top of page

Measuring Economic Integration

We measure economic outcomes (including income, unemployment, poverty, working poor, full time work, homeownership, and rent burden) of foreign-born New Yorkers to account for the effects of nativity status, race, gender and the intersections of these forces in the lives of foreign-born New Yorkers. Each comparative analysis controls for educational attainment levels and English proficiency. The comparative analyses are presented first in all education levels in general and then in three different education levels of (1) high school diploma and below, (2) some college and associate degree, and (3) bachelor’s degree and above to account for the impact of education in the comparisons.

 

Click on a comparative analysis below to see results by county and region:

(1) Effects of nativity status: Comparison of outcomes of foreign-born and native-born New Yorkers.

(2) Effects of race: Comparison of outcomes of foreign-born of color and foreign-born Whites.

(3) Effects of gender: Comparison of outcomes of foreign-born females and foreign-born males.

(4) Effects of nativity status and race: Comparison of outcomes of foreign-born of color and native-born Whites.

(5) Effects of nativity status and gender: Comparison of outcomes of foreign-born females and native-born males.

(6) Effects of nativity status, race and gender: Comparison of outcomes of foreign-born females of color and native-born White males.

(7) Effects of language proficiency: Comparison of outcomes of foreign-born Limited English proficient and foreign-born English proficient New Yorkers. 

 All analyses use American Community Survey Data. Data for the last 10 years (2012-2021).

IMMIGRANT INTEGRATION INDEX

To offer insights into where regions and counties relatively stand within New York State, the index score for each region is calculated based on the standardization process of getting a z-score for each region and county. Based on the differences in the above seven economic outcomes between foreign-born and native-born New Yorkers, we get a standardized score for seven outcomes using the mean and standard deviation of each outcome across the state and then average these scores for seven scores to compose one general index score indicating the overall disparity between foreign-born and native-born New Yorkers in each region. As some outcomes (e.g., personal income, full-time work, and homeownership) are positive while the rest (e.g., poverty, unemployment, working poor, and rent burden) are negative, we reverse the standard scores for positive outcomes, so that higher index scores can be interpreted as positive and lower scores can be interpreted as negative. As a result, from the score zero, where the disparity of a region/county equates to the state average, higher scores indicate foreign-born New Yorkers are relatively faring better, and lower scores indicate they are relatively faring worse in other regions/counties in New York.

NOTES ON METHODOLOGY

We focus on seven major economic outcomes of New Yorkers and their households: personal income, % of people under poverty, % of people unemployed, % of people working full-time, % of working poor, % of homeownership, and % under rent burden. To understand how well immigrants fare in New York, we estimate the disparity between foreign-born and native-born New Yorkers by comparing the above economic outcomes between the two groups while controlling for nativity status, race/ethnic minority status, gender, language, and education.

New Yorkers in our analysis are from 16 to 64 years old and are proficient in English except for those (limited English proficiency) in the comparison focusing on the most vulnerable population in the state. New York State’s ten geographic regions and counties (groups of counties where the estimation is not available within a single county) are used in the comparisons for the above economic outcomes and the calculation of the standardized index scores representing the immigrant integration of each region and county.

The data used in the comparisons and the index score calculations are the American Community Survey (ACS) 1-year Public Use Microdata Sample (PUMS) from 2012 to 2021 collected and managed by the US Census.

We measure the economic outcomes for each indicator using the following protocols:

  1. For total income, we average personal income after adjusting to constant dollars across the groups of our comparison.

  2. For the percentage of people under poverty, we use the Income-to-poverty ratio of 150 as our threshold to determine the state of under poverty. In each comparison, we estimate the percentage of people living in poverty by groups of our interests for comparison in each region/county.

  3. For the percentage of people unemployed, we calculate the percentage of people who are unemployed by groups of our interests for comparison in each region/county.

  4. For the percentage of people working full-time, we calculate the percentage of people who work 35 hours or more per week by groups of our interests for comparison in each region/county.

  5. For the percentage of working poor, we calculate the percentage of people who work full-time (35+ hours) but still are in poverty by groups of our interests for comparison in each region/county.

  6. For the percentage of homeownership, we calculated the percentage of households residing in homes they own within groups in our comparison. As we measure this outcome at the household level, unlike our previous individual-level outcomes, nativity status, race/ethnic minority status, and gender are determined by those of the reference persons of a household.

  7. For the percentage under rent burden, we calculate the percentage of households paying gross rent more than 50% of their household income within groups in our comparison. We also use household-level information to measure this outcome, like homeownership. Thus, nativity status, race/ethnic minority status, and gender are also determined by those of the reference persons of a household.

All outcomes are measured after weighting the responses using the replicate weights provided in the survey data.

bottom of page