Immigrant Labor: Facts And Trends 2011-2015

Summary

From 2011 through 2015, the percentage of workers who were born outside the US increased modestly from 16.7% to 17.3%. The percentage varies tremendously by state, however, from a high of 34% in California to a low of 2% in West Virginia. Where new immigrants come from also varies significantly by state.

New immigrants earn significantly less than would be expected based on  their age, education, occupation, sex and race characteristics (“AEOSR” average wage). The greatest wage discount is for educated immigrants. Most of the wage discount vanishes quickly within a few years as immigrants establish US work experience.

States with a high cost of living and high average wages have the highest percentage of non-US born workers, and they also attract the most new immigrants. This suggests that immigrants provide an economic benefit to residents of high-cost/high-wage states because they help keep the costs of products and services in those states from rising even further.

Political support for immigration is highest in states that have the highest percentage of non-US born workers. This support may be correlated with the economic value of immigrants in high-cost/high-wage states relative to states where wages are low. Support may also related to greater familiarity with immigrants and/or simply the greater likelihood that survey respondents in those states may be immigrants.

From an economic perspective, immigration has both benefits (higher economic growth and lower product costs) and costs (displacing non-immigrant workers and higher costs for social services). In a non-economic context, there are value-based arguments for immigration because America has traditionally identified itself as a land of opportunity and a destination for those seeking freedom. Those who argue for restraining immigration are often concerned about cultural assimilation.

Immigrant Worker Characteristics

This discussion is based on analyzing data from US Census data made available by the University of Minnesota through IPUMS-USA.[1] The subset of data I used was all persons who had wage income (INCWAGE>0) for years 2011 to 2015. I report on characteristics of all non-US born workers as well as two subsets: all workers who are not citizens, and those new non-citizens who immigrated during 2011 - 2015.

New immigrants are somewhat younger -- with a mean age of 32 years compared to the mean age of 42 years for all US workers. I defined new immigrants as those who have been in the US for 1 year or less. The median non-citizen worker has been in the US for 12 years, while 18 years is the median number of years in the US for all workers not born here.

New immigrant workers had a “bimodal” educational makeup. They had a higher percentage of “No High School” than all workers, but also a higher percentage of college degrees and beyond. Among new immigrants, the high number of college+ (Bachelor’s degree and beyond) educated appears to be a change in trend, however, as all non-citizens and all foreign born workers show a lower percentage of college+ than all workers.

Table 1: Educational Composition of Immigrant Workers

Education	New Non-Citizens	All Non-Citizens	All Foreign Born	All Workers
No High School	16.7	35.6	24.3	10.2
Regular high school diploma	14.3	22.1	19.8	21.8
GED or alternative credential	1.1	2.0	2.3	3.3
Some college; but less than 1 year	3.4	3.0	3.8	6.8
1 or more years of college credit; no degree	8.5	9.7	12.1	17.8
Associate's degree; type not specified	3.5	4.0	6.3	8.7
Bachelor's degree	29.0	12.7	18.0	20.1
Master's degree	15.2	7.3	8.8	8.1
Professional degree beyond a bachelor's degree	3.1	1.3	2.2	2.0
Doctoral degree	5.1	2.2	2.3	1.3

Figure 1 illustrates that new immigrants start off making less than their AEOSR-expected income (the expected AEOSR income ratio would have a value of 1.0). As a whole, all new immigrants make about 72% of what they would be expected to earn. Those with a Doctoral degree only made 64% of the wage they would be expected to make. As shown, those relative wage differentials are erased over time.

Figure 1: Immigrant Wage Discount By Years In US

The initial wage differential may be partly due to the inability of some new immigrants to speak English. Table 2 shows 37% of new immigrants do not speak English or speak it “Not Well”. For college+ workers, that number is 11% (and drops to 7% for those who have been in the US for 7 years). The percentage that does not speak English well drops to 23% for all non-US born workers (who have been in the US for longer).

The initial wage differential is erased over time, presumably because immigrants gain bargaining power with more US work experience on their resume.

Table 2: Immigrant Ability To Speak English

Immigrant Workers By Country of Origin

Table 3 shows the country of origin has changed modestly with India and China playing a larger role as a source for new immigrants during 2011-2015.

Table 3: Immigrants By Country of Origin

Rank	Birthplace	% New Non-Citizen	Birthplace	All Non-Citizen	Birthplace	All Non-US Born
1	India	15.3	Mexico	40.0	Mexico	27.6
2	Mexico	14.4	Central America	11.0	Central America	8.2
3	China	8.5	India	6.7	India	6.8
4	Central America	5.1	West Indies	5.2	West Indies	6.8
5	Canada	3.5	China	4.3	China	5.3
6	West Indies	3.4	Philippines	2.9	Philippines	5.0
7	Philippines	3.4	Cuba	2.0	Vietnam	3.0
8	Germany	2.7	Canada	1.8	Germany	2.6
9	Japan	2.5	Korea	1.5	Other USSR/Russia	2.3
10	Other USSR/Russia	2.4	Other USSR/Russia	1.4	Korea	2.3

Composition of new immigrants varies quite a bit when viewed at the state level. Tables 4 and 5 compare Florida (with a significant non-US born workforce) and South Dakota (with a small non-US born workforce).

Table 4: Florida New Immigrants By Country of Origin

 

Table 5: South Dakota New Immigrants By Country of Origin

What Determines Which States Immigrants Move To?

Table 6 shows the top and bottom states based on new immigrant migration. While it is tempting to conclude that immigrants choose states in which there are already a high percentage of immigrants (going where friends and relatives have already settled) or coastal states that are easier to get to, statistical modelling suggests that is not true.

A state's average wage, cost of living and employment growth are the statistically significant predictors of which states new immigrants move to. After accounting for those three factors, the percentage of non-US born workers in a state has no statistically significant power to predict new immigration. This suggests that immigrants are not simply following other immigrants. It indicates that economic demand for immigrants and their value to a state's economy drives immigration patterns.

Table 6: Top 10, Bottom 10 States By New Immigrants As Percent Of Workforce

State	% New Immigrants	% Non-Citizens	% Non-US Born	Avg. Wage: All Workers	Cost of Living Index	Emp. Growth 2010-2015	PRRI Support For Immigrants
Hawaii	1.1%	8.6%	22.7%	$45,697	168	10.3%	65
Washington	0.7%	8.7%	18.1%	$50,180	105	13.2%	68
Massachusetts	0.7%	8.5%	18.8%	$56,163	133	10.8%	66
Virginia	0.7%	7.3%	16.3%	$53,158	101	7.3%	62
Utah	0.7%	6.7%	11.8%	$41,446	93	18.9%	63
Texas	0.7%	12.8%	21.5%	$45,778	91	16.6%	61
California	0.6%	16.9%	34.4%	$51,112	136	14.7%	64
Colorado	0.6%	6.9%	13.0%	$48,990	104	16.2%	63
New York	0.6%	12.0%	27.5%	$53,986	131	9.5%	67
Florida	0.6%	11.2%	25.2%	$42,561	98	15.3%	62
Maine	0.3%	1.6%	4.7%	$40,509	111	3.2%	61
Indiana	0.3%	3.6%	6.2%	$40,256	90	10.5%	58
Ohio	0.3%	2.4%	5.3%	$42,146	94	8.5%	58
South Dakota	0.3%	2.0%	4.0%	$37,885	104	8.2%	46
Wyoming	0.3%	2.4%	4.5%	$43,564	92	3.6%	59
Arkansas	0.3%	4.2%	7.0%	$37,960	87	5.6%	55
Wisconsin	0.3%	3.2%	6.1%	$41,784	97	7.2%	64
Alabama	0.2%	2.8%	5.3%	$40,363	91	5.5%	54
Mississippi	0.2%	1.7%	3.4%	$36,518	86	5.3%	59
West Virginia	0.2%	0.8%	2.4%	$38,711	93	1.5%	54

Figures 2 and 3 show the relationship between new immigration and cost of living and the average income of all workers by state respectively.[2] Both explain approximately 45% of new immigration variation by state, but the two factors are highly correlated.

Figure 2: New Immigrant Migration vs. Cost of Living

Figure 3: New Immigrant Migration vs. Average State Wage

When these two factors are combined with state employment growth (which explains about 10% of net immigration variations by state) as shown in Figure 4, these three factors together explain about 60% of new immigration variations by state.

Figure 4: New Immigrant Migration vs. Predicted By COL, Avg. Worker Wage and State Employment Growth

Because new immigrants are less expensive to hire relative to their credentials, they are very valuable in high-wage states because immigrants' lower wages help restrain the (also) high cost of products and services. New immigrants are less valuable (in a relative sense) in states that already have low cost labor (i.e., where average worker wages are already low). Strong state employment growth -- which increases the demand for labor -- magnifies the value of new immigrants.

One might imagine political opposition to immigration would be highest in states with the highest percentage of immigrants because "native" workers in those states face the most competition from immigrants. The data indicates that the opposite is true: political support for immigrants is higher in states that have a high percentage of non-US born workers.

The survey group PRRI asked people if they believed illegal immigrants should have a path to citizenship if they met certain requirements. Figure 5 show the relationship between the percentage of people who feel there should be a path to citizenship versus the percentage of the state workforce that is not born in the US.

Figure 5: Political Support vs. Percentage of Workforce Not Born In US

It appears residents of high immigration states are more focused on immigrants' role in restraining those states' high cost of living than they are on the impact immigrants have on wages and job availability. Support in states with significant non-US born workers may also be driven by greater familiarity with immigrants and/or the increased likelihood that survey respondents are immigrants or are related to immigrants.

The Debate Over Immigration

From an economic perspective, immigration has both costs and benefits. Determining the net advantage can be difficult and can depend on immigrant characteristics (i.e., high skill versus low skill; English speaking versus non-English speaking, etc.). Arizona has been somewhat of a “test case” for measuring the impact of reducing low-skilled illegal immigration. After passing a tough series of anti-immigration laws, Arizona’s illegal immigrant population fell by approximately 40% from 2007 to 2012. Although difficult to separate from the effect of the national Great Recession that also occurred during that period, researchers believe the decline in the immigrant population had some negative economic effects like lower state GDP growth, lower employment, and lower home prices. There were some offsets, however, in increased wages for US-born Arizona workers, and lower social services costs related to education and health care. Notably, the Arizona case does not speak to the net benefits or costs related to high-skilled immigration desired by health care and high-tech companies that generally lobby for immigration.

In a non-economic context, the debate over immigration arises from differences in cultural values. America is  a country which was founded by immigrants. Many have continued to see it as a “country of immigrants” as witnessed by the poem inside the pedestal of the Statue Of Liberty which reads: “Give me your tired, your poor, your huddled masses yearning to breathe free.”

On the other hand, those who wish to restrict immigration also cite cultural values: primarily a desire to preserve American culture as it exists today. They are concerned about the difficulty of cultural assimilation when the number of new immigrants is large. They also cite recent significant periods of cultural dislocation within European countries that have seen large waves of uncontrolled immigration.

For those interested in further exploration of this topic, I highly recommend listening to a civil discussion between Harvard University’s George Borjas and Stanford University’s Hoover Institute Research Fellow Russ Roberts.

Transparent and reproducible: All of the labeled Figures and Tables can be generated by using the free, publicly-available R program and the R code available in “International Migration.r” on github to analyze the publicly available data obtainable from the links in the article.

[1] IPUMS-USA, University of Minnesota, www.ipums.org. The data is provided freely but is subject to their licensing restrictions.

[2] Alaska is omitted from the data used to estimate relationships affecting net retiree migration as Alaska has unique income characteristics.