Ethnicity in Israel is a complex and multifaceted phenomenon. Both Jews and Arabs are subdivided into ethnic sub-groups, and there are important differences in socioeconomic status among Arab Muslims, Arab Christians, and Arab Druze.

It is commonplace to attribute much of Israel’s domestic tensions to supposed Jewish discrimination against the country’s Arab citizens.[1] Nearly every Israeli Arab nongovernmental organization insists that such discrimination characterizes the Jewish state in general and its labor markets in particular.[2] The Israeli media routinely interview Israeli Arabs (and non-Ashkenazi Jews) who claim to have been victims of discrimination. These allegeations are echoed by Jewish Israeli academics, think tanks, and journalists, especially on the political Left, not to mention the international anti-Israel movement and the boycott, divestment, and sanctions campaign. Indeed, the U.S. Department of State has even joined the growing outcry concerning Israel’s alleged racist discrimination against its Arab citizens.[3]

Of course, in reality, Israel is the only Middle Eastern entity that is not an apartheid regime, and the apartheid slander holds no water whatsoever save in the minds of the Jewish state’s enemies and defamers. Yet discrimination is a scientifically empirical question subject to testing and not a matter of subjective personal opinion. Stripping away the venomous anti-Israel rhetoric, the legitimate question remains whether and how much discrimination really exists in Israel.

Ethnicity in Israel is a complex and multifaceted phenomenon. Both Jews and Arabs are subdivided into ethnic sub-groups, making exploration and analysis of ethnic disparities a complex challenge. In official statistical data on income, Israeli Arabs are treated as a single population group, but this is somewhat misleading. There are important differences in socio-economic status and performance among Arab Christians, Arab Muslims, and Druse. Those sub-categories are in fact amalgams of even smaller divisions. For example, there are interesting differences between “ordinary” Arab Muslims and Bedouins. The Israeli Income Survey sample does not include the Arab population of the “occupied territories,” except for East Jerusalem and the small population of the Golan Heights, both of which are formally annexed to Israel.

Ethnicity among Jews is even more complex. It is commonly measured in Israel for statistical purposes based upon the continent of birth of the person or the person’s father. Jews born in Asia and Africa (or the children of fathers born there) correspond roughly to Sephardic or Mizrahi Jews. Those born in Europe, the United States, or Australia (and their children) correspond roughly to Ashkenazi or Western Jews. These distinctions are imperfect as there are Ashkenazi Jews who come from Asia and Africa (including South Africa and some Egyptian Jews) and Sephardic Jews who come from Europe (including from Greece, Yugoslavia, and Bulgaria). Over time this “continent of birth” criterion for defining ethnicity is losing its validity because of the rapid increase in native-born Israelis who are themselves sons and daughters of native-born Israelis. In addition, the high intermarriage rate among Jews in Israel from different communities is blurring ethnic distinctions.[4]

Before tackling the specific patterns of ethnic inequality and discrimination in Israeli labor markets, it is necessary to dispose of certain myths and superstitions, beginning with the assumption that heterogeneity proves discrimination. It is a common but mistaken belief that, in the absence of discrimination, the numerical representation in any profession or wage range for all groups in a society should be the same as the proportion of that group in the general population. This might be called the false axiom of “natural homogeneity.” Thus if Group A is over-represented in a profession, compared with its weight in the general population, then it must be the beneficiary of discrimination in its favor. If Group B is under-represented, it must be suffering from discrimination against its members. Many then conclude that affirmative action quotas are needed to remedy the problem. This is known as the “disparate impact” standard or pseudo-evidence.[5]

But the axiom of natural homogeneity is completely specious. Nowhere in the real world does fair competition produce homogeneous representation in any market. Indeed, the only way in which such homogeneity can be achieved is through a rigid anti-competitive system of assignments in hiring or admissions by quota, one that suppresses individual interests, skills, culture, economics, family, educational and regional backgrounds, and meritocracy.

The world is full of examples of radical departures from numerical homogeneity in representation that clearly have nothing at all to do with discrimination: Jews around the world are over-represented among those admitted into universities relative to their numbers in the general population even in countries that have official policies of discriminating against Jews. Asian Americans are also over-represented among U.S. college students but not because these colleges discriminate against non-Asians. American blacks are not prominent in sports because of anti-white discrimination. About 60 percent of the medical students in Israel are women, and this is not because the medical schools discriminate against men. Israeli Arabs are grossly over-represented among students in schools of pharmacy, and it is not because these schools discriminate against Jews. Men are enormously over-represented among the prison populations in all countries, and it is not because of gender discrimination. And so on and so forth.

About 60 percent of Israeli medical students are women while Israeli Arabs are over-represented in schools of pharmacy. This is not because these schools discriminate against male Jews. Israeli Arabs own proportionately twice as many cars as Israeli Jews; no one has suggested that this attests to discrimination against Jews.

The fallaciousness of the idea that discrimination is proven by deviation from numerical homogeneity in representation cannot be over-emphasized. It crops up in almost every debate about ethnic or gender discrimination. When feminists, media commentators, and even many academics wish to prove that discrimination exists, their proof usually consists of presenting numbers that show departure from homogeneity. Such figures are selected when they serve the agenda of the commentator or advocate. Yet it turns out that Israeli Arabs own proportionately twice as many cars as Israeli Jews;[6] no one has suggested that this attests to discrimination in Israel against Jews.

In 2013, the Israeli newspaper Haaretz ran an exposé about supposed discrimination against Israeli Arabs by Israeli banks, which quickly became the focus of a parliamentary investigation.[7] The alleged evidence was that Israeli Arabs were paying, on average, higher bank fees than Jews for certain services. But a closer look showed that Arab bank accounts tend to be held in small rural banks with higher per-unit costs and may both be smaller on average and in different sorts of accounts than those held by Jews. For example, Arabs hold fewer long-term provident savings or retirement accounts, in part because the age structure of the Arab population is younger than its Jewish counterpart. All this results in different arrays of fees being charged but has nothing to do with discrimination. However, such an explanation would provide little sensationalist grist for the media or headline-grabbing power for politicians.

If numerical representation and deviation from natural homogeneity add nothing in terms of understanding discrimination, what about analyzing differences in wages and salaries directly? It would seem that if discrimination does indeed exist in a society, the most promising arena to seek it out is the labor market. But here, too, problems exist.

Analysis of possible discrimination as reflected in labor market wages has the advantage of being able to utilize a rich data set, unlike other markets in which discrimination is alleged. It also matters more. Few, including Arab leaders, would care very much if, after controlling for all the other possible explanations, Israeli Arabs were really paying higher bank fees than Jews. But everyone would think it is important if Arabs were the victims of wage discrimination. Having noted this, it still needs to be emphasized that the mere documentation of a disparity in wages between Jews and Arabs does not in and of itself prove anything, much less discrimination.

Consider the following situation: Suppose that it is found that 45-year-old Israeli Jewish software engineers with postgraduate degrees earn several times the wages of 20-year-old Arab youths who never finished high school. Would this datum be evidence of discrimination against Arabs in the labor market?

Of course, 45-year-old engineers in any ethnic population generally earn far more than 20-year-old high school dropouts. Their labor is simply worth more, and the market prices it accordingly. If one controls for education, age, and field of study, it is possible to compare 45-year-old Jewish engineers with Arab engineers, or 20-year-old Jewish with Arab high school dropouts, to see if there are any residual gaps in wages. There could also be other factors not yet taken into account that explain observed residual disparities, for example, disparities between wages in rural/peripheral labor markets and those in metropolitan areas. Any suspected ethnic discrimination is tentative and needs to be assessed in light of many other non-ethnic factors that affect wages.

Special attention needs to be paid to differences in labor force participation rates. Arab women in Israel, especially married Muslim women, have very low participation rates. This means that most employed Arab women are young and not yet married, which in turn generates a considerable gap in earnings levels when compared with Jewish women (and men of all groups). Gender differences in wages must be separated out to understand patterns of ethnic inequality.

It has been demonstrated in other countries that something as innocuous as age structure may often explain a considerable portion of disparities in earnings across ethnic/racial groups. For example, the eminent economist Thomas Sowell has demonstrated that one of the major causes for racial inequality in the United States is age difference, with the black and Hispanic population considerably younger than the white population for a variety of demographic reasons. He pointed out that “Mexican Americans and Puerto Ricans have median ages of less than twenty years while the average Irish American or Italian American is more than thirty years old, and Jewish Americans are over forty.”[8] Since 40-year-olds invariably earn far more than 20-year-olds, a significant portion of earnings disparities among American ethnic groups reflects nothing more than age structure differences.

Age structure also explains part of the earning differences in Israel since Israeli Jews are on average considerably older than Israeli Arabs, particularly Israeli Muslims. It is estimated that the median age of Muslim Israelis is 19 while the median age of Jewish Israelis is 31.[9] (Interestingly, Christian Arabs have an age structure similar to that of Jews, with median age 30, and also have mean earnings very close to those of Jews.) So an age-explained earnings gap similar to that in the United States arises where age explains part of ethnic inequality.

The Israeli Central Bureau of Statistics (CBS) conducts an annual survey of income and wages. It is a large, scientifically-designed, representative survey that covers the entire Israeli population excluding the population in the “occupied territories,” foreign temporary workers, and tourists. The CBS is staffed with professional statisticians of the highest caliber, and its operations are in line with international standards of professionalism and integrity.

Part of the income survey is based on households (N = 14,996) and measures income at the household level from various sources. Another is based on income from salary and other sources for individual earners (N=35,680) aged over 15. A household can have multiple earners. Income measured includes that from salaries, self-employment, capital, pension, alimony, social insurance, governmental support, and other categories.[10] Other variables contained in the survey include age, marital status, schooling, ethnicity, occupation, and location of residence.[11]

What does the income survey show about ethnic inequality in Israel? One can begin to digest the data starting with the raw numbers and measures of earnings, not adjusted for variables such as age and years of schooling. These numbers explain little about actual patterns of income inequality in Israel but offer a starting point for exploration.

In the Israeli “Income Survey of 2011,” the average salary for the entire population of Israeli Arab males was 50.2 percent of the mean for the entire population of Jewish males. Jewish females on average earned salaries that were 61.8 percent of those of Jewish males. Arab females earned only 34 percent of the salaries of Arab males and 28 percent of the salaries of Jewish females,[12] but this was no doubt in part because of part-time employment common among Arab women. Raw household income disparities follow a somewhat different pattern because salaries are only one component of household income. Household income for Arabs was about 55 percent that of Jews. While these raw disparities appear large, they are not unusual when comparing across ethnic populations within countries. The real question remains what is causing them.

There are also disparities in the raw figures among subgroups of Jews, to some extent caused by age structure. The groups with the highest salaries and household incomes are native-born Jews. Those born elsewhere are usually divided between recent immigrants and earlier immigrants. The dividing line for distinguishing recent immigrants is necessarily arbitrary; in the discussion here, the cutoff used is 1990. In the last two decades, the largest group of new immigrants has been from the former Soviet Union. A separate smaller group, about whose economic performance relatively little is known, consists of Jewish immigrants from Ethiopia. These will be separated out here from other immigrants by distinguishing them as recent immigrants born in Africa. This, too, is an imperfect measure, and some Jews from North African countries and from South Africa are probably mixed into this sub-sample definition as well.

Among native-born Israelis, the Ashkenazi males earn 16 percent more than the Mizrahi/Sephardic males. Ashkenazi and Mizrahi females earn exactly the same average salaries, which are about 40 percent lower than for native-born Ashkenazi males. Among foreign-born Jews, Mizrahim earn average salaries 32 percent lower than Ashkenazim for males, and 39 percent lower for females. Women in all population groups earn less than men in the same groups.

So the starting point is a set of seemingly wide disparities in average earnings across Israeli ethnic groups. Jews earn more than Arabs, in fact twice as much on average; women earn less than men; Mizrahim earn less than Ashkenazim. Two additional caveats need to be mentioned. First, these numbers are based on reported salaries. While survey respondents were told the information was confidential and would not be passed on to the tax authorities, it is possible that some of the salary numbers are in fact under-reported. Israel is thought to have a significant underground or unreported economy where cash is earned under the table. For a variety of reasons, including concentrations in occupations in which non-reporting is easier and more common, it is generally believed that non-reporting of income is higher among Arabs than among Jews.

An additional caveat is that disparities across ethnic groups in salaries and in household income are different from disparities in household expenditures. Standards of living are ultimately measured in real consumption rather than in monetary terms, and in Israel, gaps in levels of expenditure among the ethnic groups are considerably smaller than those in salaries or incomes. In addition, intentional under-reporting of income is unlikely to affect reported levels of expenditure, and so these data may be more reliable. The bottom line is that raw inequality among Israeli ethnic groups is considerably smaller when measured in terms of expenditures rather than incomes.

To understand properly the role of ethnicity in explaining disparities in earnings, one needs to take into account other non-ethnic factors that affect earnings, notably gender, age, education (measured in several different ways), marital status, number of persons in household, immigration status (new immigrant vs. not), membership in certain elite professions such as manager or engineer, and geographic variables (residence in one of the large cities, in medium-sized towns, etc.). Statistical estimates of the impact upon earnings by individuals of a variety of ethnic, demographic, and other factors are presented in Table 1 below.

First, after controlling for age, education, and other non-ethnic explanatory variables, is it really the case that Arabs underperform in the Israeli labor market when compared with Jews? The answer is generally, no. It does depend on which definition of earnings is being used.

When estimating only salaries for both men and women together (not shown in the table), Arabs do indeed underperform when compared with Jews. The difference is not very large (approximately 450 shekels a month or a bit over $100), and this is very small when compared to the raw disparity between earnings of Jews and Arabs, seen above as being approximately a 100 percent difference. The disadvantage in salary earnings for Arabs is about the same as that experienced by Jewish new immigrants in Israel.

But salaries are only one component of individual earnings. Salaries are what employees receive from employers while “all individual earnings” include things such as self-employed income by artisans or shop-owners or owners of proprietary establishments. Such self-employed and proprietary income is probably more common among Arabs than Jews, the latter being more likely to be salaried employees. The numbers in the table here show the results when analyzing all individual earnings from all sources, including such non-salary sources. When controlling for age, schooling, and the other non-ethnic factors, Israeli Arabs outperform Jews on average, earning more than Jews of similar age and schooling levels. Indeed, on average Arabs earn more than both Ashkenazi and Sephardic Jews, about 9 percent higher, other things being equal.

The fact that the labor market disadvantage of Israeli Arabs disappears entirely when total individual earnings (as opposed to salaries alone) are analyzed may be because many Arabs are self-employed.[13] In any case, it turns out that not only do non-ethnic factors explain the bulk of the raw disparity in earnings between Israeli Arabs and Jews, but in many cases they explain more than the total disparity. In the case of total individual earnings income, they explain more than 100 percent of the raw disparity (meaning that, after controlling for explanatory variables, Arabs actually outperform Jews).

The picture becomes clearer when men and women earners are analyzed separately. This has the advantage of removing gender differences in labor force participation rates from the analysis of the role of ethnicity. The gap in earnings for Arab women compared with Jewish women is quite small when controlling for other variables; it is only about 2 percent to the advantage of Jews. But for males, Arabs are at a 10 percent advantage over Jews in total individual earnings. Again, Arabs outperform Jews.

It is important to distinguish between salaries and earnings. For example, Israeli Arab males may make on average 50 percent less than Israeli Jewish males in salary, but in earnings (which include income sources such as self-employment), they out-perform Israeli Jews by approximately 9 percent on average.

Arabs also have a disadvantage compared with Jews when it comes to total household earnings (not shown in the table), as opposed to total individual earnings. But the wider gap at the household earnings level is due to factors outside the labor market. Jews have higher savings rates than Arabs, and thus have higher levels of household capital income.[14] Jews are also older and so receive on average higher amounts of retirement income. These disparities in non-labor income at the level of households largely reflect differences between Jews and Arabs in savings behavior and household composition and cannot be attributed to labor market discrimination.

What about disparities across ethnic sub-groups of Israeli Jews? The first notable pattern is this: The main group that over-performs compared with others is native-born Israeli Jews or sabras. Being born in the country confers a distinct earnings advantage in Israel as it does in most other countries. There is a modest advantage in income, about 8 percent for men and 2 percent for women, for those who are native-born Israeli Jews, compared with those who are foreign-born. And this is true for both Ashkenazi and Mizrahi Jews.

When controlling for other non-ethnic factors, Ashkenazim have a small advantage over Mizrahim among men, about 2 percent for total individual income and 4 percent for salary alone, much smaller than the gap in the raw earnings numbers, and much smaller than the premium enjoyed by native-born Jews. For women, Ashkenazim slightly underperform Mizrahim. More generally, because of the advantage of being a sabra, a native-born Mizrahi Jew would generally outperform a non-native Ashkenazi Jew, other things being equal. When men and women are separated in the analysis of earnings, the “natives” retain an earnings advantage among both genders. Mizrahi Jewish women are outperforming the Ashkenazi Jewish women.

Recent immigrants in Israel are at an earnings disadvantage compared to the other population groups. Controlling for age, education, and the other non-ethnic factors, recent immigrants earn about 5.5 percent less in total individual earnings while for salary alone (not shown in the table), they earn 10-14 percent less than other Israelis. The earnings disadvantage is larger for men than for women. Interestingly, immigrants from Africa (mainly Ethiopians) do not suffer from any special earnings disadvantage as compared with the earnings levels of all recent immigrants. All immigrants are at a modest disadvantage in the labor market, but Ethiopians no more so than non-Ethiopian immigrants. When men and women are analyzed separately, Ethiopians slightly outperform the other immigrants.

Economists like to describe schooling and degrees as “human capital,” and it is possible to measure the returns or market rewards to this capital using statistical methodologies. One issue that has frequently been debated in Israel is whether educated Arabs are at a market disadvantage, since—because of discrimination—they are less capable of capitalizing upon their educational achievements.[15]

Once again, the presumption of discrimination does not survive empirical statistical analysis. The truth is quite the opposite: The return on schooling for Israeli Arabs is generally considerably higher than it is for Israeli Jews. In almost every estimate, using different measures of schooling and of earnings, the return on education appears to be higher for Arabs after controlling statistically for other variables.[16] This is true both for salaries and for all individual earnings. Since the reward for educational achievement is, if anything, higher for Arabs than for Jews, this rules out the claim of systematic discrimination against Arabs who accumulate human capital and capitalize upon it in the labor market.

The return on schooling is not the same, however, as the reward for membership in elite professions. Arabs, like Jews, who are members of managerial or other professional groups (lawyers, doctors, engineers, etc.) enjoy a significant earnings advantage over those who are not members of these groups. The bonus or premium for Arabs, however, is lower than that for Jews. Discrimination cannot be ruled out as a causal factor here although other factors unrelated to discrimination could also explain these disparities, including differences in distribution among professions within the broader elite professional categories.

The most surprising conclusion from the econometric analysis of ethnic earnings disparities in Israel is how many of the stereotypical characterizations of Israel turn out to be false. Ethnicity in Israel simply does not play a large role in the labor market, in contrast with gender or schooling.

While it is widely presumed that the Arab minority underperforms in the labor market of the Jewish state, either because of discrimination or other structural or cultural disadvantages, this turns out not to be so. That accusation is central to the claim that Israel is some sort of apartheid regime. While the raw mean earnings of Arabs are considerably lower than those of Jews, the two populations differ in many significant ways, including age and schooling, and little can be concluded from this raw comparison on its own. When education, age, marital status, geographic location, and professional group membership are taken into account, Arab-Jewish earnings disparities all but disappear, and in some cases, they even invert, so that the Arabs outperform the Jews. This is particularly true of male earners.If the data fail to show a clear pattern of Arab underperformance in earnings compared with Jews with similar levels of schooling, the stereotype of Ashkenazi Jews outperforming Mizrahi or Sephardic Jews appears just as inaccurate. Once education and the other explanatory variables are controlled, there is very little difference between Ashkenazim and Mizrahim earnings, and in a few cases, particularly for women, Mizrahim outperform Ashkenazi women. The Ashkenazi-Mizrahi distinction certainly appears to be less important in explaining earnings differences than the distinction between native-born Jews and foreign-born Jews or recent immigrants. Here again, there are differences between men and women. Ashkenazi women slightly underperform Mizrahi women, other things being equal, while Ashkenazi men slightly outperform compared with Mizrahi men. The bottom line is that the data do not support the presumption that Mizrahim are systematically disadvantaged in Israeli labor markets.

While new immigrants underperform relative to other Jewish Israelis, other things being equal, Ethiopians do not appear to suffer from any special earnings disadvantage compared with other immigrants. If Ethiopian immigrants earn low levels of salary, it is because they have low levels of schooling. But given their level of schooling, they earn the same on average as immigrants from Russia, South Africa, and Argentina. When estimating total individual income separately for men or for women, the Ethiopians even slightly outperform the other immigrants.

In spite of what statistical analyses have to show, the subject of discrimination in Israel continues to fill the media, which seem to be obsessed with it even while refusing to examine actual data. For example, in the summer of 2013, a television documentary on Israel’s Channel Ten, produced by popular journalist Amnon Levy, triggered considerable media debate inside Israel. It claimed to have investigated and discovered that anti-Mizrahi discrimination is as bad as it had been back in the early decades of Israeli independence.[17] Real data show otherwise.

The problem is not just in the media. The academic careers of many in Israel, particularly in sociology, have been constructed entirely upon unsubstantiated allegations of Israeli racism. Israeli sociologists in general tend to accept at face value the notion that any documented disparity in earnings or numerical representation between Israeli Jews and Arabs must be due to discrimination.[18] Perhaps the most notorious example is that of Yehouda Shenhav, a sociologist at Tel Aviv University. Shenhav is father of the notion that “Oriental Jews” are in fact “Arabs of the Mosaic faith,” and together with Arabs, share a victimhood imposed upon them by racist Ashkenazi Zionists.[19] Shenhav and those of similar ideological orientation operate the Mizrahi Democratic Rainbow, dedicated to liberating “Oriental Jews” from Ashkenazi bigotry and capitalism.[20]

In Israel’s media, it is considered common knowledge that Arabs, Mizrahim, and Ethiopians are victims of harsh discrimination.[21] The accusations of apartheid may be malicious, disingenuous, and over-the-top—or so most Israeli commentators and sociologists would agree—but the presumption of an underlying widespread pattern of discrimination is, to their minds, undeniable. The extent to which some in Israel go to manufacture evidence of discrimination can be awe-inspiring. For example, the ordinarily prestigious Israel Democracy Institute (IDI), a left-wing think tank, published a study in May 2013 that claimed to have discovered unambiguous proof of widespread discrimination in Israel against Arabs.[22] Composed by IDI legal staffer Tanya Steiner under the supervision of Hebrew University professor Mordechai Kremnitzer, the study’s evidence was the number of complaints about discrimination submitted to the Israeli Commission on Equal Opportunities in Employment. Yet while numerous complaints from women reached the commission, only 3 percent of the complaints it received were from Israeli Arabs, who represent about 18 percent of the labor force. Of these, only three of the complaints received in the entire 2011 year by the commission about alleged anti-Arab discrimination were deemed worthy of investigation. So instead of concluding that the evidence points to an absence of discrimination, the IDI’s conclusion was that it all proves how badly discriminated Israeli Arabs are in Israel; after all, they are so victimized that they do not even file complaints about discrimination.

There is no evidence that points to ethnic discrimination against Israeli Arabs or Mizrahi Jews in Israeli labor markets. Recent immigrants appear to be the one group in the country at an earnings disadvantage. But it would be difficult to make a case that even their disadvantage is due to discrimination since immigrants in all societies are at a competitive disadvantage compared with natives.

There could be other groups in Israeli society that are victims of discrimination, but the data are not available in a form that allows for investigation. In particular, a plausible case for such discrimination may be that against ultra-Orthodox Jews. Gender discrimination also cannot be ruled out, but that is a separate and difficult methodological question beyond the scope of the discussion here.

The nearly complete absence of evidence of ethnic discrimination in Israeli labor markets does not, of course, preclude its existence in other markets or aspects of society. As was shown here, Arabs earn a higher return on education than Jews. But this does not rule out possible discrimination against Arabs in admissions to universities and colleges. It should be noted, however, that Israeli universities routinely implement affirmative action preferences in favor of Arabs and sometimes in favor of Mizrahim (and women).[23] The only other documented university discrimination is that which grants some preferences to army veterans, a practice found in most countries.

There have also been allegations that Israel discriminates in its fiscal allocations and revenue sharing where Arab towns and villages are underfunded. But an empirical analysis of the question found just the opposite; if anything, the Arab local authorities were being over-funded.[24] Evidence regarding other alleged forms of discrimination by Israel tends to be just as skimpy. Some accusations are based upon Israel’s granting automatic citizenship to Jews under its “Law of Return.” But such citizenship entitlements are not unusual in the world and can be found in many other countries, such as Armenia, Greece, Bulgaria and Romania, and are guaranteed under the United Nations Universal Declaration of Human Rights and the International Covenant on Civil and Political Rights.[25] Another indictment of Israel concerns the discriminatory nature of its military conscription. Jews and Druse are conscripted into the Israeli military while Arabs may volunteer for service but are not conscripted. Again, this practice may indeed constitute discrimination but that discrimination is against Jews, not against Arabs.

None of this proves that discrimination never exists in Israel against Arabs, against Mizrahi Jews, or anyone else. But the very fact that empirical evidence of discrimination is so hard to discern or observe must itself serve as an important warning indicator about its magnitude or lack thereof.

[1] See, for example, Ayal Kimhi, “Jewish Households, Arab Households, and Income Inequality in Rural Israel: Ramifications for the Israeli-Arab Conflict,” Defence and Peace Economics, Aug. 2010, pp. 381-94.

[2] For example, “The Inequality Report: The Palestinian Arab Minority in Israel,” Adalah: The Legal Center for Arab Minority Rights in Israel, Haifa, Mar. 2011.

[3] Digital Journal (Toronto), May 1, 2013.

[4] Yinon Cohen, Yitzhak Haberfeld and Tali Kristal, “Ethnicity and Mixed Ethnicity: Educational Gaps among Israeli-born Jews,” Ethnic and Racial Studies, Sept. 1, 2007, pp. 896-917.

[5] Knesset member Zehava Galon of the leftist Meretz party recently introduced a bill that would require all proposals of new legislation in Israel to contain estimates of disparate impact. See Haaretz (Tel Aviv), Apr. 17, 2014.

[6] Ibid., Sept. 12, 2012.

[7] Ibid., June 23,; The Jerusalem Post, June 8, 2013.

[8] Thomas Sowell, “The American Mosaic,” Ethnic America: A History (New York: Basic Books, 1981).

[9] “Projections of population in Israel for 2010–2025, by sex, age and population group,” Israel Central Bureau of Statistics (Jerusalem and Tel Aviv); “The Arab Population in Israel,” idem, p. 2.

[10] “Economic Characteristics,” Israel Central Bureau of Statistics, accessed Dec. 19, 2013.

[11] “Income Survey, 2010,” Israel Central Bureau of Statistics, Feb. 2012. Only people earning at least 100 NIS per month in salary are counted in the analysis below, with the others presumed to be absent from the labor force.

[12] Percentages computed by author from data found here: “Income of Individuals (Income survey),” Israel Central Bureau of Statistics, Table 25.

[13] Yossi Shavit and Ephraim Yuchtman-Yaar, “Ethnicity, Education, and Other Determinants of Self-Employment in Israel,” International Journal of Sociology, Spring, 2001, pp. 59-91.

[14] Haaretz, June 23, 2013.

[15] See, for example, “Israel Must End Discrimination against Arab College Graduates,” Haaretz, June 15, 2012.

[16] Pnina O. Plaut and Steven E. Plaut, “Income Disparities by Ethnicity in Israel,” Israel Affairs, forthcoming.

[17] Panim Amitiyot: Pirakim Milayim, Aug. 22, 2013, Nana 10 web site.

[18] See, for example, Noah Lewin-Epstein and Moshe Semyonov, Stratification in Israel: Class, Ethnicity, and Gender (New Brunswick, N.J.: Transaction Books, 2003), pp. 175-281; idem, “Local labor markets, ethnic segregation, and income inequality,” Social Forces, June 1992, pp. 1101–19; Sammy Smooha and Yohanan Peres, “The Dynamics of Ethnic Inequalities: The Case of Israel,” in Ernest Krausz, ed., Studies of Israeli Society (New Brunswick, N.J.: Transaction Books, 1980), vol. no. 1.

[19] Yehouda Shenhav, “Spineless Bookkeeping: The Use of Mizrahi Jews as Pawns against Palestinian Refugees,” +972 e-magazine (Israel), Sept. 25, 2012.

[20] “Hakeshet Hademocratit Hamizrahit,” web site, accessed Dec. 10, 2013.

[21] For example, Yitzhak Laor, “The Glorious State of Israel and Its Anti-Arab Discrimination,” Ha’aretz, Apr. 15, 2013.

[22] Talya Steiner, Combating Discrimination against Arabs in the Israeli Workforce, Policy Paper No. 97 (Jerusalem: Israel Democracy Institute, 2003).

[23] Haaretz, Nov. 19, 2009; John Rosenberg, “Affirmative Action … In Israel,” Discriminations Blog, Sept. 3, 2002; Noga Dagan-Buzaglo, “Non-discriminatory hiring practices in Israel towards Arab Citizens, Ethiopian Israelis and new immigrants from Bukhara and the Caucasus,” Adva Center, Tel Aviv, Nov. 2008.

[24] Tal Shahor, “Fiscal Allotment Policy vis á vis Minorities: An Empirical Measurement of the Way in Which Israel’s Majority Government Makes Its Fiscal Allotments to the Arab Minority,” Metodološki zvezki (Ljubljana, Slovenia), no. 1, 2010, pp. 73-93; Efraim Karsh, “Israel’s Arabs: Deprived or Radicalized?” Israel Affairs, Jan. 2013, pp. 1-19.

[25] Universal Declaration of Human Rights, United Nations General Assembly, New York, Dec. 10, 1948, art. 14.

[26] The effects of isolated changes in individual factors while holding all other factors constant. The “default” or base case upon which the ethnic increments are computed is for “Foreign-born Mizrahi Jews.” The figures in the table should be taken as the best estimate for changes in earnings caused by isolated changes in each individual explanatory factor (ethnicity, gender, and so on) while holding all other factors constant. This shows the isolated effect for Arabs, for example, on earnings while holding schooling, age, and other factors constant. The schooling variable is measured differently for the men-only column (where the effects of achieving degrees are estimated) than for the women-only column (where the effect of an additional year of schooling is estimated). The estimates allow us to see the “clean” effects or impacts of ethnicity and other factors upon earnings in Israel because these effects are statistically isolated from the many intermingled effects of the other variables. Estimates taken from regression analysis equations that are elaborated and appear in full in Pnina O. Plaut and Steven E. Plaut, “Income Disparities by Ethnicity in Israel,” Israel Affairs, forthcoming.