TRENDS IN IMMIGRATION FROM 1980 TO 2008:

THE EFFECTS ON INCOME AND EDUCATION IN THE UNITED STATES

Sung B. Park

University of Michigan

Econ 466: Economics of Population

April 2010

 
1  
 
INTRODUCTION

Immigration reform is a controversial issue in the US because it directly and indirectly

affects millions of people. The Obama administration has expressed a strong desire to tackle this

issue but it tends to ignite a fierce debate between those who support more immigration into the

US and those who support less. It is easy to underestimate the magnitude of the issue. Often, the

immigration argument seeps into the debate on other issues like healthcare reform because the

argument essentially comes down to who should be allowed into this country and by how much.

Furthermore, the economic recession of 2008 and the massive job losses in the US have added

fuel to the debate about how immigrants affect the economy. In particular, the American public

is concerned that legal and illegal immigrants take jobs away from US citizens and depress

wages because immigrants are willing to work for lower salaries and wages. The motivation for

my research proposal is to address this concern every American has in the back of his or her

mind. My research question originates from a broad desire to understand how changes in the

flow of immigration and immigration demographics over the past three decades have affected the

rate of unemployment in the US. If immigration does have an effect on the unemployment rate, I

am also interested in any further consequences on labor that concern Americans such as wages.

Hypothesis

My hypothesis is that from 1980 to 2008, (1) unskilled immigrant workers have had some

effect on the unemployment rate of unskilled native workers. Admittedly, this proposition will be

difficult to prove due to many other variables that account for the unemployment of unskilled

native workers, especially the economic conditions at any moment in time. I expect the U.S.

Census to reveal that (2) skilled worker immigration will have no effect on the unemployment

rate of skilled workers. This proposition will be difficult to prove as well along the same

 
2  
 
reasoning as unskilled workers, however, greater employment stability inherent in skilled

occupations should lessen variations in unemployment rates compared to unskilled occupations. I

expect to see (3) a gradual increase in the educational attainment of skilled native workers in

comparison to skilled immigrant workers in order for them to remain competitive in the skilled

labor market. Finally, I believe that (4) the wages of all unskilled workers will decline while the

wages of skilled workers will remain constant because of a greater elasticity of demand for

skilled labor as opposed to unskilled labor. The focus of this research will be on propositions 3

and 4 more than propositions 1 and 2. The difficulty in using the Census data for analysis on

unemployment is substantial. One significant reason is that the data is only available in ten year

intervals between 1980 and 2008. Hence, the four data sets that represent this twenty year period

are not the most qualified data for analyzing unemployment rates.

Dividing the labor market into skilled and unskilled labor will be important in

understanding the specific effects of immigration on wages. My hypothesis of the effects of

immigration on unskilled worker wages is based on the expectation that the unskilled immigrant

workers increase the supply of labor for unskilled workers immediately. Due to the low barriers

for entry into the unskilled labor market, unskilled immigrant workers will directly compete with

unskilled native workers for these jobs. By increasing the supply of labor, the wages of unskilled

labor will decline as employers can more easily replace unskilled workers for cheaper workers as

necessary. My hypothesis of the effects of immigration on skilled workers is based on the

expectation that the labor market demand for skilled labor is more inelastic than the demand for

unskilled labor. The labor demand for skilled labor differs from unskilled labor because skilled

labor is scarce and specific according to employer needs and therefore, employers cannot replace

skilled workers as readily as unskilled workers.

 
3  
 
 By understanding the relationship between employment and productivity, this leads to the

expectation that educational attainment is correlated to the increase in real wages and constant

level of unemployment rates for skilled native workers in the context of immigration. Also, the

lack of educational attainment is correlated with stagnant real wages and fluctuating levels of

unemployment rates for unskilled native workers in the context of immigration.

My research question centers on immigration, employment, and wages but I also believe

education will play an important role in understanding the ultimate effects of immigration in the

US. How native workers cope with increased competition in the labor market will depend on

their education and experience because a reasonable assumption is that skilled and unskilled

immigrants will remain at a fixed level of education. What I predict is that immigrants come to

the US with little education or they come to the US with a given amount of education in which

the opportunities to increase educational attainment is significantly lower than skilled native

workers.

LITERATURE REVIEW

To begin, this research question must be broken down into several components. The first

problem that must be addressed is to identify the immigrants that are the focus of this proposal.

The proposal is limited by the Census data, and therefore, not all immigrants in the US will be in

all the Census analyses. Not all immigrants should be expected to complete and return the

Census due to language barriers, temporary residence, security reasons, and a lack of interest.

Specifically, Passel’s (1986) immigrant categorization is helpful in understanding the motives

and intentions of immigrants for the purposes of this study. He categorizes immigrants into

sojourners and settlers, those who intend to stay in the US and those who intend to leave. This

 
4  
 
applies to both legal and illegal immigrants. The Census does not differentiate between legal and

illegal immigrants and as a result, this categorization will not be relevant for the study. What is

important is to identify those immigrants who are likely to be a part of the Census data and those

who are not.

Who are the sojourners, which of them are unlikely to be a part of the Census data and

how will they limit the scope of this study? Several groups of immigrants may qualify but the

sojourners of highest interest are those with some economic consequences. Specifically, seasonal

migrant workers should be included in this category, especially those involved in the agriculture

industry. Torok and Huffman’s (1986) study on US-Mexican trade involving agriculture and

illegal immigration provide some indication that the relationship between immigration,

employment, and wages are linked in a complicated way. They find that the economic conditions

of US and Mexico must be taken into account in order to explain rises and downturns in the

normal levels of immigration of seasonal migrant workers into the US. They state that the “push”

and “pull” factors involved in the employment of seasonal workers is food prices, unemployment

rates in the US and Mexico, wages in the US and Mexico, and the decrease in the native US

population. Although these seasonal migrant workers are unlikely to directly affect any Census

analyses, it is worth noting that the causal relationship between employment and immigration

can go both ways.

Who are the settlers, how likely are they going to be a part of the Census data and what

impact could they have on wages and employment? Broadly speaking, immigrants who are

considered to be “settlers” would most likely have residence in the US and a permanent source of

income. These settlers may also be divided into skilled and unskilled workers, those immigrants

who have high educational attainment upon their arrival and those immigrants who do not.

 
5  
 
According to Chiswick’s (1989) model of human capital, the wages of the native population with

high human capital are thought to decrease in the short-term due to skilled immigrant workers

being substitutes. However, in the long run, the native population accumulates more human

capital and they also have a mobility advantage within their native labor market which gives

them a net gain. The summary of his argument is that an increase in skilled immigration tends to

incentivize greater human capital investment among natives and therefore, the impact of skilled

immigration really depends on the supply elasticity of human capital. The implication of his

argument to this research is that for any increase in the flow of highly skilled workers,

unemployment of highly skilled native workers might increase in the short-run, however, in the

long-run, the unemployment rate will adjust to some equilibrium level before the increase in

highly skilled immigrants. The Census might also show an increase in the educational attainment

of native workers from the time of the increase in highly skilled immigrants to the time in which

unemployment rates stabilize.

The effects of lower skilled immigration on wages and employment tell a different story

compared to skilled immigration and it should be relevant in any future Census analysis. Fix and

Passel (1994) summarizes that lower skilled immigration has to some extent contributed to the

decline of low-skilled native unemployment and wages. However, the decline in low-skilled

native unemployment should be understood in the context of both immigration and trade

policies. Using Census data from 1980 and 1990, they estimated that an increase of ten to twenty

percent in the labor market due to immigration only decreases native employment participation

by less than one percent. Simon (1989) theorizes that the low skilled job market is not fixed but

flexible because of a multiplier effect in which one job creation leads to more job creation. Her

study of the Australian labor market provides an example which demonstrates that higher

 
6  
 
immigration should not be viewed as a sign of added pressure upon native workers, but rather a

sign of increased job opportunities and diminished unemployment for natives. Fairlie and

Meyer’s (2003) study of immigration and self-employment seems to provide indirect evidence of

the job multiplier effect. They find that self-employed immigrants do not push current self-

employed natives out of business but rather decrease the future self-employment opportunities

for natives. Therefore, self-employed immigrants can potential add more jobs while not

discouraging current job opportunities in a labor market.

Some other consequences of immigration that may be relevant to this research question

comes from Borja’s (1994) extensive paper on the economics of immigration in which he

examines the relationship of wages over a four decade period. What Borja’s findings indicate

over the period of 1960s to the 1990s is that US wages have been declining, which was not

necessarily due to immigrants accepting lower wages but rather a historic change in the US wage

structure. During the 1980s, there is evidence of an increase in the wage gap between skilled and

unskilled workers which affected native workers and immigrants differently (Levy and Murnane,

1992). As the rate of return for skilled labor increased, the wages of unskilled native and

immigrant workers declined. However, the relative wages of unskilled immigrants experienced a

higher percentage decline because immigrants’ skill levels often remain constant. Furthermore,

every decade in the US, from 1960s to the 1990s, an overall decline in the skills of successive

immigrants is noticeable.

The literature on immigration and economics shed greater light on the complexity of the

relationship between the two. Borja’s paper sets up a good starting point to examine with direct

evidence from the 1980 to 2008 Minnesota Population Center’s Integrated Public Use Microdata

Series (IPUMS), whether immigration depresses wages or whether some other factors are

 
7  
 
affecting wages that are independent of immigration (1994). The samples I will be using are the

following: 1980 1% sample, 1990 1% sample, 2000 1% sample, and the 2008 ACS sample

(Ruggles, et. al., 2009). In light of my hypothesis, the variables that will be the focus of my study

from these samples will be in the areas of immigration, occupation and employment, income,

and education. I will limit my study to the three decades between 1980 and 2008 because of the

lack of variables of interest before 1980.

CATEGORIZATION

The grouping of the observations into specific categories is necessary to compare and

contrast important trends and correlations. The purpose of this research focuses on full-time

workers. Full-time workers will be considered as those who identify themselves as working a

minimum of thirty-five hours per week and a minimum of forty weeks a year. The first

categorization is between US citizens and non-citizens. Distinguishing between citizens and non-

citizens is important because it provides an effective way of dividing workers born in the US and

workers born outside the US. Those workers who were born in the US or were born overseas to

parents of US citizenship are labeled as natives. Those workers who were not born in the US or

are naturalized citizens are labeled as immigrants.

The second categorization distinguishes between skilled and unskilled workers. Skilled

labor includes the following occupations: executive, administrative, and managerial occupations;

management related occupations; professional specialty occupations such as engineers,

scientists, therapists, teachers, librarians, religious workers, lawyers, writers; technical, sales and

administrative occupations such as technologists and technicians, and sales representatives.

Unskilled labor includes the following occupations: private household occupation, food

 
8  
 
preparation and service occupations, cleaning and building service occupations, personal service

occupations, construction and trade, precision metals working occupations, operators,

fabricators, and laborers, transportation and material moving occupations. The categorization of

skilled and unskilled workers of native and immigrant groups for 1980, 1990, 2000, and 2008 is

summarized in Table 1.

INCOME

Trends in Income among Skilled Workers

In general, the average income of skilled workers should reflect the positive returns on

educational investment or the value for increased productivity due to education. The income

trend among skilled workers is predicted to rise over time as the value for skilled workers

increase in a technologically advancing economy. These two assumptions are evident in Figure

1. The average incomes for all four time periods are adjusted to 2009 dollar amounts, increasing

for native skilled workers and immigrant skilled workers. Between 1980 and 1990, the average

income of skilled native workers increased by 6.4% and the average income of skilled immigrant

workers increased by 11.4%. Between 1990 and 2000, the rates were 15.5% and 10.5%

respectively. Between 2000 and 2008, the rates were 5.7% and 9.8% respectively. It is notable

that the average income of immigrant skilled workers increased faster than native skilled workers

for all decades except for the years between 1990 and 2000. A large average income gap is also

evident between native and immigrant skilled workers. Skilled immigrant workers seem to

receive a premium in average income. Between all four years, skilled immigrant workers receive

the largest premium in 2008 by earning an average income that is $9,301 higher than the average

income earned by skilled native workers.

 
9  
 
 An in-depth analysis of the data provides evidence for why a sizeable premium exists for

skilled immigrant workers. Two plausible reasons may account for the premium, greater work

experience and/or greater education attainment. Assuming that greater work experience must

entail more years of work experience and therefore, older age, the data shows no significant

differences in the average age of skilled immigrant workers compared to skilled native workers.

The regressions shown in Table 2.1 and Table 2.2 show that for native and immigrant skilled

workers, the effects of age (or what is assumed to be experience) and education on income are

comparable in all four decades. Holding education levels constant, both cohorts increase annual

income by roughly 1-1.5% on average due to experience. Unfortunately, the data is limited in

detailed information related to workers’ experiences such as on-the-job training, productivity

increases, and the acquirement of new skills.

However, focusing on education, a closer look at skilled immigrant workers shows that

their educational attainment may be a strong reason for the persistent gap in the income of skilled

workers of native and immigrant origins. First, Table 3 shows a regression of several variables

that quantify the correlation between income, education, and experience. Column 1 indicates

that, on average, skilled native workers earn 12% less income than skilled immigrant workers.

Column 2 indicates that an additional year of education increases the income of both cohorts by

10.2% on average. Column 3 indicates that by holding education constant, both cohorts increase

income by 1.3% on average with one additional year of age. The most important interpretation

remains in column 4. It indicates that while the increase in income for skilled native workers for

each additional year of education is 9.6%, skilled immigrant workers increase income by 0.5% in

addition to the increase in income for skilled native workers.

 
10  
 
 Looking more closely at the amount of education, I tabulated native skilled workers and

immigrant skilled workers by increasing years of education. Five categories were created: less

than high school, high school graduate, college graduate (bachelors), two years of graduate

school (masters), and more than two years of graduate school (doctorate). Within each category,

the average income was multiplied by the frequency of the category in order to observe the

proportional distribution of income according to educational attainment. The results are shown in

Figure 2. What the results show is that the persistent premium in the income of skilled immigrant

workers can be partially accounted for by the greater educational attainment of skilled immigrant

workers. The proportional distribution of income for skilled immigrant workers shows a greater

skew towards greater educational attainment than skilled native workers. This difference is large

in terms of master’s and doctorate level of education, each respectively account for an estimated

10% and 20% greater distribution than native workers.

Trends in Income Among Unskilled Workers

The average income of unskilled workers should differ from the average income of

skilled workers in several ways. The lack of educational investment and lower productivity of

unskilled workers will be reflected by lower average income compared to the average income of

skilled workers. Due to the lack of educational investment and lower productivity of unskilled

workers, the trend of the average income for unskilled workers in a technologically advancing

economy will not increase in the same way as skilled workers. This prediction is based on Levy

and Murane’s evidence from the 1980s, that the supply and demand for unskilled labor has

experienced a permanent structural shift, adversely affecting unskilled labor wages (1992).

Figure 1 provides the trend for unskilled native and unskilled immigrant workers that is

consistent with Levy and Murane’s findings in 1992 and projections beyond the 1990s.

 
11  
 
 Figure 1 shows that the average income of unskilled native workers and unskilled

immigrant workers remain relatively flat from 1980 to 2008. A slight increase in the average

income for both groups can be seen from 1990 to 2000, however, it also fell slightly for both

groups from 2000 to 2008. The stagnant average income for unskilled labor contrasts with

skilled labor. It provides clear indication that the average income of unskilled workers has not

kept up with inflation.

Similar to skilled workers, a gap exists between the average income of unskilled native

workers and unskilled immigrant workers. Unskilled native workers receive a premium in

average income compared to immigrant workers. This could be due to several factors such as

language, age, and sex, and education. Unlike skilled workers, each of these factors may

contribute to the income premium of unskilled native workers in comparison to unskilled

immigrant workers. There is a significant difference in the proportion of natives and the

immigrants workers who are English speakers which is shown in Figure 3. An observation of the

average age of each unskilled worker group shows that the average age of unskilled native

workers have increased each year from 37 in 1980 to 43 in 2008. On the other hand, the average

age of unskilled immigrant workers has remained slightly lower than native workers with the

exception of 1980.

The trend in the proportion of women among unskilled workers is of particular interest.

For native unskilled workers, the proportion of women in comparison to men has steadily

increased since 1990 and a similar trend is also observed among women of immigrant unskilled

workers. While the rate of participation for women in unskilled labor has remained below 30%,

native and immigrant women’s rate of participation in skilled labor rose above 40% by 1990 and

has steadily increased since. Skilled immigrant women workers continue to lag behind skilled

 
12  
 
native women in labor participation by an average of 4-5%. These general trends in women’s

workforce participation in both skilled and unskilled labor show in detail the widely held

perception that women are increasingly participating in the US workforce and their participation

is characterized more by skilled occupation rather than unskilled occupations.

The educational attainment distribution of unskilled workers of native and immigrant

origins provides additional reasons to justify the income gap between native and immigrant

unskilled workers. Figure 4 shows the distribution of educational attainment of native and

immigrant unskilled workers. Unskilled native workers are more likely to have completed high

school than immigrant workers. Unskilled immigrant workers are more likely to have dropped

out of school after completing elementary school or middle school than native workers.

Therefore, education and language seem to provide plausible reasons for the income gap between

unskilled native and unskilled immigrant workers.

EDUCATION

From looking at trends in income, it is clear that income is closely correlated to

education. Although this correlation was briefly explored in the section above, this section will

examine the trends in education to get a better picture of the changes in human capital

investment from 1980 to 2008. Then, the correlation between income and education will be

analyzed further concerning immigrants. Through this analysis, the results will be discussed as

they relate to my hypothesis, specifically, as they support or refute propositions 3 and 4.

Trends in Education among Skilled Workers

The trend in education among skilled workers is that skilled workers will have increased

in educational attainment in order to remain competitive in the skilled labor force. By looking at

 
13  
 
increase in skilled worker income and the advantageous distribution of immigrant educational

attainment, native skilled workers are expected to continuously catch up in educational

attainment in order to remain competitive. The evidence from the Census data shows that this

may be a possibility. Figure 5 shows the change over time of the educational attainment of

skilled native workers which shows an increasing trend towards greater education. Figure 6

shows a side by side comparison of native educational attainment trends and immigrant

educational attainment trends. A comparison of the trends shows a mixed result. The proposition

that native skilled workers are continuously catching up in educational attainment to immigrant

skilled workers does not seem to hold true by looking at the data. Rather, the trend seems to be

that native and immigrant workers are attaining higher educational levels simultaneously. The

increasing educational attainment of skilled immigrant workers may not be the cause of

increasing educational attainment of skilled native workers but other forces seem to be pushing

both groups to attain higher levels of education.

Trends in Education among Unskilled Workers

A general trend in education among unskilled workers also reveals surprising increases in

educational attainment, however, the increase in educational attainment does not seem to be as

high of a rate as skilled workers. The educational attainment of unskilled native workers,

although increasing somewhat beyond high school level, continues to peak at twelve years of

education. More unskilled native workers are completing elementary and middle school. The

educational attainment of unskilled immigrant workers was heavily distributed around an

average of nine to twelve years of education in 1980. The observable trend in the distribution has

been an increase in the standard deviation with noticeable increases in the lower distribution and

 
14  
 
higher distribution of educational attainment. By 2008, there is a significant increase in the

number of unskilled immigrant workers who have only completed up to nine years of education.

INCOME & EDUCATION AMONG IMMIGRANTS

Looking beyond the comparison between native and immigrant workers, the data lends

itself well to a focused analysis on immigrant trends in the US over the last four decades.

Immigrants who came to the US in 1980 are very different compared to the immigrants coming

to the US in 2008. Furthermore, the U.S. Census data reveals important trends in the education

and income of all immigrants that are worth exploring. In order to better understand changes in

the income and education of immigrants in this country, I perform two regressions, education

and income, for each year with the following model:

Income: logY = α +βN+ k

Education: E = α +βN+ k

The use of logY will allow me to provide an estimated percentage change in income for a given

value of N, the number of years an immigrant has lived in the US, in five year intervals.

Similarly, E will indicate the change in the number of years of education for a given value of N.

The coefficient α will be a constant term, a predicted value of income or years of education

when years of being in the US is zero. The variable k will be the residual error. The regression

will be restricted to the observations who are full-time immigrant workers, skilled and unskilled.

The results are presented in Tables 4.1 and 4.2.

 
15  
 
 In terms of income, a comparison of skilled and unskilled immigrants indicates that

unskillled immigrants have bigger gains in income for every five years of living in the US than

skilled immigrants. However, the gains in income for skilled immigrants have increase while the

gains in income for unskilled immigrants have decreased over the same time period. This

relationship seems to provide additional support for the results in Figure 6, in which recent

skilled immigrants are immigrating to the US with greater educational attainment than in the

past. By having greater educational attainment, these skilled immigrants may experience greater

increases in income as their residence in the US increases compared to past skilled immigrants

who did not have such educational attainment. While the correlation between income and

education for skilled immigrants have increased, the same correlation for unskilled immigrants

has seen little change, even a slight decrease.

In terms of education, a comparison of skilled and unskilled immigrants shows, for the

most part, negative correlations with income for all years. A raw interpretation of this evidence

can be that for every five years of living in the US, skilled and unskilled immigrants, on average,

decrease in education levels. But how can someone “decrease” their education? Referencing

back to Figure 6, I believe it is not that immigrants are losing their education but the more recent

immigrants to the US are arriving with greater education than the previous immigrants.

Therefore, the immigrants who arrived earlier and lived longer in the US, have on average, less

education than the recent immigrant cohorts. This is especially true of skilled immigrants but the

correlation is not as strong for unskilled immigrants, the correlation is not statistically significant

for 1990 and 2000.

CONCLUSION

 
16  
 
 This paper analyzes the two significant hypotheses about the effects of immigration on

wages and educational attainment. Although unemployment rates were of interest in the

comprehensive hypothesis, the limitations of the data and other factors that complicate in

isolating the effects of immigration on unemployment created substantial barriers to pursuing

this component of the hypothesis. However, the data proved to be valuable in answering the

other propositions. The third proposition was that the increased competition from highly skilled

immigrant workers would push skilled native immigrant workers to pursue higher educational

attainment to remain competitive in the skilled labor market. The fourth proposition was that the

wages of all unskilled workers would decline and the wages of skilled workers will remain

constant because of the greater supply elasticity of demand for skilled workers in comparison to

unskilled workers.

The data demonstrates that the third proposition does not necessarily hold true. Other

factors need to be taken into account before concluding that the highly skilled immigrants

contribute to the higher education levels of skilled natives through labor market competition.

However, the data points to similar trends in educational attainment which suggests other factors

are causing both cohorts to pursue more education. Although this paper held some assumptions

pertaining to the reason why skilled workers in technologically advancing economies seem to

increase in educational attainment, the exact motivation or reasoning behind the causality

remains a mystery.

In terms of the fourth proposition, my analysis of the U.S. Census has efficacy in

explaining the divergent movement of skilled and unskilled wages over the last thirty years in the

US. While unskilled worker income declined in real terms, the income of skilled workers

continued to increase over the time period. This evidence provides strong reasons to believe that
 
17  
 
the elasticity of demand for skilled workers is greater than the elasticity of demand for unskilled

workers because the income of skilled workers has risen considerably while the quantity of

skilled workers continued to increase.

Other general trends in income and educational attainment have been observed

throughout the analyses of this paper. The evidence that immigrants lower the income of

American workers does not seem to be based on fact. The Census data provides no indication

that greater immigration has had significant impact on wages. Other factors seem to play a

greater role in determining the income of American workers such as education and work

experience.

 
18  
 
 Works Cited

Borjas, George J. 1994. The Economics of Immigration. Journal of Economic Literature. 32

(December): 1667-1717.

Chiswick, Carmel U. 1989. The Impact of Immigration on Human Capital of Natives.

Journal of Labor Economics, 7 4:464-86

Fairlie, Robert W., and Bruce D. Meyer 2003. The Effect of Immigration on Native Self
Employment. Journal of Labor Economics, 21 3:619-50

Fix, Michael, and Jeffery S. Passel. 1994. Immigration and Immigrants: Setting the Record
Straight. Washington, D.C.: The Urban Institute.

Levy, Frank, and Richard Murnane. 1992. US. Earnings Levels and Earnings Inequality: A
Review of Recent Trends and Proposed Explanations. Journal of Economic
Literature, 30 3:1333-81.

Passel, Jeffrey S. 1986. Undocumented Immigration. Annals of the American Academy of

Political and Social Science, 487:181-200

Ruggles, Steven, Matthew Sobek, Trent Alexander, Catherine A. Fitch, Ronald Goeken, Patricia
Kelly Hall, Miriam King, and Chad Ronnander. 2009. Integrated Public Use Microdata
Series: Version 4.0 [Machine-readable database]. Minneapolis, MN: Minnesota
Population Center [producer and distributor].

Simons, Julian Lincoln. 1989. The Economic Consequences of Immigration, 2nd edition.
Ann Arbor: University of Michigan Press.

Torok, S.J., Huffman W.E. U.S.-Mexican Trade in Winter Vegetables and Illegal
Immigration. American Journal of Agricultural Economics. 68 2:246-60

 
19  
 
 Table 1
Categorization of Skilled and Unskilled Workers for Natives and Immigrants

Skilled  Native  Category  Summary  

 
Observations   Standard  
  (Weighed)   Mean   Deviation   Minimum   Maximum  

1980   13,523,300   0.004   0.065   0   1  

1990   13,731,227   0.008   0.089   0   1  

2000   19,064,450   0.010   0.099   0   1  

2008   22,570,020   0.012   0.108   0   1  

 
 
Skilled  Immigrant  Category  Summary  
 
Observations   Standard  
  (Weighed)   Mean   Deviation   Minimum   Maximum  

1980   928,500   2.409   0.492   2   3  

1990   1,203,651   2.417   0.493   2   3  

2000   2,330,006   2.419   0.493   2   3  

2008   3,359,101   2.390   0.488   2   3  

 
 
Unskilled  Native  Category  Summary  
 
Observations   Standard  
  (Weighed)   Mean   Deviation   Minimum   Maximum  

1980   15,096,300   0.003   0.052   0   1  

1990   14,979,499   0.006   0.077   0   1  

2000   17,171,936   0.007   0.081   0   1  

2008   16,561,781   0.010   0.098   0   1  

 
20  
 
 Unskilled  Immigrant  Category  Summary  
 
Observations   Standard  
  (Weighed)   Mean   Deviation   Minimum   Maximum  

1980   1,378,900   2.598   0.490   2   3  

1990   2,172,737   2.674   0.469   2   3  

2000   4,218,702   2.674   0.469   2   3  

2008   6,190,778   2.711   0.453   2   3  

 
 
Table 1 provides year-by-year number of observations for each worker sub-group for the relevant time periods. It
demonstrates the significant increases in immigration over the last thirty years. The categories for skilled and
unskilled occupations are elaborated in the Research Strategy/Design section.

Table 2.1
The Effect of Education and Experience on Income for Skilled Native Workers

  1980   1990   2000   2008  

Education  Level   0.0703***  0.0997***  0.0943***  0.120***  

  -­‐0.00083   -­‐0.00077   -­‐0.00077   -­‐0.00097  

Worker  Age   0.0163***  0.0124***  0.0121***  0.0113***  

  -­‐0.00015   -­‐0.00014   -­‐0.00014   -­‐0.00012  

Constant   8.986***   8.691***   8.863***   8.482***  

  -­‐0.015   -­‐0.014   -­‐0.013   -­‐0.017  

Observations   127962   128245   173780   225953  

R-­‐squared   0.11   0.17   0.12   0.11  

Standard  errors  in  parentheses      

***  p<0.01,  **  p<0.05,  *  p<0.1      

Table 2.1 provides year-by-year correlation between log income in relation to education and age for skilled native
workers. It demonstrates that education increasingly accounts for greater income percentage than experience. This
is assuming greater experience is related to increae in age.

 
21  
 
 Table 2.2
The Effect of Education and Experience on Income for Skilled Immigrant Workers

  1980   1990   2000   2008  

Education  Level   0.0815***  0.101***   0.0930***  0.111***  

  -­‐0.0033   -­‐0.0025   -­‐0.0019   -­‐0.0019  

Worker  Age   0.0146***  0.0145***  0.0123***  0.0112***  

  -­‐0.00068   -­‐0.00056   -­‐0.00043   -­‐0.00035  

Constant   8.911***   8.649***   8.926***   8.709***  

  -­‐0.063   -­‐0.045   -­‐0.034   -­‐0.036  

Observations   8603   10521   19834   30671  

R-­‐squared   0.1   0.2   0.15   0.13  

Standard  errors  in  parentheses      

***  p<0.01,  **  p<0.05,  *  p<0.1      

Table 2.1 provides year-by-year correlation between log income in relation to education and age for skilled
immigrant workers. It demonstrates that education increasingly accounts for greater income percentage than
experience. This is assuming greater experience is related to increae in age.

 
22  
 
 Table 3
Income Premium of Skilled Immigrant Workers

  1   2   3   4   5  

Citizenship   -­‐0.120***  -­‐0.0768***  -­‐0.0769***   0.0157   -­‐0.00727  

  -­‐0.0027   -­‐0.0026   -­‐0.0025   -­‐0.019   -­‐0.021  

Education  Level       0.102***   0.0970***   0.0963***   0.0962***  

      -­‐0.00039   -­‐0.00038   -­‐0.00041   -­‐0.00041  

Age           0.0134***   0.0134***   0.0135***  

          -­‐6.5E-­‐05   -­‐6.5E-­‐05   -­‐6.8E-­‐05  

Immigrant  X  Education  Level               0.00558***  0.00575***  

              -­‐0.0011   -­‐0.0011  

Immigrant  X  Age                   -­‐0.000609***  

                  -­‐0.00023  

Constant   10.99***   9.304***   8.812***   8.732***   8.753***  

  -­‐0.0026   -­‐0.007   -­‐0.0072   -­‐0.018   -­‐0.019  

Observations   725569   725569   725569   725569   725569  

R-­‐squared   0   0.09   0.14   0.14   0.14  

Standard  errors  in  parentheses        

***  p<0.01,  **  p<0.05,  *  p<0.1        

Table 3: A linear regression model that summarizes how education and age account for differences in log income
between skilled native and skilled immigrant workers using data from the U.S. Census over the following years:
1980, 1990, 2000, 2008.

 
23  
 
 Figure 1
Graph of Wages for Skilled and Unskilled Workers from 1980-2008

90000  
80000  
70000  
60000  
Skilled  NaXves  
Annual   50000  
Income   40000   Skilled  Immigrants  
in  2009   Unskilled  NaXves  
30000  
Dollars   Unskilled  Immigrants  
20000  
10000  
0  
1980   1990   2000   2010  
YEARS  

Figure 1 shows the trend of wages for skilled and unskilled workers. The income for all four years have been
adjusted to 2009 Consumer Price Index and top codes were adjusted accordingly as necessary.

Figure 2
Proportional Income Distribution of Skilled Workers by Education Category
0.5  
0.45  
Percentage   0.4  
Distribution  of   0.35  
Average   0.3  
Income  in   0.25  
Proportion  to   0.2   Skilled  Immigrant  
Observation   0.15   Skilled  NaXve  
0.1  
0.05  
0  
<High   High   Bachelors   Masters   Doctorate  
School   School  
Graduate  

Educational  Attainment  

Figure 2 provides evidence for the income gap between native and immigrant workers of skilled labor. The mean
income for per education category was multiplied by the number of observations per education category to obtain
the proportional distribution of income for each education category.

 
24  
 
 Figure 3
Percentage of Unskilled Workers’ English Fluency

100  
90  
80  
70  
Percentage  
60  
50  
40   Immigrants  
30   NaXves  
20  
10  
0  
Does  not   Yes,  speaks   Yes,  speaks   Yes,  speaks  Yes,  but  not  
speak   only  English   very  well   well   well  
English  

English  Fluency  
Figure 3 provides the level of english fluency for native and immigrant unskilled workers. This chart highlights one
possible reason for an income gap between unskilled native and unskilled immigrant workers.

Figure 4
Distribution of Educational Attainment of Unskilled Workers

70  

60  

50  
Percentage  
40  

30   NaXves  
20   Immigrants  

10  

0  
Completed   Completed   Completed   Some   Completed  
Elementary   Middle   High  School   College   College  &  
School   School   Beyond  

English  Fluency  
Figure 4 the percentile distribution of educational attainment for native and immigrant unskilled workers.

 
25  
 
 Figure 5
Trend in Educational Attainment of Native Skilled Workers

180000  
160000  
140000  
120000  
Number  of  
100000   1980  
Workers  
80000  
1990  
60000  
2000  
40000  
20000   2008  
0  
Middle   High  School   College  or   Graduate   Completed  
School  or   or  less   less   School  or   Graduate  
less   less   School  
Educational  Attainment  

Figure 5 shows the trend over time of the number of native skilled workers and the level of education they have
completed.

Figure 6
Comparison of Educational Attainment of Native and Immigrant Skilled Workers

60  

50  

40   Middle  School  or  less  

High  School  or  less  
30  
Percentage   College  or  less  
of  workers   20   Graduate  School  or  less  
Completed  Graduate  School  
10  

0  
1980   1990   2000   2008   1980   1990   2000   2008  

Native  Educational  Attainment   Immigrant  educational  Attainment  

Figure 6 shows the trend in the distribution of educational attainment of native and immigrant skilled workers.

 
26  
 
 Table 4.1
The Effects of Years in the US on Income and Education for Skilled Immigrant Workers

  1980   1990   2000   2008  

  Income   Education   Income   Education   Income   Education   Income   Education  

Years  in  the  US  (5  yrs)   0.0468***   -­‐0.215***   0.0819***   -­‐0.00664   0.0716***   -­‐0.0298**   0.0705***   -­‐0.0899***  
  -­‐0.0054   -­‐0.017   -­‐0.0045   -­‐0.016   -­‐0.0034   -­‐0.012   -­‐0.0028   -­‐0.008  
Constant   10.65***   16.53***   10.63***   16.50***   10.71***   16.51***   10.81***   17.29***  
  -­‐0.02   -­‐0.063   -­‐0.017   -­‐0.063   -­‐0.013   -­‐0.048   -­‐0.011   -­‐0.032  
Observations   8603   9285   10521   11206   19834   21077   30671   32134  

R-­‐squared   0.01   0.02   0.03   0   0.02   0   0.02   0  

Standard  errors  in  parentheses              
***  p<0.01,  **  p<0.05,  *  p<0.1              

Table 4.2
The Effects of Years in the US on Income and Education for Unskilled Immigrant Workers

  1980   1990   2000   2008  

  Income   Education   Income   Education   Income   Education   Income   Education  

Years  in  the  US  (5  yrs)   0.102***   0.157***   0.114***   -­‐0.0151   0.0957***   0.0156   0.0967***   -­‐0.0335***  
  -­‐0.0041   -­‐0.015   -­‐0.0027   -­‐0.014   -­‐0.002   -­‐0.01   -­‐0.0018   -­‐0.011  
Constant   9.786***   10.95***   9.766***   11.36***   9.821***   11.39***   9.784***   13.25***  
  -­‐0.014   -­‐0.051   -­‐0.0089   -­‐0.047   -­‐0.0068   -­‐0.034   -­‐0.0067   -­‐0.039  
Observations   13291   13789   19050   19981   36588   38967   40449   43932  

R-­‐squared   0.04   0.01   0.08   0   0.06   0   0.06   0  

Standard  errors  in  parentheses              

***  p<0.01,  **  p<0.05,  *  p<0.1              

Tables 4.1 & 4.2 show two univariate regressions for 1980, 1990, 2000, and 2008. The correlation between years of
living in the US (in 5 year intervals) with income and education are show for skilled and unskilled immigrants.

 
27  
 
 Tables and Figures Notes

Sample Selection:
1980 1% sample, 1990 1% sample, 2000 1% sample, 2008 ACS Sample

Data Source Citation

Steven Ruggles, Matthew Sobek, Trent Alexander, Catherine A. Fitch, Ronald Goeken, Patricia
Kelly Hall, Miriam King, and Chad Ronnander. Integrated Public Use Microdata Series:
Version 4.0 [Machine-readable database]. Minneapolis, MN: Minnesota Population Center
[producer and distributor], 2009.

 
28