TRENDS IN IMMIGRANT EARNINGS IN THE 1990s



Preliminary: Please

do not circulate.

THE IMMIGRANT EARNINGS TURNAROUND OF THE 1990s

George J. Borjas

Harvard University and NBER

Rachel M. Friedberg

Brown University and NBER

March 29, 2006

Abstract

This paper uses the 1960-2000 PUMS to study changes over time in the labor market performance of immigrants in the United States. While data from 1960-1990 show a continuous decline in the earnings of new immigrants, the trend reversed in the 1990s, with newcomers doing as well in 2000, relative to natives, as they had 20 years earlier. This improvement in immigrant performance is not explained by changes in origin-country composition, educational attainment or state of residence. Changes in labor market conditions, including changes in the wage structure which could differentially impact recent arrivals, also cannot account for it. Rather, the upturn appears to have been caused in part by a shift in immigration policy toward high-skill workers matched with jobs, and a shift of Mexicans away from agricultural labor. The evidence is also consistent with an improvement in immigrant quality within certain origin countries.

The authors thank Nancy Qian and David Weil for useful comments, and Ying Pan and Sheetal Sekhri for excellent research assistance. The authors’ e-mail addresses are gborjas@harvard.edu and rachel_friedberg@brown.edu.

I. INTRODUCTION

Immigration is an increasingly important force of demographic and economic change in the United States. Approximately one million legal immigrants are admitted to the country each year. Nearly half of population growth in the United States is due to immigration, and since 1970, the foreign-born share of the U.S. population has more than doubled, to 12%.

An extensive literature has examined immigrants’ performance in the labor market, including their earnings upon entry and their subsequent assimilation toward the earnings of native-born workers with similar observable characteristics (see Borjas, 1999, and LaLonde and Topel, 1997, for surveys). An important finding of this literature is that, over the period 1960-1990, there was a continuous decline in the entry wages of new immigrants. This is true both of raw earnings and of earnings conditional upon characteristics such as education and experience

A decline in immigrant labor market performance presents potentially troubling prospects. The skill composition of the immigrant population—and, particularly, how the skills of immigrant workers compare to those of native workers—is the key determinant of the economic impact of immigration on the United States. First, it determines which native workers are more likely to feel an adverse impact of immigration on their labor market opportunities. As closer substitutes in the labor market, less-skilled native workers are more vulnerable to less-skilled immigration. Second, skilled immigrants may assimilate more quickly. They may be more adept at acquiring the country-specific human capital necessary for economic success, with consequences for their fiscal as well as labor market impact. Finally, the relative skills of immigrants determine the economic benefits from immigration. The United States benefits from international trade because it can import goods that are not available or are too expensive to produce in the domestic market. Similarly, the country benefits from immigration because it can import workers with scarce qualifications and abilities.

In this paper, we use the microdata of the 1960-2000 U.S. Censuses to study trends in the entry wages of new immigrants through the 1990s. Surprisingly, we find that the downward trend of the previous three decades has reversed itself, with new immigrants earning as much in 2000, relative to natives, as they did twenty years earlier. This uptick in immigrant earnings is present in raw wages, as well as in wages that have been corrected for differences across arrival cohorts in observed characteristics.

We explore the reasons for this recent uptick in immigrant performance. We test whether the forces that could explain the previous decrease in immigrant earnings—such as origin-country composition, educational attainment, and changes in labor market conditions which may differentially impact recent arrivals, including changes in the wage structure—can also account for the subsequent reversal. Finding this not to be the case, we then go on to investigate potential alternative explanations, including changes in immigration policy (whom the U.S. admits) and changes in the opportunities presented by the labor market (whom the U.S. attracts). We also examine the issue of attenuation bias caused by an increase over time in the extent of earnings imputation in the PUMS.

We find that part of the turnaround in the relative earnings of new arrivals in 2000 can be attributed to a specific change in immigration policy, namely the H-1B temporary visa program for high-skill workers, who earn more, relative to natives, than did earlier cohorts of high-skill foreign-born workers. Another part is due to an improvement in the relative entry wages of new Mexicans, associated with their shift away from agricultural labor. We explore how factors such as changes in the pattern of settlement and selectivity may also have played a role in the observed trend in the relative earnings of all new immigrants.

We conclude by considering the implications of our findings for future patterns of foreign-born labor market performance. Was this uptick a one-time change, or is it likely to extend into a full reversal of the previous decline?

II. DATA AND EARNINGS PATTERNS

The data we use in the analysis are drawn from the Public Use Microdata of the 1960-2000 U.S. Censuses. Citizens by birth are defined to be “natives.” Non-citizens and naturalized citizens are defined to be “immigrants.” We use a 1% sample of immigrants in 1960-1970, a 5% sample of immigrants in 1980-2000, and a 1% sample of natives in all years. The sample is restricted to men aged 25-64 who are employed in the civilian sector.[1] “New immigrants” are defined as those who arrived in the five years prior to the respective Census.

Figure One plots the raw hourly earnings of new immigrants, relative to natives, in each Census year.[2] The vertical axis measures the entry wage of new arrivals, relative to the native benchmark. So for example, in 1960, immigrants who had come to the U.S. in 1955-59 earned 11% less than natives, on average. In 1970, new immigrants who had arrived in 1965-69 earned 17% less than natives. By 1990, the entry wage gap between new immigrants and natives had grown to -0.378 log points, or 31%. From 1960 to 1990, there was a continuous downward trend in the relative earnings of new arrivals.

The key point on which this paper centers is the fact that for the most recent arrival cohort—the one which came to the U.S. in 1995-99-- the trend reversed. This increase in the relative earnings of new arrivals marks a return to the level last seen twenty years earlier, in 1980. Not correcting for differences in other factors, such as age or education, new immigrants today earn 27% less than natives upon arrival.

A reason to focus on the relative earnings of immigrants is to standardize them against those of natives, whose earnings fluctuations capture general labor market conditions in the United States. However, the relative earnings of new immigrants can rise because the absolute earnings of immigrants rise, or because the absolute earnings of natives falls, and these two scenarios have very different interpretations. Figure Two plots the raw earnings of recent arrivals and natives separately. This figure shows that the change in the earnings gap between the two groups was caused by a rise in the earnings of the new arrivals themselves, rather than to a shifting native benchmark.

Finally, to see whether this trend continued beyond 2000, we conducted the same analysis on the March CPS from 1994-2003. Figure Three tracks the relative earnings of immigrants who arrived in the five years prior to each CPS survey, and compares it to the results from the PUMS. First, the wage disadvantage of new immigrants appears larger in the CPS than in the PUMS.[3] Second, though the results for 2000-2003 alone indicate a downward trend, inspection of the general pattern over the entire period reveals it is probably too noisy to be relied upon as an indicator of short-run trends.

So what caused the increase in the entry wages of immigrants in the late 1990s? The following sections consider some possible explanations.

III. OLD EXPLANATIONS

A. COMPOSITION EFFECTS

An obvious explanation for the rise in the relative earnings of new arrivals in the late 1990s could be that they were more skilled, or had more of other attributes associated with higher earnings. Since changes in the origin-country mix of immigration can explain most of the previous trend, perhaps changes in measurable characteristics can account for the 1990s reversal as well.

Table One examines changes over time in the characteristics of new arrivals. The left column describes the cohort that came to the U.S. in 1985-89 as observed in the 1990 Census, and the right column describes the cohort that arrived in 1995-99 as observed in the 2000 Census. Both cohorts are therefore measured during their first five years in the country. Over the course of the decade, there was an increase in the proportion of immigrants coming from Mexico and India, and a decrease in the proportion coming from Southeast and East Asia. The level of education rose, with a reduction in the share with less than a 5th grade education and a rise in the share holding Bachelors and Masters degrees. This rise in education was accompanied by an increase in the share working in professional/technical occupations. Far fewer settled in California and New York, and more in Texas.

Borjas (1985) shows that most of the decline in cohort quality from 1960-1980 can be explained by a shift in the origin-country composition of immigration to the United States. Following the 1965 Amendments to the Immigration and Nationality Act, fewer immigrants originated in Europe, with the majority coming instead from developing countries, particularly Latin American and Asia. Immigrants from these countries tend to be less skilled and to do worse in the U.S. labor market than other immigrants. The emphasis on family ties in the extension of new entry visas magnified this shift in origin-country mix over time.

Given the importance of country composition in explaining the previous decline in the relative earnings of new immigrants, it is natural to ask whether the late 1990s uptick can be similarly attributed to a shifting the origin mix. However, as Table One shows, with the biggest change in origin composition from the 1980s to the 1990s being a rise in the share of immigrants coming from Mexico, this cannot be the explanation. Mexicans earn less than other immigrants on average, so the increase in the number of Mexicans only makes the uptick even more puzzling.

Table Two explores whether the rise in immigrants’ entry wage can be explained by other compositional changes. The sample is comprised of the natives and new immigrants observed in the pooled 1990 and 2000 PUMS. The dependent variable is log hourly earnings. The first specification regresses log earnings on a dummy variable for immigrant status (1 for immigrants who arrived in the previous 5 years, 0 for natives), a dummy variable for the later period (1 for 2000, 0 for 1990), and the interaction of those two, “immigrant*2000.” The coefficient on this last variable measures the increase in the relative (to natives) earnings of new arrivals in 2000, compared to that of new arrivals in 1990. The coefficient of .064 means that the 1995-99 cohort had an entry wage 6.4% higher than the entry wage of the 1985-89 cohort. This is what we will refer to as the “uptick” in the earnings of recent arrivals in the 1990s.

The rest of the specifications in Table Two analyze whether the uptick goes away when we correct for other factors that influence earnings. For example, since the level of education of new arrivals rose in the 1990s, we would expect the uptick to be smaller once this is taken into account. The second and third regressions in Table Two correct for education using dummy variables for: high school graduate, some college, and college graduate (the omitted group is high school dropouts). As expected, the coefficient on the uptick variable falls, from .064 to .060, and to .056 when the return to education is allowed to vary over time and with nativity. However, correcting for age (columns 4 and 5) raises the measured uptick to over .08.[4] The estimated uptick is also larger when we correct for 5-digit country of origin, as in column 6 (coefficient of .100). This is not surprising, given the increase in immigration from Mexico. Correcting for changes in the pattern of settlement of new immigrants, using state fixed effects in column 7, also raises the coefficient (to .104).

Correcting for all of these factors together, seen in the final column of Table Two, raises the size of the estimated uptick from the 6.4% found for raw earnings to 12.7% for residual earnings. Clearly, the increase in the relative earnings of new arrivals cannot be attributed to straightforward changes in their composition, in terms of the observable characteristics of origin country, education, age, or state of residence.

B. ECONOMIC ENVIRONMENT

When immigrants and natives occupy different positions in the labor market, changes in general labor market conditions can have an effect on the outcomes of immigrants, compared to natives, even when there have been no changes in the productive characteristics of either group. Having found that the rise in the relative earnings of new immigrants cannot be explained by the broad changes in their observed characteristics, we now explore a second possibility, namely, that it was brought about by changes in U.S. labor market conditions which might differentially affect immigrants and natives.

Labor Market Conditions

Figure Four shows a time-series of the rate of real wage growth and the unemployment rate from 1980 to the present. Real wages were falling in the late 1980s. By contrast, real wages grew rapidly in the late 1990s. Similarly, though the unemployment rate was falling both periods, the average level in the late 1980s was 6.1%, while the average in the late 1990s was 4.8%. In 2000, the unemployment rate fell to 4%, the lowest level in 30 years.

One potential explanation for the rise in the relative earnings of recent immigrants in 2000 is that the tight U.S. labor market of the late 1990s somehow benefited immigrants relative to natives. If this were the case, we might expect earlier cohorts to also have done particularly well in 2000.

Table Three presents the relative earnings of each arrival-cohort in each year, 1970-2000. This table can be used to compare the earnings of the same cohort over time, or of different cohorts at the same stage of years since arrival. The top panel of Table Four shows that, although the earnings of the most recent cohort (0-5 years since migration) rose between 1990 and 2000, there was not a similar increase in earnings for the second most recent cohort (5-10 years since migration), and earlier cohorts (10-15 and 15-20 years since migration) actually experienced a substantial decline in relative earnings of about 10%.

Figure Five shows the longer-term pattern of cohort effects 1960-2000. The figure shows a uniform decline in immigrants’ relative earnings, across cohorts and across years through 1990. Seen in this larger context, the uptick in 2000 for the most recent arrivals (0-5 years) is even more striking, and the disappearance of the decline for the next most recent arrivals (5-10 years) is also noteworthy.

The bottom panel of Table Four asks whether assimilation rates, i.e., immigrants’ rates of relative earnings growth over time, were faster in the 1990s than in previous decades. The earnings path of a given arrival cohort can be seen by reading across a row in Table Three. These earnings paths are also plotted in Figure Six. We can compare the earnings growth of the cohort that arrived in 1975-79 during its first decade in the U.S. (1980-1990) to that of the cohort that arrived in 1985-89 during its first decade here (1990-2000). In fact, the first-decade assimilation rate was higher in the 1980s than in the 1990s (12.8% and 9.6%, respectively). Though the second-decade assimilation rate was slightly higher— 11.5% in the 1980s and 12.3% in the 1990s— this difference is within one standard error of the estimated coefficients, and in any case, is not of the same magnitude as the uptick for new arrivals.

In sum, based on comparisons with earlier arrival cohorts, it does not seem that the tight labor market conditions of the late 1990s benefited all immigrants relative to natives, and that this can explain the upturn in the earnings of new immigrants.

The Wage Structure

Previous work has shown that the wage gap between immigrants and natives is influenced by general changes in the structure of wages (see LaLonde and Topel, 1991; Lubotsky, 2001; and Butcher and DiNardo, 2002). Since the average immigrant falls in the lower tail of the native wage distribution, increases in the dispersion of the distribution will result in a drop in the earnings of the average immigrant, relative to the average native, and therefore a rise in the absolute size of the immigrant-native wage gap. This previous work argues that increased wage inequality in the United States since 1970 has been largely responsible for the observed decline in the relative wages of (new) immigrants.

Wage inequality in the United States continued to increase in the 1990s, this time in an asymmetric way. The upper end of the wage distribution (as measured by the 90%-50% differential) widened as rapidly as it had in the 1980s, while the lower end (the 50%-10% differential) remained fairly stable or even narrowed somewhat (Autor, Katz, Kearney, 2005a, 2006).

For the average new immigrant, situated in the bottom half of the native wage distribution, compression in lower-tail inequality could potentially imply a rise in entry wages, relative to the average native. In 1990, the average log wage of new immigrants was 2.131. This fell at the 25th percentile of the 1990 native wage distribution. Between 1990 and 2000, the wage differential between natives at the 25th and 50th percentiles shrank by 1.0%. However, the differential between those at the 25th percentile and the mean actually grew by 1.4%, so that changes in the wage structure cannot explain the observed reduction in the earnings gap between the average new immigrant and the average native in the 1990s.[5]

Quantiles

Figures 7A and 7B show the wage distributions in 1990 and 2000 for natives and new immigrants, respectively. Figure 7C shows the wage distribution by nativity in 1990. New immigrants earn less than natives on average, and the shapes of the distributions are rather different. Figure 7D shows the wage distribution by nativity in 2000. The average new immigrant went from earning at the 25th percentile of the 1990 native distribution to the 30th percentile of the 2000 native distribution.

Table Five shows quantile comparisons to explore whether the uptick occurred in some parts of the wage distribution and not others. At the bottom of the wage distribution, at the 10th percentile, new immigrants actually lost ground to natives. The gains become larger as we move up the distribution. The uptick is strongest at the 75th percentile, declining again by the 90th percentile. These results suggest that whatever was the driving force behind the rise in the relative earnings of new immigrant in the 1990s was something going on toward the upper end of the wage distribution.

Education Groups

Figure Eight breaks the sample into four education groups-- high school dropouts, high school graduates, those with some college, and college graduates. The figure charts a time-series of the earnings of new immigrants belonging to each education group, relative to natives with the same level of education. What is apparent is that the earnings of new immigrants with less than a college degree all follow a similar pattern of decline 1960-1990, followed by a modest uptick in the 1990s. In fact, the uptick regression coefficient in the sample of all those with less than a college degree is an insignificant zero.[6] But the pattern for college-graduate immigrants follows a completely different trend, increasing in the 1960s and 1970s, falling somewhat in the 1980s, and rising sharply in the 1990s. The uptick coefficient for this group is 0.099.

It is useful to consider these within-education-group results alongside across-education-group patterns, i.e. changes in the relative educational attainment of immigrants and natives. Figure 9A shows the distribution of new immigrants across the four education group over time. There has been a steady drop in the share of new immigrants who are high school dropouts, and an increase in the share of the other groups, particularly college graduates. Native education, however, has risen at an even faster pace, so that the relative education of new immigrants has fallen over time.[7] As seen in Figure 9B, this is particularly true for the dropout category. However, the 1990s did not represent a sharp break with earlier trends, so changes in relative supplies are probably not the story behind changes in relative wages, particularly the dramatic rise in the relative earnings of new immigrant college graduates. The large increase for the most-educated is consistent with the findings of the previous section, suggesting a focus on the upper end of the immigrant earnings distribution in trying to understand the reasons behind the upturn for immigrants as a whole. The next section explores some new potential reasons for the observed rise in the relative earnings of new immigrants.

IV. NEW EXPLANATIONS

Since the factors that could explain the previous decline in the entry wages of new immigrants before 1990-- changes in the educational attainment and origin-country composition of new immigrants and changes in the wage structure-- cannot account for the recent upturn, and since changes in state of residence and general labor market conditions also do not appear to have caused it, we now turn to another set of potential explanations. The first is attenuation bias, caused by earnings imputation in the PUMS; the second is changes in immigration policy (whom the U.S. admits); the third is an improvement in the selectivity of immigrants (whom the U.S. attracts), brought about by altered opportunities and incentives both in the United States and in the sending countries. The fourth section considers the special case of California.

A. ATTENUATION BIAS

This section considers whether the measured uptick might be an artifact of imputation in the Census data. There has been a substantial increase over time in the extent of data imputation in the PUMS. In the 1990 data, 11% of natives and 20% of new immigrants have allocated wage and salary income data.[8] In the 2000 data, these numbers are 24% and 31%, respectively.[9]

The Census Bureau does not include information on nativity in the formula used to impute wages. Since there are far more natives than immigrants in the population, this results in immigrants with missing wage data essentially being assigned the wages that natives with similar characteristics would earn. When income is imputed without respect to nativity, the estimated coefficient on nativity in a wage regression will therefore suffer from attenuation bias (i.e., be biased toward zero).[10] An increase in the share of observations with imputed wages therefore should result in a reduction in the measured immigrant wage disadvantage over time. However, in Table Six, comparing results for imputed and unimputed samples in the PUMS yields the surprising result that the estimated uptick is higher in the unimputed sample. Imputation therefore cannot help to explain the observed rise in immigrants’ relative earnings. The contrast between the imputed and unimputed results is a puzzle to be resolved in future work.

B. IMMIGRATION POLICY

A second new potential explanation for the rise in the entry wage of new arrivals in the 1990s is that it is somehow related to changes in U.S. immigration policy which occurred over the decade. For example, the Immigration Act of 1990 raised the share of visas allocated on the basis of skill. Though correcting for education does not explain the uptick, perhaps there is some effect of the rise in these high-skilled visas not fully captured by the increase in education. For example, arriving with a job in hand means there is less of an initial earnings gap due to search frictions or imperfect skill transferability, and a better selection of immigrants in terms of English ability, a match between their skills and U.S. labor market opportunities, and so on. This would be consistent with the finding of a large uptick in the relative entry wages of college graduates and those at the upper quantiles of the immigrant wage distribution.

If the 1990 Act raised the average skill level of immigrants, we would expect not only the immigrants who arrived in the late 1990s, but also those who arrived in the early 1990s to have high earnings in 2000. However, as previously shown in Table Four, this was not the case.

Another policy change which took place was the expansion, most notably in the late 1990s, of the H-1B visa program. H-1Bs are temporary, employer-sponsored visas for college graduates who work in “specialty” occupations. In 2001, 58% of H-1B visa holders were in computer-related occupations, with another 12% in engineering and architecture. Over 40% had at least a Masters degree, and half of them were from India. An H-1B visa can be renewed for up to six years, and people on H-1Bs can apply to remain in the U.S. permanently. Table Seven shows the number of H-1B visa recipients for selected years, 1981-2002.[11] In 1985, there were 47,322. The number increased to around 100,000 in the early 1990s, after the cap on such visas was raised by the Immigration Act of 1990. In 1996, the number increased to 144,548. The cap was raised again with the American Competitiveness and Workforce Improvement Act of 1998, when the number rose to 240,947. By 2000, the number of H-1B visas was over 355,000.

A natural question to ask is what share of new arrivals in the late 1990s were H-1Bs? Though it is possible to make a rough estimate, there are some definitional issues to grapple with. The Census asks foreign-born respondents “When did you first come to the U.S. to live/stay?” The answer to this question is used to gauge the number of arrivals in a given year. By contrast, the INS counts as immigrants the number of people who became legal permanent residents in a given year. The Census measures when people first arrived, whether it was on a temporary visa, like the H-1B, or even illegally. The INS measures when those people got green cards (and includes special cases like the IRCA legalization). Strictly speaking, the Census measures the number of foreign-born arrivals, while the INS measures the number of permanent legal immigrations. H-1B visa holders are foreign-born arrivals on temporary work permits. Many, though not all, of these H-1Bs are likely to be sponsored by their employers for green cards.

With these caveats in mind, the INS Statistical Yearbook lists just over 800,000 H-1B visa applications accepted for the period 1995-99 (see Table Seven), not including the missing data for 1997. This number includes both new visas and renewals, so individuals who renew are double-counted. If we make the conservative assumption that half of these represent new visas granted, about 400,000 people entered the U.S. on H1-B visas in the late 1990s. This can be compared to the roughly 5 million people who obtained legal permanent residence in the U.S. over the same period. In other words, H-1Bs made up about 8% of the overall flow of new “immigrants” in this period.

Table Eight shows regressions which focus on “high-tech” workers, defined as computer scientists and engineers. The first column replicates the basic regression documenting the uptick in the entry wage in the 1990s. The second column simply adds a dummy variable for whether an individual is a high-tech worker. This correction reduces the uptick coefficient from .063 to .039. In other words, about one-third of the increase in entry wages over the 1990s can be explained by the fact that there were more high-tech workers among the new arrivals in 1995-99 than there were in 1985-89—11.1% compared to under 5%.

The third column of Table Eight shows further, that when the return to being a high-tech worker is allowed to differ for natives and immigrants, the uptick falls to a marginally significant .011. The last column of the table allows wage growth over the 1990s to be different for native and immigrant high-tech workers. The result is that the estimated uptick disappears completely.

An equivalent way of seeing this result is to split the sample into high-tech and non high-tech workers and run the basic regression separately for the two groups. This is shown in the bottom panel of Table Eight. The first regression shows that, compared to native high-tech workers, immigrant high-tech workers in the late 1990s earned 16.3% more on entry than did those that arrived in the late 1980s. This is consistent with the idea that H-1B visa holders arrive with an earnings advantage, even relative to other high-skill immigrants. In the second column, which excludes high-tech workers, the uptick seen in the aggregate data has disappeared.

Figure Ten shows the trend in the relative earnings of new arrivals 1960-2000, with and without engineers and computer scientists. In both cases, there is a continuous decline 1960-1990. However, in the 1990s, including high-tech workers yields an uptick, and excluding them, the trend goes flat.

It is important to emphasize that it is not just that there were more high-tech workers in this most recent wave of arrivals, but that most of the uptick is due to these high-tech workers earning more, relative to native high-tech workers, than used to be the case. This improvement in immigrant earnings is consistent with a story in which arriving with a job in hand, as is the case with the H-1B visa program, eliminates some of the initial labor market disadvantage of new immigrants. It could also be due to an increase in the supply of well-trained foreign science and technology workers, and the development of networks connecting places like the Indian Institute of Technology to the U.S. high-tech industry. A final possibility is that the quality of these immigrants was higher because the U.S. high-tech boom attracted the best of the international pool of high-tech workers. The issue of improved selectivity is addressed in the next section.

C. POSITIVE SELECTION

Apart from changes in whom the U.S. admitted in the 1990s, another new explanation is that there were changes in whom the U.S. attracted in this period. Borjas (1987) applies the Roy model to immigrants selecting destination countries. He shows that, holding mean wages constant, more-skilled immigrants will be attracted to countries with higher returns to skill, while less-skilled immigrants will be attracted to countries with lower returns to skill.

As discussed above, in the 1990s, overall earnings inequality in the United States rose asymmetrically. The upper half of the wage distribution widened, while the lower half held steady or narrowed, depending on the measure used. By the logic of the Borjas model, these changes would have provided an increased incentive for highly skilled immigrants to come to the U.S., as well as a stable or increased incentive for less skilled immigrants. Could it be that, on net, the changes in the U.S. wage structure attracted a better selection of immigrants, and that this can explain the overall upturn in immigrant earnings?

Considering first the cross-country composition of immigrant flows, the biggest change in the 1990s was the substantial increase in immigration from Mexico, whose immigrants are on average less-skilled than U.S. natives. There was also a smaller increase in immigration from India, whose immigrants are more educated than the native U.S. population. On net, the composition shifted toward Mexico, which is why the uptick in entry wages was found in Table Two to be even more of a puzzle once changes in the country composition of new arrivals were taken into account.

Since cross-country changes cannot help to explain the rise in immigrant earnings, it must be due to something that occurred within countries of origin, rather than across them. In fact, the entry wages of immigrants from some countries rose significantly, while others did not. Running the uptick regression separately by five-digit country of origin yields just seven countries with statistically significant uptick coefficients.[12] Those countries are six Asian countries-- India, China, Taiwan, Hong Kong, Korea, the Philippines-- and Mexico.[13]

Asia

Half of the immigrants from the group of six Asian countries with significant upticks are Indian, and another quarter are Chinese. India and China (as well as Korea and Taiwan) have income distributions more equal than that of the United States. According to the Borjas theory, immigrants from such countries will be positively selected from the population of the home country. Indeed, among new immigrants from these six Asian countries, in 1990, 47% were college graduates, which is much higher than the share in their home countries (or, for that matter, the United States). Consistent with an increasing degree of positive selection motivated by rising U.S. skill prices and in particular, the rise in the return to a college degree, that share grew to 72% in 2000. However, over the 1990s, the relative earnings of immigrants from this group of six countries grew by an enormous 47%, only one-half of which can be explained by the rise in their relative educational attainment and in the return to college in this period. This finding is consistent with increased positive selectivity in unobserved skills in response to the rise in U.S. residual upper-tail inequality. With the share of high-tech workers among this group increasing from 8.2% to 38.5%, it is also consistent with the H-1B visa story described in the previous section.[14]

Mexico

Mexico is by far the most important source country for immigration to the United States. Over 30% of new immigrants in 2000 came from Mexico, and there was a substantial increase in that share in the 1990s. The rise in the supply of Mexican immigrants can be largely attributed to population growth and weak economic conditions-- including a 20% drop in real wages-- in Mexico in the 1990s (Card and Lewis, 2005). Compared to the cohort of Mexicans who arrived in the U.S. in the late 1980s, those who came in the late 1990s earned 6.2% more, relative to natives.[15] With only 5% of new Mexican immigrants having graduated from college, the high-skill story told above clearly does not apply to them. Is there a reason to believe that a better selection of Mexicans is now coming to the United States?

According to the Borjas model, it is the relative degree of inequality in the sending and receiving country that determines the degree of immigrant selectivity. [16] Compression in the lower tail of the U.S. wage distribution in the 1990s would have attracted a worse selection of Mexicans to the United States. And in Mexico, despite a macroeconomic crisis in 1994 which had a disproportionate impact on more-skilled workers and brought a halt to inequality growth, wage inequality was still higher in the late 1990s than it had been in the late 1980s (Airola and Juhn, 2005). Unless potential immigrants responded to the rate of change in inequality, rather than just the level, this would also suggest a diminished incentive for skilled Mexicans to emigrate in the later period.

Nevertheless, compared to the cohort that arrived in the late 1980s, new Mexican immigrants in the late 1990s were more educated. Eight percent of them went from being high school dropouts to high school graduates, with that share rising from 30% to 38%. Most of this can be attributed to rising educational attainment in Mexico, where the high school graduation rate rose by six percentage points.[17] Still, the rise in the education level of Mexican immigrants to the United States was even greater than the rise in Mexico, leaving room for the possibility of increased positive selection. Regardless, since U.S. native educational attainment rose even faster during this period, correcting for education only increases the size of the unexplained earnings uptick for Mexicans.

In terms of other observables, Mexicans in the 1990s were much less likely to settle in California (dropping from 58% in 1990 to 24% in 2000) and more likely to live in Texas, as well as new destinations like North Carolina and Georgia, with a 15 percentage point increase in the share settling in these three states.[18] However, while natives earn substantially more in California than elsewhere, Mexicans do not.[19] So this geographic redistribution did not affect immigrants’ relative earnings at the national level.

Finally, new Mexicans were less likely to work in agriculture, and more likely to work in construction, with 9% shifting from the former to the latter. Adding a dummy variable for whether an individual is a farm laborer causes the Mexican uptick coefficient to fall to an insignificant .014, so that this shift appears to fully explain the rise in the relative earnings of new Mexican immigrants.[20] Whether this was purely a demand-side change, with recruiters pulling Mexicans into these new jobs and locations, or whether it was linked to the increasingly skilled supply of immigrants from Mexico remains for future work to determine.

Cross-Country Analysis of Within-Country Results

To look at the question of selection in a broader set of countries, Figure Eleven displays data for 36 of the biggest sending countries. These two figures consider how cross-country variation in the uptick in immigrant earnings varies with the initial level and subsequent change in each country’s degree of income inequality, relative to the level and change in the United States.

The vertical axis in Figure 11A measures the 1990-2000 change in the relative earnings of new immigrants from a given country, and the horizontal axis measures that country’s Gini coefficient in 1990. The value of the Gini coefficient for the United States in 1990 is marked with a vertical line at 38. Countries to the left of the line have income distributions more equal than that of the United States, predicting positive selection in migration, while countries to the right have more unequal distributions, predicting negative selection. Weighting by the size of the 1995-2000 arrival cohort, the correlation between these two variables is -0.57 (the unweighted correlation is -0.21). This suggests that the more equal was a country’s initial income distribution, the greater was the increase in relative entry wages.

Figure 11B shows the relationship between the uptick in immigrant earnings and the change in the Gini coefficient, 1990-2000. Income inequality, as measured by the Gini coefficient, rose by about 2 points in the United States in this period, and this value is marked in the graph with a vertical line. Weighting by the size of the 1995-2000 arrival cohort, the correlation between these two variables is 0.38 (0.17 unweighted). This suggests the greater the rise in a country’s inequality, the greater was the increase in relative entry wages.

The change in selectivity over time should depend, not only on how inequality changed in the sending country compared to the United States, but also on whether the sending country was initially more or less equal than the United States. More work is needed to understand the significance of these cross-country patterns, including the use of finer measures of inequality that distinguish between upper-tail and lower-tail patterns.

D. CALIFORNIA

For many years, California was by far the most important destination state for new immigrants. But in the 1990s, there was a major shift in the pattern of settlement of new immigrants, with substantial decline in the share of new immigrants settling in California. This was particularly true of Mexicans, with the fraction of new immigrants settling there falling from 58% in the late 1980s to only 24% in the late 1990s.[21]

Table Nine shows results for California versus the rest of the country. The first column, for California alone, shows an enormous uptick of .229. In the fourth column, which excludes California, there is no uptick at all. However, this result merely reflects the geographic redistribution of Mexicans, who earn less than other immigrants, away from California. The share of Mexicans among new immigrants in California held steady at about 40%, while their share among new immigrants outside of California doubled, from 15% to 30%.[22] When we correct for this composition effect, as expected, the uptick for California is unchanged, and an uptick appears outside of California as well. In the case of immigrants living outside of California, the uptick for Mexicans is fully explained by the drop in the share working as farm laborers, and the remainder of the uptick is explained by the wage growth of high-tech immigrants.

However, for new immigrants living inside California, these factors cannot explain away the rise in their relative earnings. Mexicans in California were not much less likely to be farm laborers than before. And even allowing for a separate pattern for high-tech immigrants, there remains a substantial uptick of over 20% for California immigrants taken as a whole.

What factors lie behind the substantial rise in the relative earnings of new immigrants to California— which occurred across educational groups and metropolitan areas? One possible link is to the drop in the share of new immigrants settling in California.

Native wages are higher in California than in the rest of the country. But in the 1990s, the bottom end of the California wage distribution collapsed, both in absolute terms and particularly when compared to the rest of the United States.[23] This might help to explain why fewer new arrivals, especially the least-skilled, chose to settle in California in the 1990s.

Upper-tail inequality increased in California in this period, which would have been attractive to the minority of new immigrants who are high-skilled. The educational attainment of immigrants outside of California was unchanged in the 1990s. But inside California, there was an 11 percentage point drop in the share of immigrants who were high school dropouts, and a 14 percentage point rise in the share of college graduates. Still, these changes in educational attainment can explain only a small part of the rise in the relative earnings of new arrivals.

There is other evidence to suggest that California “deflected” new immigrants. Research on Los Angeles suggests that an intentional tightening of housing ordinances raised rents and reduced the accessibility of Los Angeles to poorer immigrants, thus deflecting further inflows, while maintaining its attractiveness to better-off immigrants (Light, 2006). The drop in the supply of new immigrants, particularly less-skilled ones, is consistent with the observed rise in their relative wages.

V. SUMMARY

Following a thirty-year decline in the earnings of new immigrants relative to natives, the immigrants who came to the United States in the late 1990s earned more on arrival than the cohort that preceded them. This paper explores the factors behind the recent improvement in immigrant performance.

We find that, while new arrivals in the late 1990s were more educated than were new arrivals a decade earlier, this change alone can explain only a small part of the rise in entry wages. Taking into account other changes in the composition of new immigrants-- such as the rise in the share originating in Mexico and the decline in the share settling in California-- only makes the rise in entry wages more difficult to explain. Based on the experience of other cohorts, neither does the rise in entry wages appear to have been the result of an economic environment or wage structure generally favorable toward immigrants.

Having found that the explanations which accounted for the previous decline in the earnings of new immigrants cannot account for the recent rise, we turn to a new set of explanations. The first, that the measured upturn is an artifact of attenuation bias due to imputation in the PUMS, is not supported by the evidence.

Rather, we find the upturn was due in part to an improvement in the earnings of new immigrant computer scientists and engineers, relative to comparable natives, compared to earlier cohorts of immigrant high-tech workers. It should be emphasized that it is not just that there were more high-tech workers in this most recent wave of immigrants, but that these high-tech workers earned more, relative to native high-tech workers, than used to be the case. This improvement in immigrant earnings is consistent with a story in which arriving with a job in hand, as is the case with the H-1B visa program, eliminates some of the initial labor market disadvantage of new immigrants. It is also possible that the high-tech boom and rising upper-tail wage inequality attracted a better selection of workers to the United States.

Another part of the recent upturn in immigrant earnings was the improvement in the entry wages of Mexicans. Most of this change can be explained by the movement of Mexicans out of agricultural labor, though there is also evidence consistent with an improvement in the quality of immigrants coming from Mexico in this period.

Has the U.S. found a recipe for getting better immigrants, or for improving immigrant performance? First, it remains to be seen how the cohort of the late 1990s will perform in the future. This depends, first, on whether the most successful members of this arrival-cohort remain in the U.S. after the six-year limit on their temporary H-1B visas expires. It is likely that most would be sponsored for green cards by their employers. Yet, with the collapse of the U.S. high-tech sector, some may not be sponsored, and others may choose to leave for third countries, like Germany. Second, if the high entry earnings of this cohort were due, not to positive selectivity, but simply to the advantage of arriving with a job in hand and thus bypassing the slow process of wage assimilation toward some long-run level, the earnings advantage of this cohort may be only a short-run phenomenon.

At the other end of the immigrant skill distribution, the flows and future performance of Mexican immigrants in the United States will depend on pull factors like demand growth in industries that employ less-skilled labor, push factors like economic conditions in Mexico, and changes in the U.S. and Mexican wage structures that affect both the composition of Mexicans emigrating to the United States as well as their relative earnings once here.

The United States is currently considering a major reform of immigration policy. One of the proposals on the table would create a temporary guest-worker program aimed at the 11 million undocumented workers in the country—a kind of low-skill H-1B program. Such a program would surely increase the opportunities of these workers, but the general equilibrium effects are not obvious. Another proposal is to raise the cap on the number of H-1B visas, which was cut drastically in 2004. The analysis in this paper shows H-1B recipients do very well in the U.S. labor market, even when compared to native high-skill workers, though at a much higher cap, an increase in their supply might eventually lead to a reduction in their relative earnings.

BIBLIOGRAPHY

Airola, Jim and Chinhui Juhn, “Wage Inequality in Post-Reform Mexico,” University of Houston Working Paper, January 2005.

Angrist, Joshua, Victor Chernozhukov and Ivan Fernandez-Val, “Quantile Regression Under Misspecification, with an Application to the U.S. Wage Structure,” Econometrica, forthcoming.

Autor, David H., Melissa S. Kearney, and Lawrence F. Katz, “The Polarization of the U.S. Labor Market,” NBER Working Paper 11986, January 2006

Autor, David H., Melissa S. Kearney, and Lawrence F. Katz, “Rising Wage Inequality: The Role of Composition and Prices,” NBER Working Paper 11628, September 2005.

Autor, David H., Melissa S. Kearney, and Lawrence F. Katz, “Trends in U.S. Wage Inequality: Re-Assessing the Revisionists,” NBER Working Paper 11627, September 2005.

Barro, Robert, and Jong-Wha Lee, “International Data on Educational Attainment: Updtaes and Implications,” Harvard University Center for International Development, Working Paper 42, April 2000.

Borjas, George J. "The Economic Analysis of Immigration," in Handbook of Labor Economics, Volume 3A, edited by Orley Ashenfelter and David Card, North-Holland, 1999, pp. 1697-1760.

Borjas, George J. “Self-Selection and the Earnings of Immigrants.” American Economic Review, Sept. 1987, pp.531-553.

Borjas, George J. “Assimilation, Changes in Cohort Quality, and the Earnings of Immigrants.” Journal of Labor Economics, Oct. 1985, pp.463-89.

Borjas, George J., “Assimilation and Changes in Cohort Quality Revisited: What Happened to Immigrant Earnings in the 1980s?” Journal of Labor Economics, vol.3 no.2, 1995.

Borjas, George J., “The Rise of Low-Skill Immigration in the South,” unpublished manuscript, July 2004.

Butcher, Kristin F., and John E. DiNardo, “The Immigrant and Native-Born Wage Distributions: Evidence from the United States Censuses” Industrial and Labor Relations Review, vol.56, no.1, October 2002, pp.97-121.

Card, David and Ethan G. Lewis, “The Diffusion of Mexican Immigrants During the 1990s: Explanations and Impacts,” NBER Working Paper 11552, August 2005.

Chiquiar, Daniel and Gordon H. Hanson, “International Migration, Self-Selection, and the Distribution of Wages: Evidence from Mexico and the United States,” Journal of Political Economy, vol.113, no.2, April 2005, pp.239-281.

Deininger and Squire. Cross-Country Measures of Income Inequality. (www url).

Funkhouser, Edward and Stephen J. Trejo, “The Labor Market Skills of Recent Male Immigrants: Evidence from the Current Population Survey,” Industrial and Labor Relations Review, vol. 48, no. 4, July 1995.

Hanson, Gordon H. “Emigration, Labor Supply, and Earnings in Mexico,” NBER Working Paper 11412, June 2005.

Hirsch, Barry T. and Edward J. Schumacher, “Match Bias in Wage Gap Estimates Due to Earnings Imputation,” Journal of Labor Economics, vol. 22, no. 3, July 2004, pp.689-722.

Jasso, Guillermina, Mark R. Rosenzweig, and James P. Smith, “The Changing Skill of New Immigrants to the United States: Recent Trends and Their Determinants” in Issues in the Economics of Immigration, George J. Borjas, ed. Chicago: University of Chicago Press, forthcoming.

Jasso, Guillermina, Mark R. Rosenzweig, and James P. Smith, “The Earnings of U.S. Immigrants: World Skill Prices, Skill Transferability and Selectivity,” unpublished manuscript, March 2002.

LaLonde, Robert J. and Robert H. Topel, “The Assimilation of Immigrants in the U.S. Labor Market.” In Immigration and the Work Force: Economic Consequences for the United States and Source Areas, edited by George J. Borjas and Richard Freeman. Chicago: University of Chicago Press, 1992, pp.67-92.

LaLonde, Robert J. and Topel, Robert H. "Economic Impact of International Migration and the Economic Performance of Migrants," M. R. Rosenzweig and O. Stark, Handbook of Population and Family Economics. Handbooks in Economics, vol. 14. Amsterdam; New York and Oxford: Elsevier Science North-Holland, 1997, pp. 799-850.

Light, Ivan. Deflecting Immigration: Networks, Markets, and Regulation in Los Angeles. New York: Russell Sage Foundation, 2006.

Lubotsky, Darren, “The Effect of Changes in the U.S. Wage Structure on Recent Immigrants’ Earnings,” Princeton University IR Section Working Paper #458, Sept. 2001.

Munshi, Kaivan, “Networks in the Modern Economy: Mexican Migrants in the U.S. Labor Market,” Quarterly Journal of Economics, May 2003, pp.549-599.

U.S. Department of Homeland Security, Office of Immigration Statistics, “Characteristics of Specialty Occupation Workers (H-1B): Fiscal Year 2003,” Nov. 2004.

U.S. Immigration and Naturalization Service (later, Citizenship and Immigration Service). Statistical Yearbook of the Immigration and Naturalization Service. Washington, D.C.: Government Printing Office, various years.

Zuniga, Victor, and Ruben Hernandez-Leon. New Destinations: Mexican Immigrants in the United States. New York: Russell Sage Foundation, 2005.

TABLE ONE: Composition of New Arrivals

TABLE 1A: Continent of Birth

| |1990 |2000 |Change |

|Mexico |24.1 |31.1 |7.0 |

|Central America |8.4 |5.7 |-2.7 |

|Caribbean |7.4 |6.5 |-0.9 |

|South America |7.7 |6.8 |-0.9 |

|Northwest Europe |6.4 |5.1 |-1.3 |

|Southeast Europe |7.4 |7.3 |-0.1 |

|Former Soviet U. |1.4 |3.1 |1.7 |

|East Asia |12.5 |7.8 |-4.7 |

|Southeast Asia |8.8 |5.6 |-3.2 |

|Southwest Asia |7.5 |10.6 |3.1 |

|Other |8.4 |10.4 |2.0 |

|Total |100.0 |100.0 |0.0 |

Note: The data for 1990 are for immigrants who arrived 1985-89, and the data for 2000 are for immigrants who arrived 1995-99. The sample is restricted to men aged 25-64 who are employed in the civilian sector.

TABLE 1B: Education

|  |  |1990 |2000 | change |

|HIGH SCHOOL Dropout |0 |6.9 |4.6 |-2.4 |

| |1-4th |5.1 |2.5 |-2.6 |

| |5-8th |12.7 |13.4 |0.7 |

| |9th |4.0 |4.9 |0.9 |

| |10th |2.9 |2.4 |-0.4 |

| |11th |2.0 |2.0 |0.0 |

|High School |12th |5.8 |5.8 |0.1 |

| |HS |17.7 |17.9 |0.2 |

|Some |Some |10.0 |9.1 |-0.9 |

|college |Associate |4.7 |3.3 |-1.4 |

|COLLEGE GRADUATE |BA |14.6 |17.4 |2.8 |

| |MA |7.4 |9.9 |2.5 |

| |professional |3.3 |3.5 |0.3 |

| |PhD |3.2 |3.4 |0.2 |

|Total |  |100.0 |100.0 |0.0 |

TABLE 1C: Occupation and Industry

| OCCUPATION |1990 |2000 |change |

|Professional, Technical |16.8 |22.7 |6.0 |

|Farmers |0.3 |0.1 |-0.2 |

|Manager |10.3 |9.2 |-1.1 |

|Clerical |7.2 |6.3 |-0.9 |

|Sales |3.2 |2.7 |-0.5 |

|Crafts |14.0 |14.9 |0.9 |

|Operatives |17.7 |17.5 |-0.2 |

|Service |16.8 |13.8 |-2.9 |

|Farm labor |3.8 |2.3 |-1.5 |

|Laborers |10.0 |10.4 |0.4 |

|Total |100.0 |100.0 |0.0 |

| INDUSTRY |1990 |2000 |change |

|Agriculture |6.9 |5.1 |-1.8 |

|Construction |11.9 |14.4 |2.4 |

|Durables Manufacturing |12.9 |12.7 |-0.2 |

|Nondurables Manufacturing |9.4 |8.8 |-0.6 |

|Transportation, Communication |4.6 |4.9 |0.2 |

|Trade |24.0 |21.0 |-3.0 |

|Finance, Insurance, Real Estate |4.2 |3.4 |-0.9 |

|Services |24.1 |28.2 |4.1 |

|Public |2.0 |1.7 |-0.4 |

|Total |100.0 |100.0 |0.0 |

TABLE 1D: State

|  |1990 |2000 |Change |

|Arizona |1.4 |2.4 |1.0 |

|California |33.2 |19.0 |-14.1 |

|Colorado |0.6 |1.9 |1.4 |

|Connecticut |1.5 |1.4 |-0.2 |

|Florida |7.3 |8.4 |1.1 |

|Georgia |1.2 |3.6 |2.3 |

|Illinois |4.7 |5.5 |0.8 |

|Indiana |0.4 |1.0 |0.6 |

|Maryland |2.1 |1.8 |-0.3 |

|Massachusetts |3.2 |2.7 |-0.5 |

|Michigan |1.3 |2.3 |1.1 |

|Minnesota |0.6 |1.1 |0.6 |

|Nevada |0.7 |1.1 |0.4 |

|New Jersey |6.2 |5.3 |-0.9 |

|New York |16.0 |10.3 |-5.7 |

|North Carolina |0.8 |3.1 |2.3 |

|Ohio |0.9 |1.4 |0.4 |

|Oregon |0.6 |1.0 |0.4 |

|Pennsylvania |1.4 |1.6 |0.2 |

|Texas |6.5 |10.6 |4.0 |

|Virginia |2.3 |2.5 |0.2 |

|Washington |1.5 |2.2 |0.7 |

TABLE TWO: Correcting for Composition

|Immigrant |-.378 |

| |1970 |1980 |1990 |2000 |

|1995-99 | | | |-.314 |

| | | | |(.004) |

|1990-94 | | | |-.320 |

| | | | |(.004) |

|1985-89 | | |-.378 |-.282 |

| | | |(.006) |(.004) |

|1980-84 | | |-.323 |-.200 |

| | | |(.005) |(.005) |

|1975-79 | |-.319 |-.192 |-.112 |

| | |(.007) |(.006) |(.005) |

|1970-74 | |-.214 |-.099 |-.061 |

| | |(.007) |(.007) |(.007) |

|1965-69 |-.188 |-.086 |.014 |.046 |

| |(.010) |(.008) |(.008) |(.008) |

|1960-64 |-.049 |-.000 |.088 |.122 |

| |(.011) |(.009) |(.009) |(.011) |

|1950-59 |.053 |.058 |.177 |.169 |

| |(.008) |(.007) |(.009 |(.012) |

|pre-1950 |.090 |.099 |.232 |.260 |

| | (.007) |(.009) |(.017) |(.030) |

Note: Each column represents a separate regression. The coefficients shown are the unconditional log earnings, relative to natives, of the specified arrival-cohort in the specified Census year. Sample sizes are 301,510 in 1970; 471,305 in 1980; 586,927 in 1990; and 767,258 in 2000.

TABLE FOUR

Cross-Cohort Comparisons

Did the uptick occur for earlier cohorts as well? No.

|Years Since Migration |Arrival Cohort |1990 |2000 |Change |

|0-5 |y85 vs y95 |-.378 |-.314 | .064 (.009) |

|5-10 |y80 vs y90 |-.323 |-.320 | .002 (.009) |

|10-15 |y75 vs y85 |-.192 |-.282 |-.090 (.009) |

|15-20 |y70 vs y80 |-.099 |-.200 |-.100 (.010) |

Was assimilation particularly fast in the 1990s? No.

First decade assimilation rates:

y85 over the 1990s: -.378 to -.282 = .096 (.009)

y75 over the 1980s: -.319 to -.192 = .128 (.011)

Second decade assimilation rates:

y80 over the 1990s: -.323 to -.200 = .123 (.009)

y70 over the 1980s: -.214 to -.099 = .115 (.012)

Note: The table compares outcomes of different five-year arrival cohorts. y95 indicates immigrants who arrived 1995-99, y90 indicates those who arrived 1990-94, etc.

TABLE FIVE

Quantile Comparisons

|Quantile of the 1990 |Quantile of the 1990 |Quantile of the 2000 |Change in Quantile of the|

|Immigrant Wage |Native Wage Distribution |Native Wage Distribution |Native Wage Distribution |

|Distribution | | | |

|10% |3.5% |3.0% | -0.5 |

|25% |7.3% |7.9% | 0.6 |

|50% |20.2% |23.7% | 3.5 |

|75% |51.5% |64.8% |13.3 |

|90% |85.3% |89.3% | 4.0 |

| |1990 |2000 | | |Gain of Immigrants |

| | | |Native Wage |Immigrant Wage Growth |Relative to Natives |

| | | |Growth | | |

| |Native log wage|Immigrant log |Native log wage|Immigrant log | | | |

| | |wage | |wage | | | |

|10% |1.730 |1.322 |1.745 |1.314 |.015 | -.008 | -.023 |

|25% |2.136 |1.609 |2.146 |1.671 |.010 |.062 |.052 |

|50% |2.528 |2.040 |2.528 |2.108 |0 |.068 |.068 |

|75% |2.900 |2.560 |2.904 |2.739 |.004 |.179 |.175 |

|90% |3.219 |3.090 |3.290 |3.267 |.071 |.177 |.106 |

|mean |2.510 |2.131 |2.534 |2.219 |.024 |.088 |.064 |

Note: The numbers in the final column of the lower panel are the same as the coefficient on the variable immigrant*2000 generated by a quantile regression of log wages on dummy variables for immigrant, 2000, and immigrant*2000.

TABLE SIX

Attenuation Bias

| |Full Sample |Imputed Sample |Non-Imputed Sample |

|Immigrant | -.378 (.007) |-.216 (.017) |-.402 (.008) |

|2000 | .024 (.001) |.084 (.004) |.031 (.001) |

|Immigrant*2000 | .064 (.009) |-.050 (.020) |.076 (.011) |

|Constant |2.51 (.001) |2.37 (.003) |2.53 (.001) |

| | | | |

|R2 |.007 |.008 |.007 |

|Natives |896,468 |165,673 |730,687 |

|Immigrants |21,765 |6,104 |15,769 |

Note: Data are the pooled 1990 and 2000 PUMS, natives and new immigrants only. Sample includes men aged 25-64 who are employed in the civilian sector. Observations are classified as “imputed” if they contain imputed nativity or allocated wage data. All regressions are weighted by the PUMS sample weights.

TABLE SEVEN

H-1B Visas

|Fiscal Year |Number of H-1Bs |

|1981 |. |

|1985 |47,322 |

|1989 |89,856 |

|1990 |100,446 |

|1991 |114,467 |

|1992 |110,223 |

|1993 |92,795 |

|1994 |105,899 |

|1995 |117,574 |

|1996 |144,458 |

|1998 |240,947 |

|1999 |302,326 |

|2000 |355,605 |

|2001 |384,191 |

|2002 |370,490 |

|2003 |360,498 |

|2004 |386,821 |

Source: INS Statistical Yearbook, 1996 and 1999; U.S. Department of Homeland Security, Office of Immigration Statistics Yearbook, 2004.

TABLE EIGHT

Computer Scientists and Engineers

|Immigrant |-.378 (.007) |-.380 (.007) |-.398 (.007) |-.393 (.007) |

|2000 |.023 (.001) |.016 (.001) |.017 (.001) |.017 (.001) |

|Immigrant*2000 |.063 (.009) |.039 (.009) |.011 (.009) |.001 (.009) |

|High-Tech | |.440 (.003) |.423 (.003) |.430 (.004) |

|High-Tech*Immigrant | | |.407 (.015) |.278 (.035) |

|High-Tech*2000 | | | |-.010 (.006) |

|High-Tech*Immigrant*2000 | | | |.162 (.039) |

|Constant |2.50 (.000) |2.49 (.000) |2.49 (.000) |2.49 (.001) |

|R2 |0.007 |0.028 |0.029 |0.029 |

|Natives |896,468 |896,468 |896,468 |896,468 |

|Immigrants |21,765 |21,765 |21,765 |21,765 |

| |High Tech |Excluding |

| |Only |High-Tech |

|Immigrant |-.114 (.025) |-.393 (.007) |

|2000 |.006 (.004) |.017 (.001) |

|Immigrant*2000 |.163 (.028) |.001 (.009) |

|Constant |2.92 (.003) |2.49 (.001) |

|R2 |0.001 |0.008 |

|Natives |43,542 |852,933 |

|Immigrants |1,968 |19,790 |

Note: Data are the pooled 1990 and 2000 PUMS, natives and new immigrants only. Sample includes men aged 25-64 who are employed in the civilian sector. All regressions are weighted by the PUMS sample weights.

TABLE NINE

California

| |California Only |Excluding California |

|Immigrant |-.669 (.013) |-.444 (.015) |

R2 |0.038 |0.048 |0.053 |0.005 |0.007 |0.009 | |Natives |91,585 |91,585 |91,585 |804,486 |804,486 |804,486 | |Immigrants |5,848 |5,848 |5,848 |16,314 |16,314 |16,314 | |

Note: Data are the pooled 1990 and 2000 PUMS, natives and new immigrants only. Sample includes men aged 25-64 who are employed in the civilian sector. All regressions are weighted by the PUMS sample weights.

FIGURE ONE

Log Earnings of New Immigrants Relative to Natives

[pic]

Note: Figure derived from Table Three. Each point represents the average log earnings of immigrants who arrived in the five years prior to the specified Census year, relative to average native earnings in that year.

FIGURE TWO

Log Earnings of Natives and New Immigrants

[pic]

Note: Each point represents average log earnings in the specified Census year. New immigrants are those who arrived in the five years prior to the specified year.

FIGURE THREE

CPS Estimates of the Relative Earnings of

New Immigrants, 1994-2003

[pic]

Note: The 1960-2000 line was calculated using the PUMS. The 1994-2003 line was calculated using the March CPS.

FIGURE FOUR

U.S. Labor Market Conditions

[pic]

Source: U.S. Bureau of Labor Statistics.

FIGURE FIVE

Trends in the Relative Earnings of

Immigrant Arrival Cohorts

Notes: The relative wage is calculated in the sample of working men aged 25-64 who are not enrolled in school and who worked in the civilian sector. The hourly wage rate is defined as the ratio of total income earned annually to annual hours worked in the calendar year prior to the Census.

FIGURE SIX (A)

Earnings Assimilation by Arrival Cohort

Notes: The relative wage is calculated in the sample of working men who are not enrolled in school and who worked in the civilian sector. The hourly wage rate is defined as the ratio of total income earned annually to annual hours worked in the calendar year prior to the Census.

FIGURE SIX (B)

Earnings Assimilation by Arrival Cohort

[pic]

Note: Figure derived from Table Three. Each point represents the average log earnings of immigrants from a particular arrival cohort, relative to natives, in the given Census year. The cohorts are defined in five-year intervals, i.e., “y95” denotes immigrants who arrived 1995-99, “y90” those who arrived 1990-94, and so forth, with the exception that “y50” denotes those who arrived 1950-59, and “y40” denotes immigrants who arrived prior to 1950.

FIGURE SEVEN (A)

The Native Wage Distribution in 1990 and 2000

[pic]

FIGURE SEVEN (B)

The New Immigrant Wage Distribution in 1990 and 2000

[pic]

FIGURE SEVEN (C)

Wage Distribution by Nativity in 1990

(Natives and New Immigrants)

[pic]

FIGURE SEVEN (D)

Wage Distribution by Nativity in 2000

(Natives and New Immigrants)

[pic]

FIGURE EIGHT

Earnings of New Immigrants Relative to Natives

By Level of Education

[pic]

FIGURE NINE (A)

Educational Attainment of New Immigrants

[pic]

FIGURE NINE (B)

Educational Attainment of New Immigrants,

Relative to Natives

[pic]

FIGURE TEN

The Relative Wages of New Immigrants:

High-Tech and Non-High-Tech

Notes: The relative wage is calculated in the sample of working men aged 25-64 who are not enrolled in school and who worked in the civilian sector. The hourly wage rate is defined as the ratio of total income earned annually to annual hours worked in the calendar year prior to the Census.

FIGURE ELEVEN (A)

Changes in Entry Wages and

the Level of Origin-Country Inequality

[pic]

Note: The Gini coefficient ranges from 0 to 100, with higher numbers representing greater levels of income inequality. Source is the Deininger and Squire dataset. The vertical line in the graph represents the U.S. value of 38.

FIGURE ELEVEN (B)

Changes in Entry Wages and

Changes in Origin-Country Inequality

[pic]

Note: The source of the Gini coefficient data is the Deininger and Squire dataset. The vertical line in the graph marks the U.S. value of 2.

-----------------------

[1] The sample also excludes those with zero reported weeks of work, no wage income, the self-employed, those enrolled in school, and those living in group quarters.

[2] Hourly earnings are calculated from annual wage and salary income, divided by weeks worked per year, divided by hours worked per week.

[3] Future work should address the reason for this difference.

[4] There was no change in the age distribution of new arrivals between 1990 and 2000 (the average age in this sample was 36), but the native population aged. Another characteristic which did not change across these arrival cohorts was English fluency. The share reporting speaking English “very well” held steady at 38%.

[5] Another potential effect of changes in U.S. earnings inequality is to change the selection of immigrants who choose to migrate to the United States. We address this issue in section IVC.

[6] When the regression is run separately for each of these categories, the only statistically significant result is a 0.059 uptick for dropouts, which is driven by a 0.046 reduction in the wages of native dropouts over this period.

[7] As their numbers dwindle, it may well be that native dropouts are increasingly negatively selected, contributing to the declining wages observed for this group.

[8] A sizeable share of observations coded as new immigrants have imputed nativity status as well.

[9] Imputation in the CPS-MORG also increased over the 1990s, from 4% in 1990, jumping to 23% in 1995, and to 30% in 2000.

[10] Hirsch and Schumacher (2004) study the effect of earnings imputation in the 1996-2001 CPS-ORG on wage gap estimates, and find substantial attenuation bias. The estimated coefficient on an immigrant dummy goes from -.063 to -.081 when allocated earners are excluded from the sample.

[11] The H-1B visa category as such was formally established by the 1990 Act. The number of new H-1Bs was capped at 67,000 in 1997, rising to 115,000 in 1999, and to 195,000 in 2001. In 2004, the cap was brought back down to 65,000 (plus 20,000 in 2005 for foreigners holding U.S. graduate degrees).

[12] Care should be taken in treating statistical significance as the metric of the importance of the uptick, since countries with more immigrants have larger sample sizes in the PUMS, providing greater power to reject the null of no change in entry wages.

[13] The uptick coefficients for the individual countries are: India .434 (s.e. .041), China .319 (.051), Taiwan .219 (.107), Hong Kong .407 (.138), Korea .276 (.063), Philippines .168 (.048), and Mexico .062 (.019).

[14] Despite their high education, the earnings of the average immigrant from each of these countries falls in the lower half of the native wage distribution in all cases, except India and Taiwan in 2000, so a mechanical effect of rising upper-tail inequality boosting immigrant earnings (as in section IIIB) cannot explain the uptick for this group.

[15] Considering again the mechanical effect of compression of the U.S. wage distribution, in 1990, the average new Mexican earned at the 10th percentile of the native wage distribution.  Over the 1990s, the mean/10% differential shrank by .8 percentage points, so this mechanical effect can account for only one-eighth of the Mexican uptick.

[16] With a more unequal income distribution than the United States, the theory predicts Mexican immigrants to the United States would be negatively selected. Chiquiar and Hanson (2005) provide evidence that Mexican immigrants to the United States are more educated than non-emigrant Mexicans, and are drawn from the middle of Mexico’s wage distribution. Hanson (2005) finds emigration rates to be highest among Mexicans with earnings in the top half the Mexican wage distribution.

[17] The share of men over the age of 25 who had completed high school rose from 5.9% in 1985 to 11.8% in 1995 (Barro and Lee, 2000).

[18] See Borjas (2004) for more on the new immigration to the South, and Zuniga and Hernandez-Leon (2005) on the legal changes that paved the way for the geographic shift.

[19] Mexicans earned 2.7% more in California than elsewhere in 1990, and 2.8% less in 2000.

[20] Though not within the state of California. See section IVD.

[21] The shift was not, however, limited to Mexicans. Among non-Mexicans, the share fell from 26% to 17%.

[22] Much of this is due to the redistribution of Mexicans away from California. The educational attainment of Mexicans inside versus outside California was similar in both years.

[23] Considering the mechanical effect on the relative wages of new immigrants, in 1990, the average new immigrant in California earned at the 13th percentile of the California native wage distribution. Over the 1990s, real earnings at the 13th percentile fell by 5%, relative to the California native mean. This makes the observed 20% upturn in relative entry wages all the more striking.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download