A new series of maps by the Center for Immigration Studies based on Census Bureau data provides detailed information on the nation's immigrant population (legal and illegal) at the county level in 1990, 2000, and 2014. The analysis focuses on adults because they have the most immediate impact. Adults directly affect the job market as workers, impact politics as constituents and potential voters, and begin to reshape the culture in receiving communities as soon as they arrive. Census Bureau data, which includes legal and illegal immigrants, shows that growth in the adult immigrant population in some counties has been nothing short of astonishing, while other areas have seen little growth. The findings make clear that Washington may set immigration policy, but it is local communities that feel the impact.
Among the findings:
- In 1990, immigrants were at least 20 percent of the adult population (18-plus) in just 44 counties; by 2014 they were at least 20 percent of the adult population in 152 counties.
- In 1990, only one out of eight Americans lived in a county in which at least 20 percent of adults were immigrants; by 2014, nearly one in three Americans lived in such counties.
- Since 1990, the immigrant share of adults has more than quadrupled in 232 counties.1
- Examples where the immigrant share of adults more than quadrupled from 1990 to 2014 include:
- In Georgia: Stewart County, <1 percent to 23 percent; Echols County, 2 percent to 21 percent; and Gwinnett County, 6 percent to 32 percent.
- In North Carolina: Mecklenburg County, 4 percent to 17 percent; Durham County, 4 percent to 16 percent; and Duplin County, 2 percent to 15 percent.
- In Kansas: Scott County, 2 to 15 percent; and Hamilton County, 3 percent to 21 percent.
- In Nebraska: Colfax County, 3 percent to 30 percent; Dawson County, <1 percent to 24 percent; and Dakota County, 6 percent to 28 percent.
- In Minnesota: Nobles County, 2 percent to 24 percent; and Watonwan County, 3 percent to 13 percent.
- In Oklahoma: Texas County, 2 percent to 28 percent; and Harper County, 2 percent to 14 percent.
- In Virginia: Manassas Park City, 7 percent to 40 percent; and Loudoun County, 7 to 30 percent.
- In Texas: Garza County, 5 percent to 48 percent; and Dallam County, 3 percent to 17 percent.
- Other examples include: Buena Vista County, Iowa, 3 percent to 22 percent; and Jerome County Idaho, 4 percent to 22 percent.
- While the immigrant share of adults has often increased the most in counties with smaller populations, growth since 1990 has also been dramatic in many large counties with over a million residents:
- Dallas County, Texas: 13 percent to 29 percent;
- King County, Wash.: 11 percent to 25 percent;
- Clark County, Nev.: 11 percent to 27 percent;
- Alameda County, Calif.: 21 percent to 38 percent;
- Sacramento County, Calif.: 12 percent to 25 percent;
- Fairfax County, Va.: 18 percent to 37 percent; and
- Montgomery County, Md.: 22 percent to 40 percent.
Map 1: 1990 Immigrant Share of Adult Population
of Adult Population
Map 2: 2000 Immigrant Share of Adult Population
of Adult Population
Map 3: 2014 Immigrant Share of Adult Population
of Adult Population
between 2010 and 2014
Map 4: Increase in the Immigrant Share of Adult Population, 1990 to 2014
of Adult Population 1990 to 2014*
between 2010 and 2014
Method and Data Source
The data for this analysis comes from the 1990 and 2000 Census and the 2010 to 2014 five-year file of the American Community Survey (ACS). We refer to the five-year data as 2014 data, though technically it reflects the population 2010 to 2014. In almost all cases the data comes from American FactFinder on the Census Bureau website.2 We have four different maps. The first three maps show the immigrant share of the adult population in 1990, 2000, and 2014 differentiated into five categories.3 The fourth map shows the growth in the immigrant share of the adult population from 1990 to 2014 with seven categories. The percentage increase in the adult population is calculated in a straightforward manor. For example, if 10 percent of adults in a county were immigrants in 1990 and it increased to 15 percent by 2014, then the adult share of the population grew 50 percent.4 It is also important to note that in the growth map all counties in which immigrants are less than 5 percent of adults are included in the bottom category no matter how much the percentage increased. So, for example, a county that grew from 1 percent to 4 percent immigrant technically increased 300 percent, and we do not report those numbers. But it would be included in the bottom category in our growth map.5
In 2014 there were 3,142 "county equivalents" as they are referred to by the Census Bureau and other government agencies. (In 2013, there were still 3,143 counties because Bedford City Virginia ended its independence in that year and some publications still report the higher figure.) County equivalents include parishes in Louisiana, independent cities (primarily in Virginia), the District of Columbia, and a few other administrative areas that are considered equivalent to counties by the U.S. government. Our analysis only has data for 3,140 counties in 2014, not 3,142, because the Hoonah-Angoon Census Area and Skagway Municipality were combined in 2000, but were separated in 2014. We recombine them in the map for 2014 so we can make comparisons (Map 4).6 Our number is reduced by one additional county because we combine Wrangell City & Borough and Petersburg Borough in Alaska, which were one county equivalent in 1990 and 2000. By combining them we can make comparisons with 1990.
We do not have data for 1990 for a few counties. In these cases we compare growth 2000 to 2014 on our growth map. This is the case for Skagway Municipality & Hoonah-Angoon Census Area, Alaska; Denali Borough, Alaska; and Yakutat City and Borough, Alaska.
Generally, counties do not change their borders, but from 1990 to 2014 there were some changes and our analysis tries to deal with them by creating as much continuity as possible to allow for comparisons. Broomfield County, Colo., did not exist in 1990 or 2000 because it did not become a county until 2001. In 1990 and 2000 it is part of other counties; in 2014, 9.2 percent of its population was immigrant and is shown accordingly on the 2014 map. In 1990 and 2000, Bedford City and Clifton Forge City were independent cities in Virginia and the Census Bureau reported information for them separately. By 2014, these independent cities had merged with their surrounding counties — Bedford with Bedford County and Clifton Forge with Allegheny County. However, in all three time-periods we report these cities with their counties to create continuity in the data and allow for a comparison 1990 to 2014.
1 As explained in our methods statement, figures exclude counties in which immigrants were less than 5 percent of the adult population in 2014. So, for example, an increase from 1 percent to 4 percent is a quadrupling, but because the population is still under 5 percent in 2014 it is not including in our count of counties that increased four-fold.
2 We were not able to obtain data for 1990 from FactFinder. The immigrant population data comes from tables made on the Data Ferret website from the 1990 Summary File 3. We were not able to obtain information by country for each county in 1990. We were able to use the public-use file of the 1990 Census to generate the top three sending countries for 402 counties in 1990.
3 The first category is under 5 percent, with rounding to the hundredths place. So, for example, all counties where adult immigrants are 4.94 percent of the population or less are in the bottom category; those counties that are 4.95 percent to 9.94 percent are in the second category; those that are 9.95 to 14.94 percent are in the third category; and so on. The last category is those counties in which the immigrant population is 19.95 percent or more.
4 As in the other maps, values are rounded based on the hundredths place. So, for example, 49.4 percent growth is rounded down to 49 percent, while 49.5 percent growth is rounded up to 50 percent.
5 There were 18 counties in 1990 with zero immigrants, so it is not possible to calculate growth in those counties from 1990 to 2014. But in all but one of these cases, immigrants are still less than 5 percent of the adult population in 2014 and so they are in the bottom category in our growth map. The one exception is Daggett County, Utah. Although zero in 1990, the immigrant population was relatively large by 2014. We include this county in our growth map by measuring growth in the immigrant adult share of the population from 3.5 percent in 2000 to 8.1 percent in 2014.
6 We do not have data for these areas in 1990.