementarity between Palestinian workers and the more-skilled Israeli Arab workers is puzzling. One would expect a net substitution e¤ect, as the 0-11 schooling group is treated like homogeneous. Notwithstanding, our result suggests that, while such a substitution e¤ect may exist due to similar work duties within this skill group, there is a gross complementarity e¤ect which outweighs the substitution e¤ect. One possible explanation of this outcome is the bilingualism of Israeli-Arab workers. Israeli-Arabs are likely to know Hebrew, beside their native Arabic language, given that this is the (…rst) o¢ cial language of the country, and that it is mandatory at Arab schools starting around the third grade. On the other hand, Arabic is not a compulsory subject at Jewish schools, and is not taught there regularly. Anecdotal evidence, added to the fact that between 0.6%-1.7% and a wage decrease of previous immigrants of 6.7%, due to immigration. The larger e¤ects found here might be reasoned out by the fact that Palestinians do not dilute the e¤ect through demand for goods, by the very nature of their temporary presence in Israel. Additionally, internal migration in Israel is not likely to play a role in di¤using these e¤ects. That said, overreaction of the market in response to a major violent event like the intifada, is not surprising, at least not in the medium run studied here.

27

Arabic is the only o¢ cial language in the Palestinian Authority and its schooling system, however, hint that most Palestinians do not speak Hebrew.20 Bilingual Israeli-Arab workers, therefore, can serve as a link (in management, communication, and transportation) between the Israeli-Jewish employers and the Palestinian-Arab workers.21 The evidence in this study is consistent with the implications of the bilingualism hypothesis, but by no means proves it or renders it unique.22 The fact to be stressed, nevertheless, is that the 6-11 group responds di¤erently (oppositely) from the 0-5 group, although the absolute e¤ects on this group are not always statistically signi…cant. Secondly, the statistically signi…cant e¤ects for the more-skilled group seem to cluster around the immediate run, suggesting that the e¤ects are short-lived. One possible reason for this …nding, given that these are unskilled jobs— most likely matched and performed in a spot-market setting— wages across industries are very likely to converge pretty quickly. It is also possible that foreign workers replace Palestinians within a few 20

“Amidst the damage of the second Intifada, the number of Palestinians who speak, understand, and

read Hebrew has become even more limited,” Maan Network (2007). Articles in newspapers relate to this fact as well (for example, Ben-Zur (2011), Keret (2004), and Menassat (2008)). Likewise, other reports lend support to this proposition [see Altinok (2010)]. 21

“If the Palestinian doesn’t speak Hebrew and a private company is to employ him or her, some-

one in the company needs to speak some Arabic.” Palestine Primer, Economy. accessed October 8, 2011. 22

Other possible explanations of this phenomenon— e.g., trade-based explanation, or the selective with-

drawal hypothesis, were examined and refuted in this context. See Asali (2008).

28

quarters after a drastic fall in their numbers.23 When comparing the results for Jewish and Israeli Arab workers, the evidence in this study is also consistent with the hypothesis that labor market discrimination against Israeli Arab workers intensi…es with the Israeli-Palestinian con‡ict, and thus the negative e¤ect on the young skilled Israeli Arab group (as the positive substitution e¤ects are experienced only by a very small, old, unskilled group of workers), as opposed to the neutral or positive e¤ect on Jewish workers. Miaari et al. (2012) presented evidence showing that the Arab-Jewish (political and economic) relationship in Israel worsened after the outbreak of the intifada; they also argued that this rise in tensions between Israeli Arabs and Jews (and discrimination against Arab workers) constituted the most plausible explanation for the increase in the job separation rate of Israeli Arab workers experienced after the outbreak of the intifada, lending support to our interpretation of the di¤erent estimated reactions (between Arab and Jewish workers). Employment and Robustness Checks I use income data to de…ne the employment status of the worker. The worker is considered to be employed if he reports a positive income, and unemployed otherwise. “Employment,” 23

If the number of foreign workers in the industry is not controlled for, we …nd that the e¤ects are estimated

to last for the whole period. Since the e¤ects in this study, where I do control for the number of foreign workers, are more short-lived, we tend to believe that foreign workers do indeed replace Palestinian workers, in line with this explanation. This is supported in previous studies like Miaari & Sauer (2011) and Friedberg & Sauer (2003).

29

thus de…ned, is then used as the dependent variable in regressions similar to those used earlier to analyze the hourly wage of workers (equations 1 and 2). The industrial a¢ liation of the unemployed, or recent labor force dropouts, pertains to the industry of their last job. Table 6 presents results of the employment regression for the pooled samples of both groups: Israeli Arab and Jewish workers. The total e¤ect of the intifada on the employment of Jewish workers appears with mixed signs, and is not statistically di¤erent from zero. The e¤ects on the employment of Arab workers are negative and mostly statistically signi…cant. However, carrying out this exercise within subgroups of education (0-5 versus 6-11), bears di¤erent results: these are reported in Table 7. Note that in Table 7 I only report the coe¢ cients that we focus on hereafter, namely, those of I

P

H ( 7 ); of I

P ( 4 ); and their sum (

4

+

7 ).

The evidence brought in

Table 7 is consistent with the previous …ndings about the e¤ect on wages. Jewish workers (of 0-11 schooling years) seem not to be the focal point of the e¤ect of the intifada on labor market outcomes. Their employment, as was found with their wages, is not a¤ected by the intifada, not even if the 0-5 group is distinguished from the 6-11 group. Israeli-Arab workers, on the other hand, are a¤ected by the intifada: a large, positive, and statistically signi…cant e¤ect on the least-skilled (0-5) group (0.59 right after the intifada, down to 0.08 a year later); and a statistically signi…cant negative e¤ect on the 6-11 group relative to their least-skilled counterparts. The absolute e¤ect on the 6-11 group is also negative and statistically signi…cant at the 1% level. The …ndings are consistent with a

30

I

Table 6: The E¤ect of the Intifada on the Employment of Native Groups Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 Arab Workers P :018 :050 :163 :201 :263 (:0280)

I

:002

(:0210)

P

N Adj. R2

I

P

I

N Adj. R2

:078

(:0235)

(:0611)

:0152

(:0312)

(:0912)

:189

(:0534)

(:1095)

:223

(:0681)

:103

:186

:107

:059

:012

(:3535)

(:2759)

(:2507)

(:2164)

(:2033)

1,257 .359

2,057 .324

2,839 .326

3,660 .334

4,606 .343

:005

(:1633)

:027 (:0201)

P

(:0248)

:074

Jewish Workers :146

(:0929)

(:1335)

:094

:157

(:0788)

(:1175)

:221

:313

(:1621)

(:1916)

:198

(:1332)

:222

(:1488)

:317

:232

:223

:213

:170

(:2325)

(:1609)

(:1311)

(:1226)

(:1135)

2,083 .168

3,688 .133

5,406 .133

7,153 .146

8,844 .151

NOTE.— * p<10%, ** p<5%, *** p<1%. The dependent variable is a dummy for employment, which takes on the value 1 if the individual has positive reported income and zero otherwise. Standard errors are clustered by industry and reported in parentheses. See notes of Table 3 for more details about the variables and periods.

31

Table 7: The E¤ect of the Intifada on the Employment of Natives, by Skill Group Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 Arab Workers P H :661 :593 :620 :504 :379

I I

P 4

+

7

P

I

H

P 4

+

(:1825)

:593

:503

(:2973)

N Adj. R2 I

(:3114)

7

N Adj. R2

(:0933)

:407

(:1788)

(:0441)

(:1016)

:251

(:0414)

(:1100)

:077

(:0717)

:068

:089

:212

:253

:301

(:0356)

(:0279)

(:0651)

(:1002)

(:1231)

1,257 .362

2,057 .326

2,839 .330 Jewish Workers :076

3,660 .340

4,606 .348

(:0631)

(:0695)

:081

(:0914)

:211

(:1551)

:037

:059

(:1125)

:056

:273

:071

:180

:360

(:1919)

(:2121)

(:1413)

(:1402)

(:1495)

(:1525)

:025

:062

:146

:217

:301

(:0865)

(:1385)

(:1678)

(:1978)

2,083 .171

3,688 .134

5,406 .135

7,153 .147

8,844 .152

NOTE.— * p<10%, ** p<5%, *** p<1%. The dependent variable is a dummy for employment status, which takes on the value 1 if the individual has positive reported income and zero otherwise. Standard errors are clustered by industry and reported in parentheses. See notes of Table 4 for more details.

32

simple supply-demand story, where the 0-5 Israeli-Arab workers are the only close substitutes for Palestinian workers. As a robustness check, de…ning skill along other dimensions, like occupation (de…ning ‘unskilled’as someone whose occupation is “unskilled-workers in industry and agriculture”); or by years of experience (unskilled having less than 25 years of potential experience), whether over samples covering all schooling levels or those limited to the 0-11 schooling group, yielded qualitatively similar results. Namely, the intifada was estimated to have a positive e¤ect on the least skilled Israeli-Arab workers, and a relative and absolute negative e¤ect on the slightly more skilled group; these e¤ects, however, were not always statistically signi…cant: probably the small, narrowly-de…ned samples contributed to this fact, but also these dimensions are by no means equivalent measure of language pro…ciency; hence, it comes as no surprise that there is no perfect match between the results based on the di¤erent de…nitions of skill (schooling versus occupation or experience).24 Finally, I carried out falsi…cation tests in two ways. First, I examined the sample of highly-skilled workers (those with 13 or more years of schooling). Alternatively, I analyzed the same sample of unskilled workers (0-11) but, as a placebo, I rede…ned I (the start of the intifada variable) to be Q3/2000 onwards, that is, one quarter before it actually started. Results of these analyses are provided in Tables 9–11, in the appendix. 24

While the results based on occupation or experience are not identical to those based on schooling years,

they are qualitatively equivalent. I thank the referee for pointing out these other dimensions of skill as a supporting evidence.

33

Panels A and B of Table 9 show the wage and employment results for the highly-skilled (13+) Arab and Jewish workers. As apparent from panel A of the table, in most cases the e¤ects for this group are statistically insigni…cant, both for Israeli Arab and Jewish workers.25 The insigni…cant …gures were mostly negative for Arab workers and positive for Jewish workers. Panel B of the table shows negative e¤ect of the intifada on the employment of highly skilled Arab workers— these e¤ects are statistically signi…cant in periods 2-4. There were no signi…cant employment e¤ects for Jewish workers, despite the much larger samples of this group. The negative, albeit insigni…cant, wage and employment e¤ects for Arab workers are in line with previous …ndings, given that this group is uniformly more educated. Tables 10 and 11 provide placebo analyses for the e¤ect on wages and employment, respectively. Each panel of the tables, shows results from both the short (pooled) and the long functional forms. Here we assume I = 1 for Q3/2000 onwards— i.e., intentionally (mistakenly) assuming the intifada started in Q3/2000, before the actual event which started in Q4/2000.26 Essentially I here represents a “no-event” variable, and thus, if the previous results were driven by the intifada, we expect no signi…cant e¤ect whatsoever of this newly de…ned variable. The fact that the estimated e¤ects on wages and employment, from Tables 10 and 11, are diminished in magnitude and statistically not di¤erent from zero, both for Arab and 25

The estimated e¤ect for Arab workers was negative, and signi…cant at the 10% level in the fourth period;

and that for Jewish workers was positive and signi…cant at the 10% level in the …rst period. 26

In this regard, results are similar regardless of any earlier period we choose for I = 1:

34

Jewish workers, lend support to the hypothesis that the e¤ects estimated earlier should be interpreted as the causal e¤ects of the intifada.

5

Concluding Remarks

The second intifada, starting on September 30, 2000, and resulting in a drastic drop in the number of Palestinian workers in Israel, provides a natural experiment for the study of the e¤ect of immigration on labor market outcomes of natives. Using Israeli income surveys, this study analyzes the e¤ect of this drastic drop in the number of Palestinian workers on the wages and employment of unskilled (with 0-11 years of schooling) Israeli Arab and Jewish workers. When the Israeli-Arab and Jewish unskilled workers are analyzed as a whole group each, the second intifada seems to have no e¤ect on labor market outcomes of Jewish workers, but a slight negative e¤ect on the wage and employment of Arab workers. Studying the e¤ect for subsamples of workers by schooling levels, however, reveals a clear substitution e¤ect between Palestinians and the least-skilled (0-5 schooling years) Israeli Arab workers: as a result of the intifada, this group experiences an increase in their wages and employment. These e¤ects are most intense in the immediate run. The slightly more-skilled (6-11) Israeli Arab workers experience a decrease in their wages and employment, both in absolute values and relative to their least-skilled counterparts. This is a suggestive evidence of the complementarity between the 6-11 Israeli Arab workers and Palestinian workers. 35

The decomposition of Jewish workers by degrees of schooling gives similar results to those from the pooled sample, namely, the intifada is found not to have an e¤ect on their wages or employment. Bilingualism of Israeli Arabs, who speak Arabic and Hebrew, as well as labor market discrimination against them (intensi…ed by the intifada), are two possible explanations consistent with the evidence provided in this study. The e¤ect of bilingualism on labor market outcomes and its share in explaining di¤erences in the e¤ect of immigration among natives are an interesting venue for future research.

6

Appendix

Table 8 below reports the distribution of observations, of workers with strictly positive income, across education groups: the least skilled (0-5) and the slightly more skilled (611), by ethnic group. Table 9 shows the e¤ect of the intifada on the highly-skilled workers, and Tables 10 and 11 present placebo analyses for the e¤ect of the intifada on wages and employment, respectively.

36

Table 8: The Distribution of Workers by Education Level Arab Jewish Quarter 0-5 6-11 0-5 6-11 Q1/1999 5.7 94.3 4.2 95.8 Q2/1999 3.7 96.3 4.8 95.2 Q3/1999 9.2 90.8 4.2 95.8 Q4/1999 10.6 89.4 4.9 95.1 Q1/2000 8.2 91.8 6.1 93.9 Q2/2000 6.9 93.1 6.0 94.0 Q3/2000 7.6 92.4 5.1 94.9 Q4/2000 4.8 95.2 7.9 92.1 Q1/2001 12.6 87.4 5.3 94.7 Q2/2001 10.0 90.0 5.6 94.4 Q3/2001 7.6 92.4 6.3 93.7 Q4/2001 7.3 92.7 4.5 95.5 NOTE.— Source: author’s calculations from the income survey …les. Samples are restricted to 21–64 years old, working males and females, with strictly positive earnings, and with 11 or less years of schooling. “0-5”refers to workers with 0 to 5 years of schooling, and “6-11” to workers with 6 to 11 years of schooling. Reported portions describe the actual (not the weighted) number of observations in the working samples.

37

Table 9: The E¤ect of the Intifada on the Highly-Skilled Workers Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 A. Log Hourly Wage Arab Workers I P :017 :114 :074 :143 :097 (:091)

(:080)

(:062)

(:074)

(:079)

N Adj. R2

245 .279

481 .313

992 .301

1,267 .318

I

:044

P

N Adj. R2

(:024)

(:041)

:019

730 .337 Jewish Workers :015 (:046)

(:047)

(:049)

3,498 .172

6,661 .190

9,983 .187

13,240 .188

16,842 .186

:092

:069

:033

B. Employment I

P

N Adj. R2 I

P

N Adj. R2

:125

Arab Workers :166

(:132)

(:054)

(:033)

(:033)

(:037)

434 .242

761 .255

1,118 .231 Jewish Workers :122

1,512 .231

1,939 .222

:122

:018

:148

:134

:001

:142

(:015)

(:087)

(:123)

(:150)

(:165)

5,189 .112

9,333 .088

13,753 .087

18,101 .088

22,756 .083

NOTE.— * p<10%, ** p<5%, *** p<1%. Main entries are the coe¢ cients of the interaction term I P , from the respective wage and employment regressions. Standard errors are clustered by industry and reported in parentheses. See notes of Tables 3 and 6 for more details.

38

I

Table 10: The E¤ect of the Intifada on Wages: Placebo Analyses Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 Arab Workers :138 :041 :046 :041

P

Adj. R2 I

P

I

H

P

(:0661)

(:0618)

(:0484)

(:0439)

.144

.130

.138

.147

:007

:145

:116

:267

(:1936)

(:2116)

(:1338)

(:1694)

:136

:093

(:2168)

(:1345) 4

+

:143

7

N Adj. R2 I

P

I

P

H

P 4

+

(:047)

(:034)

669 .142

1,001 .130 Jewish Workers :053

1,352 .141

1,736 .149

:010

:010

(:1205)

(:1355)

(:1099)

(:1048)

.147

.153

.157

.154

:177

:068

(:1166)

:112

(:1901) 7

N Adj. R2

:061

(:060)

(:2248)

I

:057

:206

(:1836)

(:077)

:062

Adj. R2

:051

:059

(:1334)

:065

:106

(:1395)

:005

(:0884)

:113

:104

:013

(:1316)

(:1050)

(:0781)

:045

:002

:008

(:138)

(:142)

(:119)

(:110)

1,949 .147

3,002 .154

4,010 .159

5,041 .156

NOTE.— * p<10%, ** p<5%, *** p<1%. The dependent variable is the log of hourly wage. Standard errors are clustered by industry and reported in parentheses. See notes of Tables 3 and 4 for more details.

39

I

Table 11: The E¤ect of the Intifada on Employment: Placebo Analyses Q3/00-Q4/00 Q2/00-Q1/01 Q1/00-Q2/01 Q4/99-Q3/01 Q3/99-Q4/01 Arab Workers :150 :035 :074 :142

P

Adj. R I

2

P

I

H

P 4

+

(:0832)

.322

.314

.318

.324

:137

(:1705)

H

P +

:039

N Adj. R2

:025

(:1833)

:087

:099 (:1827)

:153

(:075)

(:081)

(:088)

2,057 .321

2,839 .315 Jewish Workers :073

3,660 .322

4,606 .328

:149

:239

(:0560)

(:1072)

(:1247)

(:1500)

.127

.120

.130

.134

:422

(:2580)

:458 7

:054 (:1376)

:032

:175

:123

:074

(:1266)

(:0809)

(:0941)

:090

(:2230)

(:3021) 4

:113 (:1285)

(:063)

:034

P

:071 (:1331) (:1975)

:157

7

Adj. R2

I

(:0834)

:019

P

I

(:0734)

(:1216)

N Adj. R2 I

(:0489)

:036

:085

:030

:162

(:1511)

(:1523)

:154

:236

(:063)

(:110)

(:130)

(:154)

3,688 .129

5,406 .123

7,153 .132

8,844 .135

NOTE.— * p<10%, ** p<5%, *** p<1%. The dependent variable is a dummy for employment status, which takes on the value 1 if the individual has positive reported income and zero otherwise. Standard errors are clustered by industry and reported in parentheses. See notes of Tables 6 and 7 for more details.

40

References Altinok, Nadir. 2010.

An Analysis of Schooling Quality Di¤erences In Israel

and Palestinian National Authority During the Second Intifada. EFA Global Monitoring Report 2011.

UNESCO,

Accessed October 2011. Available at

http://unesdoc.unesco.org/images/0019/001907/190773e.pdf. Angrist, Joshua D. 1996. Short-Run Demand for Palestinian Labor. Journal of Labor Economics, 14(3), 425–453. Aranki, Ted N, & Daoud, Yousef. 2010. Competition, Substitution, or Discretion: An Analysis of Palestinian and Foreign Guest Workers in the Israeli Labor Market. Journal of Population Economics, 23, 1275–1300. Asali, Muhammad. 2008. Essays on Discrimination, Immigration, and Poverty. Ph.D. thesis, Columbia University: New York, NY. Asali, Muhammad. 2010. Jewish-Arab Wage Gap: What are the Causes?

Defence and

Peace Economics, 21(4), 367–380. Ben-Zur, Raanan. 2011. Risk.

Don’t Speak Hebrew?

Ynet, March 7, 2011.

Caution:

Your Freedom is at

7 March. Accessed October 2011. Available at

http://www.ynet.co.il/articles/0,7340,L-4038981,00.html. Borjas, George, & Katz, Lawrence. 2007. The Evolution of the Mexican-Born Workforce in 41

the United States. In: Borjas, George (ed), Mexican Immigration to the United States. Cambridge, MA: National Bureau of Economic Research. Borjas, George J. 2003. The Labor Demand Curve is Downward Sloping: Reexamining the Impacts of Immigrationon on the Labor Market. Quarterly Journal of Economics, 118(4), 1335–1374. Borjas, George J, Grogger, Je¤rey, & Hanson, Gordon H. 2010. Immigration and the Economic Status of African-American Men. Economica, 77, 255–282. Card, David. 1990. The Impact of the Mariel Boatlift on the Miami Labor Market. Industrial and Labor Relations Review, 43(2), 245–257. Card, David. 2001. Immigrant In‡ows, Native Out‡ows, and the Local Labor Market Impacts of Higher Immigration. Journal of Labor Economics, 19(1), 22–64. Card, David. 2005. Is the New Immigration Really so Bad?

The Economic Journal,

115(507), F300— -F323. Card, David, & Lewis, Ethan. 2007. The Di¤usion of Mexican Immigrants During the 1990s: Explanations and Impacts. In: Borjas, George (ed), Mexican Immigration to the United States. Cambridge, MA: National Bureau of Economic Research. Cohen-Goldner, Sarit, & Paserman, M. Daniele. 2011. The Dynamic Impact of Immigration

42

on Natives’Labor Market Outcomes: Evidence from Israel. European Economic Review, 55(8), 1027–1045. Friedberg, Rachel M. 2001. The Impact of Mass Migration on the Israeli Labor Market. Quarterly Journal of Economics, 116(4), 1373–1408. Friedberg, Rachel M, & Hunt, Jennifer. 1995. The Impcat of Immigrants on Host Country Wages, Employment, and Growth. Journal of Economic Perspectives, 9(2), 23–44. Friedberg, Rachel M, & Sauer, Robert M. 2003. The E¤ects of Foreign Guest Workers in Israel on the Labor Market Outcomes of Palestinians from the West Bank and Gaza Strip. The Maurice Falk Institute for Economic Research in Israel, Discussion Paper No. 03.08. Hunt, Jennifer. 1992. The Impact of the 1962 Repatriates from Algeria on the French Labor Market. Industrial and Labor Relations Review, 45(3), 556–572. Jaeger, David A., Klor, Esteban F., Miaari, Sami H., & Paserman, M. Daniele. 2012. The Struggle for Palestinian Hearts and Minds: Violence and Public Opinion in the Second Intifada. Journal of Public Economics, 96, 354–368. Keret, Etgar. 2004. Speak Hebrew or Shut Up! The Guardian, August. August 11. Accessed October 2011. Available at http://www.guardian.co.uk/world/2004/aug/11/israel. Klinov, Ruth. 2010. icy.

In:

Palestinian Workers and Foreign Workers:

Decisions and Pol-

Lisk, Moshe (ed), Milestones in the Decisions of the Government 43

of Israel. able

Jerusalem: at

The Van-Leer Institute.

Accessed October 2011. Avail-

http://economics.huji.ac.il/facultye/klinov/Palestinian%20Workers%20-

%20Migrant%20workers.pdf. Lach, Saul. 2007. Immigration and Prices. Journal of Political Economy, 115(4), 548–587. Maan Network. 2007. Hebrew Language Course for Palestinian Journalists. Accessed October 2011. Available at http://www.maannet.org/EnglishSite. Mansour, Hani. 2010. The E¤ects of Labor Supply Shocks on Labor Market Outcomes: Evidence from the Israeli-Palestinian Con‡ict. Labour Economics, 17(6), 930–939. Menassat. 2008. When Israelis speak in Arabic and Palestinians in Hebrew. November. Accessed October 2011. Available at http://www.menassat.com/?q=en/news-articles/5209when-israelis-speak-arabic-and-palestinians-hebrew. Miaari, Sami, & Sauer, Robert. 2011. The Labor Market Costs of Con‡ict: Colsures, Foreign Workers, and Palestinian Employment and Earnings. Review of Economics of the Household, 9(1), 129–148. Miaari, Sami, Zussman, Asaf, & Zussman, Noam. 2012. Ethnic Con‡ict and Job Separations. Journal of Population Economics, 25(2), 419–437. Ottaviano, Gianmarco I P, & Peri, Giovanni. 2012. Rethinking the E¤ect of Immigration on Wages. Journal of the European Economic Association, 10(1), 152–197. 44

Zussman, Asaf, Zussman, Noam, & Nielsen, Morten Orregaard. 2008. Asset Market Perspectives on the Israeli-Palestinian Con‡ict. Economica, 75, 84–115.

45

The Effect of Immigration on Unskilled Native Workers

*International School of Economics, Tbilisi State University, 16 Zandukeli Street, ... taken so far, and suggested that the effect in question is best estimated at the ...... The decomposition of Jewish workers by degrees of schooling gives similar ...

338KB Sizes 2 Downloads 153 Views

Recommend Documents

On the Macroeconomic Effects of Immigration: A VAR ...
Jun 9, 2017 - Immigration shocks, as well as technology shocks are identified through long- .... of the revisions can be directly and solely linked to new information on ... timate of the degree and speed of capital adjustment following an ..... in t

The Effect of Crossflow on Vortex Rings
The trailing column enhances the entrainment significantly because of the high pressure gradient created by deformation of the column upon interacting with crossflow. It is shown that the crossflow reduces the stroke ratio beyond which the trailing c

the impact of young workers on the aggregate ... - Semantic Scholar
years, from various years of the Statistical Abstract of the United States. The following observations are missing: 16–19 year olds in 1995 and 1996; 45–54 year old men in Utah in 1994; and women 65 and over in. Delaware, Idaho, Mississippi, Tenn

the impact of young workers on the aggregate ... - Semantic Scholar
An increase in the share of youth in the working age population of one state or region relative to the rest of the United States causes a sharp reduction in that state's relative unemployment rate and a modest increase in its labor force participatio

Estimating the Impact of Immigration on Output and Technology ...
We study the impact of immigration-induced changes in labor supply within agriculture in the. US during ... workers seems to be occurring via changes in technology, with shifts in the output mix playing a lesser role: ..... such as, for instance, the

The Effect of Crossflow on Vortex Rings
University of Minnesota, Minneapolis, MN, 55414, USA. DNS is performed to study passive scalar mixing in vortex rings in the presence, and ... crossflow x y z wall. Square wave excitation. Figure 1. A Schematic of the problem along with the time hist

Rethinking the Effects of Immigration on Wages
Aug 1, 2006 - The second question is more “micro” ..... the capital-output ratio in the US do not exhibit any trend in the long run (Kaldor, 1961). .... κt.2) One cannot discern any pattern of opposite co-movements (negative correlation) of the 

Estimating the Impact of Immigration on Output and Technology ...
We study the impact of immigration-induced changes in labor supply within agriculture in the. US during ... workers seems to be occurring via changes in technology, with shifts in the output mix playing a lesser role: ..... such as, for instance, the

On the political economy of immigration
Of course the preferences of natives concerning immigration policies will also be ... 0014.2921/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved ... (b) If Re ki have a higher.

on the political economy of immigration and income ...
internet newsletter titled the “Migration News” that reports on worldwide immigration .... movement of labor into the domestic economy, other things equal, will raise the ..... marginal product of capital, and therefore raises their capital incom

On the determinants of workers' and firms' willingness ...
are unnecessary for identification, are rejected by the data, and lead to biased ... more easily employed in other jobs or allocated to different tasks within the ...

On the Effect of Bias Estimation on Coverage Accuracy in ...
Jan 18, 2017 - The pivotal work was done by Hall (1992b), and has been relied upon since. ... error optimal bandwidths and a fully data-driven direct plug-in.

On the Effect of Bias Estimation on Coverage Accuracy in ...
Jan 18, 2017 - degree local polynomial regression, we show that, as with point estimation, coverage error adapts .... collected in a lengthy online supplement.

Effect of native and alien vertebrate frugivores on seed ...
13°W–18°W, located less than 100 km off the Atlantic coast of ..... The rather scarce data, compared with other fru- .... USDA Forest Service of California, USA.

On the determinants of workers' and firms' willingness ...
and future jobs (see e.g. Lynch, 1994; Blanchflower and Lynch, 1994; Booth and. Bryan .... scribed by the predetermined wage model while other interactions are best described ..... via telephone interviews using computer-aided techniques.

The effect of mathematics anxiety on the processing of numerical ...
The effect of mathematics anxiety on the processing of numerical magnitude.pdf. The effect of mathematics anxiety on the processing of numerical magnitude.pdf.

The effect of mathematics anxiety on the processing of numerical ...
The effect of mathematics anxiety on the processing of numerical magnitude.pdf. The effect of mathematics anxiety on the processing of numerical magnitude.pdf.

The effect of ligands on the change of diastereoselectivity ... - Arkivoc
ARKIVOC 2016 (v) 362-375. Page 362. ©ARKAT-USA .... this domain is quite extensive and has vague boundaries, we now focused only on a study of aromatic ...

The Effect of Recombination on the Reconstruction of ...
Jan 25, 2010 - Guan, P., I. A. Doytchinova, C. Zygouri and D. R. Flower,. 2003 MHCPred: a server for quantitative prediction of pep- tide-MHC binding. Nucleic ...

Effect of earthworms on the community structure of ...
Nov 29, 2007 - Murrell et al., 2000). The development and application of suitable molecular tools have expanded our view of bacterial diversity in a wide range ...

The effect of Quinine on Spontan.Rhythmic contrac. of Rabbit Ileal ...
The effect of Quinine on Spontan.Rhythmic contrac. of Rabbit Ileal smoot. musc..pdf. The effect of Quinine on Spontan.Rhythmic contrac. of Rabbit Ileal smoot.

Effect of Torcetrapib on the Progression of Coronary ...
29 Mar 2007 - additional use of these data to understand the mechanisms for adverse cardiovascular outcomes observed in the suspended torcetrapib trial. Methods. Study Design. The Investigation of Lipid Level Management Us- ing Coronary Ultrasound to

Effect of Torcetrapib on the Progression of Coronary ...
Mar 29, 2007 - Pinnacle Health at Harrisburg Hospital, ... of Lipid Level Management to Understand Its Im- ...... College of Cardiology Task Force on Clin-.

An examination of the effect of messages on ...
Feb 9, 2013 - regarding promises rather than testing guilt aversion under double-blind procedures or discriminating among various models of internal motivation. (5) In CD, messages were sent before As made their decisions, and Roll choices were made