South Asian turnout, 1998 to 2024

South Asian here means a citizen born in India, Pakistan, Bangladesh, Sri Lanka, Nepal or Bhutan, or a citizen with a parent born there. The CPS does not ask about detailed Asian origin, so birthplace is the only way to identify the group. Grandchildren of immigrants, and Indo-Caribbean voters born in Guyana or Trinidad, are not counted. Chapter 1 and chapter 7 instead use single-race responses in the ACS; chapter 7 finds that 91% of adult citizens born in South Asia reported a South Asian race in 2022 to 2024.In the November 2024 Current Population Survey, 67.4% of South Asian American citizens aged 18 and older said they voted (90% interval ±4.0 points), compared with 65.3% of all citizens [1]. In 2000 the figures were 42.8% and 59.5%, though the 2000 South Asian estimate covers only Indian, Pakistani and Bangladeshi Americans (see Methods). In midterm years the group still votes less than other citizens: 46.2% against 52.2% in 2022.

Turnout follows the Census Bureau's convention: respondents who did not answer the voting question count as nonvoters [2]. With that rule this file reproduces the Bureau's published national rates, such as 65.3% in 2024 [3].The Current Population Survey (CPS) is the monthly labor-force survey run by the Census Bureau. Every November after a federal election it asks whether each adult in the household registered and voted. It is the largest regular source of turnout data, with 767 South Asian citizen records in 2024 and between 649 and 792 in each election since 2014. Chapter 7 described who these citizens are; this one asks how often they vote.

Presidential turnout rose to the national rate

South Asian turnout in presidential years rose from 42.8% in 2000 (Indian, Pakistani and Bangladeshi Americans only) to 59.2% in 2012 and 72.9% in 2020. National turnout rose less over the same years, from 59.5% to 66.8%. In 2020 South Asian citizens voted at a higher rate than citizens overall; in 2024 the two rates, 67.4% and 65.3%, were within sampling error of each other. South Asian turnout has been above that of other Asian American citizens in every presidential year since 2004; in 2024 the other Asian American rate was 54.2%.

Figure 1

South Asian presidential-year turnout went from 17 points below the national rate in 2000 to at or above it in 2020 and 2024

Reported turnout among citizens 18 and older, 1998 to 2024. The shaded band is the 90% interval for South Asian citizens. South Asian points for 1998 to 2006 cover Indian, Pakistani and Bangladeshi Americans only. Hover for values.

• South Asian
• Other Asian American
• All citizens

Show data

Midterm turnout still trails

In the three most recent midterms (2014, 2018 and 2022 combined), 41.7% of South Asian citizens voted, against 49.3% of all citizens. Registration accounts for part of the difference: 62.7% of South Asian citizens reported being registered in those years, against 66.9% of all citizens. The rest comes from registered citizens who stayed home. Among registered South Asian citizens, 66.6% voted in those midterms, against 73.7% of all registered citizens.

Presidential years look different. In 2016, 2020 and 2024 combined, South Asian registration (73.1%) matched the national rate (72.2%), and 92.7% of registered South Asian citizens voted, against 89.3% of all registered citizens.

Figure 2

Registered South Asian citizens vote at high rates in presidential years and at lower rates than other registered citizens in midterms

Registration among citizens 18 and older, and voting among registered citizens, pooled over 2016, 2020 and 2024 (presidential) and 2014, 2018 and 2022 (midterm), with 90% intervals.

• South Asian
• Other Asian American
• Other naturalized citizens
• All citizens

Show data

Turnout below what education and age predict

Expected turnout applies the turnout rate of all citizens in the same year, state, age group, education level and sex to South Asian citizens, then averages over the South Asian population. It describes how South Asian turnout compares with similar citizens; it does not explain the difference.In the 2024 ACS, South Asian adult citizens had a median age of 45 against 48 for all adult citizens, and 70% of those aged 25 and older held a bachelor's degree or more, against 38% of all adult citizens that age (chapter 7). College graduates vote at higher rates than other citizens. Holding those traits fixed changes the comparison. If South Asian citizens had voted at the rates of all citizens with the same age, education, sex and state, their turnout in 2016, 2020 and 2024 combined would have been 70.7%; it was 67.8%, a shortfall of 2.9 points (90% interval ±2.2). In the 2014, 2018 and 2022 midterms the shortfall was 11.9 points (41.7% against an expected 53.6%).

The presidential shortfall has narrowed. It was between 16 and 24 points in 2000, 2004 and 2008 (the 2000 and 2004 estimates cover Indian, Pakistani and Bangladeshi Americans only). Among college graduates alone, South Asian citizens voted at 73.9% in the three recent presidential elections, against 78.5% of all citizens with a bachelor's degree.

Figure 3

South Asian turnout fell 24 points short of similar citizens in 2000 and 4 points short in 2024

South Asian turnout minus the turnout expected from all citizens of the same age, education, sex and state, percentage points, with 90% intervals. Filled points are presidential years; open points are midterms. South Asian points for 1998 to 2006 cover Indian, Pakistani and Bangladeshi Americans only.

• Presidential year
• Midterm year

Show data

Immigrants and their children

Most South Asian citizens are naturalized immigrants: 1,640 of the 2,296 South Asian records in the three recent presidential years. Their turnout rises with time in the country, from 60.5% among those who arrived less than 15 years before the election to 71.7% among those who arrived 25 or more years before. At each length of residence they vote at higher rates than other naturalized citizens, who range from 46.9% to 60.6%.

The CPS records arrival year in ranges of two to ten years, so years in the US are approximate.The second generation, born outside South Asia (almost always in the US) to a South Asian-born parent, votes at about the same rate as immigrants overall (66.0% against 68.5% in the recent presidential years), though it is much younger. Among citizens aged 18 to 29, 60.2% of second-generation South Asians voted, against 50.3% of all young citizens. The second generation is still too small above age 45 to estimate separately.

Figure 4

South Asian naturalized citizens vote more than other naturalized citizens at every length of residence

Reported turnout of naturalized citizens by years between arrival and the election, presidential years 2016, 2020 and 2024 pooled, with 90% intervals.

• South Asian naturalized citizens
• Other naturalized citizens

Show data

Figure 5

Young second-generation South Asians vote at higher rates than young citizens overall

Reported turnout by age, presidential years 2016, 2020 and 2024 pooled, with 90% intervals. Groups with fewer than 100 records are not shown.

• South Asian immigrants
• South Asian second generation
• All citizens

Show data

Differences by origin and schooling

Indian Americans are most of the sample, and their presidential-year turnout, 70.0%, sets the group average. Pakistani Americans voted at 61.7% and Bangladeshi Americans at 68.5%, with intervals of about ±6.4 and ±8.5 points. Origin here is the respondent's own birthplace or, for the second generation, a parent's.

Figure 6

Indian American citizens voted at 70.0% in recent presidential elections; the smaller origin groups have wide intervals

Turnout and registration among citizens 18 and older, presidential years 2016, 2020 and 2024 pooled.

Table of turnout and registration among citizens 18 and older, presidential years 2016, 2020 and 2024 pooled. South Asian, all: voted 67.8 percent (90% interval plus or minus 2.4 points), registered 73.1 percent, 92.7 percent of the registered voted, 2,296 records; India: voted 70.0 percent (90% interval plus or minus 2.6 points), registered 75.2 percent, 93.1 percent of the registered voted, 1,697 records; Pakistan: voted 61.7 percent (90% interval plus or minus 6.4 points), registered 69.1 percent, 89.3 percent of the registered voted, 303 records; Bangladesh: voted 68.5 percent (90% interval plus or minus 8.5 points), registered 72.6 percent, 94.4 percent of the registered voted, 158 records; Sri Lanka, Nepal or Bhutan: voted 52.3 percent (90% interval plus or minus 11.5 points), registered 55.6 percent, 94.0 percent of the registered voted, 138 records; South Asian, bachelor's degree or more: voted 73.9 percent (90% interval plus or minus 2.5 points), registered 78.1 percent, 94.5 percent of the registered voted, 1,558 records; South Asian, less than a bachelor's: voted 55.3 percent (90% interval plus or minus 4.0 points), registered 62.8 percent, 88.0 percent of the registered voted, 738 records; All citizens, bachelor's degree or more: voted 78.5 percent (90% interval plus or minus 0.3 points), registered 83.1 percent, 94.5 percent of the registered voted, 84,695 records; All citizens, less than a bachelor's: voted 57.2 percent (90% interval plus or minus 0.3 points), registered 66.5 percent, 85.9 percent of the registered voted, 163,587 records; All citizens: voted 64.5 percent (90% interval plus or minus 0.2 points), registered 72.2 percent, 89.3 percent of the registered voted, 248,282 records.

Group (citizens 18+)	Voted, %	90% interval	Registered, %	Voted, % of registered	Records
South Asian, all	67.8	±2.4	73.1	92.7	2,296
India	70.0	±2.6	75.2	93.1	1,697
Pakistan	61.7	±6.4	69.1	89.3	303
Bangladesh	68.5	±8.5	72.6	94.4	158
Sri Lanka, Nepal or Bhutan	52.3	±11.5	55.6	94.0	138
South Asian, bachelor's degree or more	73.9	±2.5	78.1	94.5	1,558
South Asian, less than a bachelor's	55.3	±4.0	62.8	88.0	738
All citizens, bachelor's degree or more	78.5	±0.3	83.1	94.5	84,695
All citizens, less than a bachelor's	57.2	±0.3	66.5	85.9	163,587
All citizens	64.5	±0.2	72.2	89.3	248,282

What these CPS data cannot show

The CPS records what people say about voting, and surveys overstate turnout. Studies that match survey respondents to voter files find that nonvoters who are politically engaged and have more resources are the most likely to say they voted [4], and in 2024 South Asian adult citizens had a median household income of $172 thousand, against $99 thousand for all adult citizens, with 70% of those 25 and older holding a bachelor's degree or more (chapter 7). The South Asian rates here may therefore be overstated by more than the national rate. Supplement nonresponse, which the Census convention counts as not voting, was 13 to 20% for South Asian citizens in 2016 to 2024, against 14 to 18% for all citizens. In 2006 to 2010 the difference was larger (22 to 24% against 12 to 16%), so the convention widens the South Asian gap in those years, and the expected-turnout comparison with it. Counting only respondents who answered the question, South Asian turnout in 2008 was 68.7% against 73.8% for all citizens, a gap of 5 points instead of 11. In 2000 the gap among respondents was 17 points, close to the 17 points under the convention.

The intervals are approximate. The public CPS file has no replicate weights and no sampling-unit identifiers, so they come from resampling households, which ignores the clustering of households in sampled areas and makes the intervals somewhat too narrow. The CPS also says nothing about how anyone voted. Vote choice is the subject of the next chapter, chapter 9.

---

Data sources

• Current Population Survey, Voting and Registration Supplement (IPUMS CPS)November Voting and Registration Supplement, 1998 to 2024, from IPUMS CPS (extract 33, with own and parents' birthplace) [1]Years: 1998 to 2024 · Weights: VOSUPPWT (supplement weight)

Methods

Sample. Citizens aged 18 and older with a positive supplement weight (VOSUPPWT), in every November supplement from 1998 to 2024. Turnout follows the Census convention: voted means the respondent reported voting, and refusals, don't-knows and nonresponse count as not voting. From 2022 IPUMS codes supplement nonrespondents as not in universe; they are kept in the denominator, which reproduces the Bureau's published turnout in every year from 2008 to 2024. Registered means the respondent reported registering or voting.

Group definition. South Asian immigrants were born in India, Pakistan, Bangladesh, Sri Lanka, Nepal or Bhutan (IPUMS BPL). The second generation was born elsewhere, almost always in the US, and has a mother or father born in one of those countries (MBPL, FBPL). Parents' birthplace is miscoded in the 1994 and 1996 IPUMS samples, so the series starts in 1998. The 1998 to 2006 samples contain no records with Sri Lanka or Nepal as own, mother's or father's birthplace; both first appear in 2008, and Bhutan first appears in 2012. South Asian estimates for 1998 to 2006, including the 2000 baseline, therefore cover India, Pakistan and Bangladesh only. Origin is the respondent's birthplace, or the father's and then the mother's for the second generation.

Expected turnout. For each year, state, age group, education level and sex, the turnout rate of all citizens is applied to South Asian citizens in that cell, and the results are averaged with South Asian weights.

Uncertainty. Intervals are 90% (1.645 standard errors) from a household-clustered Poisson bootstrap with 400 replicates, which ignores clustering above the household and understates the true error. Estimates for groups with fewer than 100 records are not shown.

Reproduction. south-asian-voters/chapters/02-turnout/build_data.py computes every number from the IPUMS extract and writes data.json and claims.json; build_page.py assembles this page.

References

1. Flood, S.; King, M.; Rodgers, R.; Ruggles, S.; Warren, J. R.; Backman, D.; Breton, E.; Cooper, G.; Rivera Drew, J. A.; Richards, S.; Van Riper, D.; Williams, K. C. W. (2025). IPUMS CPS: Version 13.0 [dataset]. Minneapolis, MN: IPUMS. doi:10.18128/D030.V13.0
2. Hur, A.; Achen, C. H. (2013). Coding voter turnout responses in the Current Population Survey. Public Opinion Quarterly, 77(4), 985–993.
3. US Census Bureau (2025). Voting and Registration in the Election of November 2024 (P20-590 detailed tables). census.gov. www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-590.html
4. Ansolabehere, S.; Hersh, E. (2012). Validation: What big data reveal about survey misreporting and the real electorate. Political Analysis, 20(4), 437–459. doi:10.1093/pan/mps023

How to cite

@misc{singh2026turnout,
  author = {Singh, Vishal},
  title = {South Asian turnout, 1998 to 2024},
  year = {2026},
  url = {https://vishalsingh.org/south-asians/turnout},
  note = {South Asians in America, chapter 8. Published 7 November 2026.}
}

Acknowledgments

Microdata are from IPUMS CPS, University of Minnesota.