Retirement Income Among American Indians and Alaska Natives in the American Community Survey

by
Research and Statistics Note No. 2015-01 (released February 2015)

John Murphy is a social scientist in the Office of Retirement Policy, Office of Retirement and Disability Policy, Social Security Administration (SSA). Brent Huggins is a former intern with SSA, sponsored by the Washington Internships for Native Students (WINS).

Acknowledgments: The authors thank Kathleen Romig, Dale Kintzel, and Bill Davis for their insightful comments and ideas. They also thank Anna Holster for providing very important subject information and expertise.

The findings and conclusions presented in this note are those of the authors and do not necessarily represent the views of the Social Security Administration or any other author affiliations.

Summary

Selected Abbreviations
ACS American Community Survey
AIAN American Indian and Alaska Native
DI Disability Insurance
OASI Old-Age and Survivors Insurance
PUMS Public Use Microdata Sample

This study uses data from the 2005–2009 American Community Survey (ACS) Public Use Microdata Sample (PUMS) to analyze annual retirement income of the American Indian and Alaska Native (AIAN) population. The overview compares annual retirement income among AIAN respondents with that of other race groups and includes separate analyses for AIANs of single-race and multiple-race backgrounds. Average annual retirement income among all AIAN respondents was significantly lower than it was for blacks. For example, mean annual retirement income (as defined in the ACS PUMS) was $4,600 for all AIAN respondents, compared with $5,800 for black respondents. However, this difference and others that are apparent in our single- and multiple-race AIAN analyses primarily reflect the characteristics of economically vulnerable single-race AIANs.

Introduction

The American Indian and Alaska Native (AIAN) population faces substantial economic challenges, many of which can play a role in the financial resources available in retirement. For example, research suggests that AIANs have lower average wages and higher rates of disability, both of which can reduce the amount of future income available through pensions, savings, and Social Security (Smith-Kaprosy, Martin, and Whitman 2012). However, little research has specifically addressed retirement income among AIANs.

This note contributes to our understanding of retirement security within the AIAN population by using the 2005–2009 ACS PUMS to compare retirement income of AIANs with that of whites and blacks in the U.S. population. This analysis provides separate findings for single- and multiple-race AIANs to identify retirement income differences within the overall AIAN population.

Our findings show that AIANs had significantly less combined retirement income (including Social Security) than did whites and blacks, as measured by the ACS PUMS. Retirement income (excluding Social Security) among the AIAN population also lagged behind that of whites and blacks in the survey. Typically, AIANs had lower annual retirement income (by between $1,200 and $2,400) than did members of the other two race groups. The gap in combined retirement income was particularly pronounced between single-race AIANs and multiple-race AIANs, reaching a disparity as high as $4,300 per year.

Data and Methodology

Data limitations are a substantial obstacle to research on AIANs. The AIAN population (alone or in combination with another race) constitutes less than 2 percent of the U.S. population, and the sample size in many surveys is inadequate for detailed analyses (Norris, Vines, and Hoeffel 2012). The ACS offers a promising resource for examining AIANs because it randomly surveys over 3 million housing units per year and oversamples American Indian reservations (HHS 2006).

To increase the accuracy of our population estimates, we use a multiyear (2005–2009), pooled version of the ACS PUMS. The 2005–2009 ACS PUMS file incorporates responses from the 2005, 2006, 2007, 2008 and 2009 1-year ACS PUMS files, with no duplicate household addresses included over that sampling period (Census Bureau 2010b). To focus on the retirement population, we restrict the analysis to individuals who were at least age 62 and no longer in the workforce (Gendell and Siegel 1992). After restricting the data file within these parameters, we are left with a total unweighted sample of approximately 2.2 million individuals for this study.

In this analysis, we use three categorizations to identify the AIAN population: (1) AIAN, (2) single-race AIAN, and (3) multiple-race AIAN. These variables are defined as follows:

  1. AIAN.1 A binary variable listing American Indian or Alaska Native identification based on self-reporting. This variable, taken directly from the PUMS file, includes all respondents who selected American Indian or Alaska Native alone or in combination with another race (Census Bureau 2013).
  2. Single-race AIAN.2 A binary, self-reported variable indicating that the respondent selected American Indian or Alaska Native alone for his or her race identification.
  3. Multiple-race AIAN.3 A binary, self-reported variable indicating that the respondent selected American Indian or Alaska Native and one or more other major race group(s) for his or her race identification.

We disaggregate the AIAN population into multiple-race and single-race categories because multiracial backgrounds are particularly prevalent among AIANs. In the 2010 census, 44 percent of the overall AIAN population reported more than one race, and this group is growing at a higher rate than the single-race AIAN population (Census Bureau 2012). Different racial backgrounds among AIANs may point to different experiences that can affect retirement security. For example, persons who are AIAN alone are more likely than AIANs with multiracial backgrounds to live in reservation communities, where labor force opportunities may be particularly scarce because of geographic location and other factors (Perry 2012).4 However, even within our segmented AIAN designations, a wide variety of social, historical, political, and cultural experiences exist. Readers should be aware that the diversity and heterogeneity of AIAN communities and tribes (Norris, Vines, and Hoeffel 2012) may reduce the ability to generalize the results to all AIAN subgroups.

Table 1 presents some basic characteristics for the race groups we analyze, which again comprise respondents aged 62 or older and out of the workforce. The 22,812 respondents in our analysis sample who identified themselves as AIAN represented 1 percent of our overall sample population. Among that group, 12,040 identified themselves as single-race AIANs, and 10,772 identified as multiple-race AIANs.

Table 1. Selected characteristics of the analysis population (aged 62 or older and out of the workforce), by race, 2005–2009
Race Analysis population Median age Median years of education Married (%) Women (%)
Number Percent
Total 2,197,098 100.0 74.3 12.2 63.0 60.0
White 1,905,936 87.0 74.5 12.4 59.0 59.0
Black 268,350 12.0 73.4 12.1 37.0 63.0
AIAN 22,812 1.0 72.3 12.0 47.0 60.0
Single-race 12,040 0.5 72.3 11.4 47.0 56.0
Multiple-race 10,772 0.5 72.2 12.8 48.0 60.0
SOURCE: Authors' calculations using the 2005–2009 American Community Survey Public Use Microdata Sample.
NOTE: AIAN = American Indian and Alaska Native.

To assess retirement income in the ACS, we analyze the following three financial measures, all of which are adjusted to 2009 dollars:

A few caveats are worth noting with regard to these variables. First, all of the values are based on self-reporting from respondents, which may be imprecise. For example, past research suggests that individuals are prone to underreporting income from Social Security (Olsen 2002). Second, those values only represent current income from the listed sources, but not the value of assets such as savings accounts and housing; they provide an incomplete picture of wealth and well-being in retirement. Third, the Social Security income variable in the ACS does not differentiate between benefits received through Social Security's OASI program and those received through the agency's DI program. Respondents in our analysis sample who were still under the full retirement age (those aged 65–66 for the included birth cohorts) may have been receiving DI benefits.

These limitations in financial reporting are significant, but necessary, trade-offs for analysis using the ACS PUMS to obtain a sufficient AIAN sample size. However, these shortcomings highlight the need for additional data and future research to understand fully the state of retirement income and well-being among AIANs.

Results

Table 2 (retirement income), Table 3 (Social Security income), and Table 4 (combined retirement income) provide mean, lower quartile, median, and upper quartile values for each race category in the study. The values are rounded to the nearest hundred, and the results are tested for statistical significance relative to white respondents. The accompanying chart to this note summarizes the combined average annual values for each of the three financial measures. Because of the constraints associated with these variables, the actual dollar amounts should not read as definitive descriptions of retirement income or economic well-being among AIANs or other race groups. Our focus is on the relative values for these variables as reported in the ACS PUMS.5

Annual Retirement Income

AIANs reported an average of $4,600 in annual retirement income (Table 2). This was $2,400 less than that of white respondents and $1,200 less than that of black respondents. However, this difference was largely driven by single-race AIANs who reported an average of $4,300 in retirement income, compared with $5,800 reported by multiple-race AIANs. Retirement income among multiple-race AIANs was lower than that of whites, but was essentially the same and not statistically different from that of blacks.

Table 2. Comparison of annual retirement income of the analysis population (aged 62 or older and out of the workforce), by race, 2005–2009 (in thousands of 2009 dollars)
Race Mean Lower quartile Median Upper quartile
White 7,000  0 0 8,400
Black 5,800* 0 0 7,200
AIAN 4,600* 0 0 4,800
Single-race 4,300* 0 0 4,900
Multiple-race 5,800* 0 0 6,900
SOURCE: Authors' derivation using the 2005–2009 American Community Survey Public Use Microdata Sample.
NOTES: Values are rounded to the nearest hundred.
AIAN = American Indian and Alaska Native.
* = statistically significant at p ≤ 0.05 when compared with whites.
The percentage of respondents with no annual income (using derivation measure) follow: whites, 58 percent; blacks, 62 percent; AIANs, 69 percent; single-race AIANs, 65 percent; and multiple-race AIANs, 58 percent.

Annual Social Security Income

AIAN respondents reported an average of $7,600 in annual Social Security income (Table 3), which was $2,200 lower than that of white respondents and $300 lower than that of black respondents (with the latter not being statistically significant at the 95 percent level). Again, this pattern was largely driven by single-race AIANs, as there was considerable variation in average annual Social Security income among AIANs of single- and multiple-race backgrounds. Multiple-race AIANs reported an average of $2,900 more in annual Social Security income than did single-race AIANs ($9,100 and $6,200, respectively). Multiple-race AIAN respondents had an average annual Social Security income that was higher than that of black respondents by $1,200, but lower than that of white respondents by $700.

Table 3. Comparison of annual Social Security income of the analysis population (aged 62 or older and out of the workforce), by race, 2005–2009 (in thousands of 2009 dollars)
Race Mean Lower quartile Median Upper quartile
White 9,800  5,800  9,900  13,900 
Black 7,900* 3,600* 7,800* 11,600*
AIAN 7,600* 4,200* 8,100* 9,200*
Single-race 6,200* 3,100* 7,300* 11,300*
Multiple-race 9,100* 5,200* 9,000  12,900*
SOURCE: Authors' derivation using the 2005–2009 American Community Survey Public Use Microdata Sample.
NOTES: Values are rounded to the nearest hundred.
AIAN = American Indian and Alaska Native.
* = statistically significant at p < 0.05 when compared with whites.

Annual Combined Retirement Income

AIAN respondents had an average of $12,200 in annual combined retirement income, which was $4,600 less than that of white respondents and $1,500 less than that of black respondents (Table 4). Among AIAN subgroups, single-race AIANs had less annual combined retirement income than did those of multiple-race backgrounds. Single-race AIANs had an average of $10,600 in annual combined retirement income, while the comparable figure for multiple-race AIAN respondents was $14,900. Annual combined retirement income among multiple-race AIANs was lower than that of whites, but higher than that of blacks.

Table 4. Comparison of annual combined retirement income of the analysis population (aged 62 or older and out of the workforce), by race, 2005–2009 (in thousands of 2009 dollars)
Race Mean Lower quartile Median Upper quartile
White 16,800  7,200  12,500  21,000 
Black 13,700* 6,000* 10,000* 17,500*
AIAN 12,200* 5,400* 10,800* 16,200*
Single-race 10,600* 4,900* 9,000* 15,300*
Multiple-race 14,900* 6,700  11,400* 18,900 
SOURCE: Authors' derivation using the 2005–2009 American Community Survey Public Use Microdata Sample.
NOTES: Values are rounded to the nearest hundred.
AIAN = American Indian and Alaska Native.
* = statistically significant at p < 0.05 when compared with whites.
Chart.
Comparison of average annual retirement incomes of the analysis population (aged 62 or older and out of the workforce), by race, 2005–2009 (in thousands of 2009 dollars)
Bar chart displaying mean income values from Tables 2, 3, and 4.
SOURCE: Authors' derivation using the 2005–2009 American Community Survey Public Use Microdata Sample.

Conclusion

AIANs had significantly less combined retirement income than did the other groups analyzed in this study, as measured by the 2005–2009 ACS PUMS. Retirement income among AIANs also lagged behind the other groups we examined in the survey. AIANs had an average of $4,600 less combined retirement income than did whites and $1,500 less than did blacks. The difference is especially stark when comparing singe-race AIAN respondents, who had an average of $6,200 and $3,100 less combined retirement income than did whites and blacks, respectively.

This note provides preliminary analysis of annual retirement income among AIANs using the 2005–2009 ACS PUMS. Our analysis suggests there is a sizable annual combined retirement income gap between AIAN respondents and the other groups under study.

Because of the limitations in the retirement income measures available in the ACS PUMS, it would be beneficial for other financial surveys that have broader and more sophisticated financial-asset variables to conduct oversampling of AIANs to gain a better picture of retirement security, savings and investments, and other household-consumption information.

Additional data are also required to better identify retirement income differences within the two AIAN subgroups. Although we were able to disaggregate single-race and multiple-race AIANs, the overall AIAN population is heterogeneous, and characteristics—such as whether an individual lives on or near a reservation, trust land, or in an urban location—may be important predictors of retirement income and well-being among this group. Additionally, this note is descriptive only, so we do not determine specific effects of key variables. Consequently, the literature would be greatly improved if future research examined AIANs further to determine whether there is an independent effect associated with being of AIAN descent.

Our preliminary analysis suggests that AIANs typically have lower retirement incomes than the overall U.S. population, and single-race AIANs may be particularly at risk of economic insecurity. However, this research is meant to provide a baseline for researchers and to inform policymakers. More exploration and research is needed to investigate the nuances and obstacles of retirement security among AIANs.

Notes

1 According to the Office of Management and Budget and the Census Bureau, American Indian or Alaska Native is defined as “a person having origins in any of the original peoples of North and South America (including Central America) and who maintains tribal affiliation of community attachment. Tabulated data for this category include people who (1) indicate their race as “American Indian” or “Alaska Native”; (2) report the name of an American Indian tribe (for example, Navajo or Blackfeet) or Alaska Native tribe (for example, Inupiat or Yup'ik); or (3) report entries such as Canadian Indian, French American Indian or Spanish American Indian” (Census Bureau 2008, 3).

2 Respondents who indicated more than one tribe are included (Norris, Vines, and Hoeffel 2012).

3 Respondents who indicated more than one tribe are included (ibid.).

4 However, Perry (2012) notes that the majority of the population living in either American Indian reservation areas or Alaska Native village statistical areas (77 percent and 66 percent, respectively) report being non-AIAN.

5 For more comprehensive presentations of retirement income, see the Social Security Administration's Income of the Population 55 or Older and Income of the Aged Chartbook, both updated regularly at http://www.socialsecurity.gov/policy.

References

Census Bureau. 2008. A Compass for Understanding and Using American Community Survey Data: What General Data Users Need to Know. Washington, DC: U.S. Government Printing Office. http://www.census.gov/acs/www/Downloads/handbooks/ACSGeneralHandbook.pdf.

———. 2010a. American Community Survey, Puerto Rico Community Survey: 2009 Subject Definitions. http://www.census.gov/acs/www/Downloads/data_documentation/SubjectDefinitions/2009_ACSSubjectDefinitions.pdf.

———. 2010b. “Chapter 4: Sample Design and Selection.” In ACS Design and Methodology. http://www.census.gov/acs/www/Downloads/survey_methodology/Chapter_4_RevisedDec2010.pdf.

———. 2012. “2010 Census Shows Nearly Half of American Indians and Alaska Natives Report Multiple Races.” News release (January 25). http://www.census.gov/newsroom/releases/archives/2010_census/cb12-cn06.html.

———. 2013. 2005–2009 ACS PUMS Data Dictionary. http://www.census.gov/acs/www/Downloads/data_documentation/pums/DataDict/PUMSDataDict05_09.pdf.

Department of Health and Human Services. 2006. “American Community Survey (ACS).” In Data on Health and Well-being of American Indians, Alaska Natives, and Other Native Americans: Data Catalog. http://aspe.hhs.gov/hsp/06/catalog-ai-an-na/acs.htm.

Gendell, Murray, and Jacob S. Siegel. 1992. “Trends in Retirement Age by Sex, 1950–2005.Monthly Labor Review (July): 22–29. http://www.highbeam.com/doc/1G1-12504816.html.

HHS. See Department of Health and Human Services.

Norris, Tina, Paula L. Vines, and Elizabeth M. Hoeffel. 2012. “The American Indian and Alaska Native Population: 2010.” 2010 Census Briefs. Department of Commerce, Economics and Statistics Administration, Census Bureau. http://www.census.gov/prod/cen2010/briefs/c2010br-10.pdf.

Olsen, Janice A. 2002. “Social Security Benefit Reporting in the Survey of Income and Program Participation and in Social Security Administrative Records.” ORES Working Paper No. 96. Washington, DC: Social Security Administration, Office of Policy, Office of Research, Evaluation, and Statistics. http://www.socialsecurity.gov/policy/docs/workingpapers/wp96.html.

Perry, Steven W. 2012. Tribal Crime Data Collection Activities, 2012. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. http://www.bjs.gov/content/pub/pdf/tcdca12.pdf.

Smith-Kaprosy, Nolan, Patricia P. Martin, and Kevin Whitman. 2012. “An Overview of American Indians and Alaska Natives in the Context of Social Security and Supplemental Security Income.” Social Security Bulletin 72(4): 1–10. http://www.socialsecurity.gov/policy/docs/ssb/v72n4/v72n4p1.html.