Income, poverty and education across South Asian American groups

South Asian here means residents who report a single race and name Indian, Pakistani, Bangladeshi, Nepali, Sri Lankan or Bhutanese as their Asian group, as in chapter 1. Residents who report more than one race are not included. Dollar figures are 2024 dollars, adjusted with the consumer price index [1].The median Indian American household had an income of $163,100 in the American Community Survey (ACS) for 2022 to 2024 (90% interval $161,300 to $164,800) [2]. Indian Americans are 80% of the 6.05 million single-race South Asian residents of the US. The median for Sri Lankan American households was $113,000, Pakistani $111,400, Nepali $104,700, Bhutanese $86,100 and Bangladeshi $81,700. The Sri Lankan median cannot be told apart from the Pakistani or the Nepali at 90% confidence; the Pakistani median is $6,700 above the Nepali (90% interval $800 to $12,600). The Bhutanese median ($66,700 to $105,500) cannot be told apart from the Bangladeshi or the Nepali. All US households had a median of $80,700, and households of other Asian origins $104,200. The Pew Research Center's analysis of the 2021–23 surveys gives the same highest and lowest of the five origins it reports income for, with medians of $151,200 for Indian and $78,400 for Bangladeshi households in 2023 dollars [3]. Chakravorty, Kapur and Singh trace the Indian American figures to the selective immigration of highly educated workers [4].

Poverty is the Census Bureau's own calculation, which compares family income with the official thresholds. For 2024 alone it gives 12.0% nationally, against 12.1% in the Bureau's published tables, which use the full sample [5].In 2022–24, 15.9% of Bangladeshi Americans lived below the poverty line, against 12.3% of all US residents and 6.4% of Indian Americans. The Pakistani American rate, 11.7% (90% interval 10.8 to 12.7), cannot be told apart from the national rate. Among children under 18, the Bangladeshi American rate was 20.0% (90% interval 17.1 to 23.0), against 15.0% for Pakistani, 4.7% for Indian American and 15.8% for all US children. The Bhutanese American rate, 8.9% (3.7 to 14.1), is a small-sample estimate; another 26.4% of Bhutanese Americans had incomes between one and two times the threshold, and their median income per adult-equivalent ($40,900) was about the same as the Bangladeshi ($43,600). 43.1% of Bangladeshi Americans had Medicaid or another means-tested health plan, against 9.9% of Indian Americans and 21.0% of all residents.

The public-use ACS file has records for about 3.4 million people a year. Pooling the 2022, 2023 and 2024 samples gives 172,495 South Asian person records, including 138,835 Indian, 17,282 Pakistani, 7,860 Bangladeshi, 6,098 Nepali, 2,022 Sri Lankan and 398 Bhutanese Americans. Every pooled South Asian and other Asian American estimate carries a 90% interval from the Census Bureau's replicate weights [6]. The Bhutanese estimates rest on 398 people in 109 households and are all marked as small samples.

Education, English ability and women's employment also differ by origin

78.0% of Indian American adults aged 25 and over had a bachelor's degree and 46.3% a graduate or professional degree. The bachelor's share was 63.7% among Sri Lankan, 59.8% among Pakistani, 52.2% among Bangladeshi, 50.3% among Nepali and 25.7% among Bhutanese American adults, against 36.3% of all US adults. 46.9% of Bhutanese and 21.0% of Nepali American adults had not finished high school, compared with 10.2% nationally, so the Nepali group combines many college graduates and many adults with little schooling.

Of the 43,531 US residents born in Bhutan (2022–24 average), 21,722 report Nepali as their Asian group and 10,895 Bhutanese. Most Bhutanese Americans came as refugees from camps in eastern Nepal; more than 92,500 were resettled in the US starting in late 2007 [7]. The Nepali row therefore includes part of that refugee population, and the Bhutanese row covers only part of it. Children born in the camps count as born in Nepal. The born-in-Bhutan row counts immigrants of any race.Among immigrants born in Bhutan, of any race, 54.8% of adults had no high school diploma and 36.6% of those aged 5 and over spoke English less than well. The share speaking English not well or not at all, a stricter measure than the Census Bureau's "less than very well," was 18.3% for Bhutanese, 17.7% for Bangladeshi, 13.9% for Nepali, 7.7% for Pakistani, 5.1% for Indian and 4.6% for Sri Lankan Americans, against 4.4% of all US residents and 15.4% of other Asian Americans.

Employment differs most among women. 60.2% of Pakistani and 61.2% of Bangladeshi American women aged 25 to 54 were employed, against 72.4% of Indian American women, 74.2% of Nepali American women and 76.6% of all US women that age. Among men the rates were 88.6%, 88.2%, 92.8%, 88.4% and 84.9%.

Figure 1

Indian Americans had the highest median household income and bachelor's share of the six South Asian origins, and Bangladeshi Americans the highest poverty rate

Pooled 2022–24 estimates with 90% intervals. Dollar figures in 2024 dollars. † fewer than 200 records, or any Bhutanese estimate (all rest on 109 households).

• South Asian origin
• All South Asian
• Other Asian American
• All US residents
• Reference group, outside the six origins or a subset

† Small sample.

Show data

Measured against their own state's median, Bangladeshi Americans are at 0.97 on household income and 0.73 after adjusting for household size

Household income also counts everyone in the household. Bangladeshi American households averaged 3.66 people and 1.75 employed members; Indian American households 2.92 people and 1.57 employed members; all US households 2.46 and 1.23. These are averages over households of the number of people in each; the Census Bureau's measure in chapter 5, household population divided by the number of households, gives somewhat different values for the same years. Dividing household income by the square root of household size, a standard equivalence scale [8], and assigning the result to every member gives income per adult-equivalent. The median Indian American had $101,000 per adult-equivalent, the median Bangladeshi American $43,600 and the median US resident $56,500.

The groups also live in states with different income levels. 44% of Bangladeshi Americans live in New York State, where the median household income was $85,200. Indian Americans are spread more widely, with the largest shares in California (20%), Texas (13%) and New Jersey (9%). Dividing each household's income by the median for all households in its state, the typical Indian American household had 1.90 times its state's median (90% interval 1.88 to 1.91), Pakistani households 1.31, Sri Lankan 1.31, Nepali 1.24 and Bangladeshi households 0.97 (0.91 to 1.04). Comparing each person's income per adult-equivalent with the median for all residents of the same state gives 1.69 for Indian, 1.28 for Sri Lankan, 1.02 for Pakistani, 0.99 for Nepali, 0.77 for Bhutanese and 0.73 for Bangladeshi Americans (0.69 to 0.76). A state median reflects the state's income level, not living costs, which also vary within states.

Figure 2

At each percentile shown, Indian Americans had more income per adult-equivalent than US residents as a whole, and Bangladeshi Americans less

Household income divided by the square root of household size, assigned to each person, 2022–24, 2024 dollars. Line: 10th to 90th percentile. Bar: 25th to 75th. Dot: median.

Show data

The ratio of the 90th to the 10th percentile of income per adult-equivalent was 8.9 for Indian, 10.7 for Pakistani, 10.4 for Bangladeshi, 7.8 for Nepali and 9.9 for Sri Lankan Americans, against 9.2 for all US residents and 10.5 for other Asian Americans. The combined South Asian ratio, 10.1, is higher than the Indian ratio because the six origins' distributions sit at different levels. Pew Research Center analyses, which also adjust household income for household size, report a 90/10 ratio of 10.7 among Asian American adults in 2016 against 8.7 for all US adults [9], and ratios above 13 for Pakistani and Sri Lankan households and below 8 for Nepalese households in 2022 [10]. Those are single-year estimates with different populations, so their levels differ from the pooled ratios here.

Figure 3

Relative to their own state's median, the typical Indian American household had 1.90 times the income and the typical Bangladeshi American household 0.97 times

Median of household income divided by the median for all households in the same state, and of income per adult-equivalent divided by the median for all residents of the same state, by origin, with intervals

Table of eight groups: the median ratio of household income to the state median and the median ratio of income per adult-equivalent to the state median, each with a 90% interval, and the states with the most residents of each origin. Indian 1.90 and 1.69 (CA 20%, TX 13%, NJ 9%); Pakistani 1.31 and 1.02 (TX 17%, NY 15%, CA 13%); Bangladeshi 0.97 and 0.73 (NY 44%, MI 8%, TX 7%); Nepali 1.24 and 0.99 (TX 15%, OH 11%, CA 11%); Sri Lankan 1.31 and 1.28 (CA 20%, NY 13%, TX 9%); Bhutanese 1.13† and 0.77† (PA 33%, OH 20%, GA 10%); All South Asian 1.76 and 1.52; Other Asian American 1.17 and 1.14 (CA 36%, NY 8%, TX 7%).

Group	Household income ÷ state median, median	90% interval	Income per adult-equivalent ÷ state median, median	90% interval	States with the most residents of this origin (share)
Indian	1.90	1.88 to 1.91	1.69	1.67 to 1.71	CA 20%, TX 13%, NJ 9%
Pakistani	1.31	1.26 to 1.36	1.02	0.98 to 1.06	TX 17%, NY 15%, CA 13%
Bangladeshi	0.97	0.91 to 1.04	0.73	0.69 to 0.76	NY 44%, MI 8%, TX 7%
Nepali	1.24	1.21 to 1.27	0.99	0.95 to 1.02	TX 15%, OH 11%, CA 11%
Sri Lankan	1.31	1.19 to 1.42	1.28	1.19 to 1.37	CA 20%, NY 13%, TX 9%
Bhutanese	1.13†	0.85 to 1.41	0.77†	0.66 to 0.87	PA 33%, OH 20%, GA 10%
All South Asian	1.76	1.75 to 1.78	1.52	1.51 to 1.54
Other Asian American	1.17	1.17 to 1.18	1.14	1.13 to 1.15	CA 36%, NY 8%, TX 7%

Holding measured traits fixed narrows the Pakistani, Nepali and Sri Lankan income gaps with Indian households by about half, and the Bangladeshi gap by 37%, on the log scale

A regression of log household income on origin, for householders aged 25 to 64, shows how much of each gap with Indian American households remains among households that look alike on other measured traits. With only survey-year indicators, Bangladeshi households had 48.3% less income than Indian households (90% interval 45.5 to 51.1). Holding the householder's sex, age and education fixed, the difference was 37.2%. Adding the number of employed people, household size, state, and the householder's years in the US and citizenship left 33.9% (31.1 to 36.7).

The same adjustments reduce the Pakistani difference from 30.9% to 17.2%, the Nepali from 38.8% to 22.8%, the Sri Lankan from 30.2% to 16.1% and the Bhutanese from 44.1% to 11.1% (interval −1.2 to 23.4, 104 households, so the Bhutanese figures are rough). Measured on the log scale, the adjustments remove 49% of the Pakistani gap, 47% of the Nepali, 51% of the Sri Lankan and 37% of the Bangladeshi gap. Sex, age and education come first. Adding the number of employed members and household size then moves the Nepali gap from 20.5% to 27.6%; it moves the other gaps by 0.5 to 5.5 points, within their 90% intervals.

The gap also appears within education levels. Among households whose householder, aged 25 or over, had a bachelor's degree or more, the 2022–24 Bangladeshi American median was $111,800, the Pakistani $141,300 and the Indian American $188,300, against $124,900 for all US households headed by someone with a bachelor's degree or more. The regression does not measure field of study, occupation, English ability or where a degree was earned, so the remaining gaps are differences this chapter can describe but not explain.

Figure 4

With sex, age, education, workers, household size, state, time in the US and citizenship held fixed, Bangladeshi households still had 34% less income than Indian households

Percent difference in household income from Indian American households, householders aged 25 to 64, 2022–24, with 90% intervals.

• Unadjusted
• Sex, age and education held fixed
• All listed traits held fixed

Show data

In 2022–24, Indian immigrants who arrived after 2019 had a poverty rate of 16.2%

Within the Indian American population, poverty is highest among the newest arrivals. In 2022–24, 16.2% of Indian immigrants of all ages who arrived in 2020 or later lived below the poverty line (90% interval 15.1 to 17.3), and 22.9% of the adults among them were enrolled in school. Earlier arrival cohorts had poverty rates of 4.2% to 4.8%. 48.9% of the women aged 25 to 54 who arrived in 2020 or later were employed, against 78.6% of those who arrived in 2000 to 2009. The ACS does not record visa type, so it cannot separate students, workers and their spouses.

The median age of Indian Americans who were born US citizens was 15 years. Those aged 25 and over, 11,894 records, had about the same bachelor's share as Indian immigrants born outside the Caribbean and Fiji (78.8% against 79.2%), and 82.0% of the women aged 25 to 54 among them were employed.

Indo-Caribbean Americans who report a race other than Asian, or more than one race, are outside this chapter's groups. Of residents born in Guyana, Trinidad and Tobago or Fiji in 2022–24, 14.0% reported a single South Asian race and 1.9% Asian Indian with another race (chapter 1).Asian Indian immigrants born in Guyana, Trinidad and Tobago, Suriname, Jamaica or Fiji, about 82,000 residents, differ from other Indian immigrants. 30.5% of their adults had a bachelor's degree, against 79.2% of other Indian immigrants, and their median household income was $92,400, against $169,100. They are older (median age 57.3 years against 41.8) and more likely to own their home (76.1% against 63.7%).

Figure 5

Among Indian Americans, recent immigrants had the highest poverty rate and the lowest employment among women

Indian Americans by birthplace, year of arrival and citizenship, pooled 2022–24, with 90% intervals.

Show data

In 2022–24, poverty among immigrants aged 25 to 64 was highest in their first five years in the US

In 2022–24, among immigrants aged 25 to 64, 25.9% of Bangladeshi immigrants who had been in the US less than five years were below the poverty line (595 records), against 10.5% of those in the US 20 years or more. The rate was also lower for the longest-resident Pakistani (16.6% to 8.0%), Nepali (16.7% to 4.5%), Indian (10.0% to 3.9%) and other Asian immigrants (17.6% to 6.9%). Among Bangladeshi immigrants the rate did not fall steadily; it was 10.0% for those in the US 5 to 9 years and 15.8% for those in the US 10 to 19 years. For Pakistani immigrants, the interval for the first five years overlaps that of a later group. These are different arrival cohorts observed in the same years, not the same people followed over time, so the pattern mixes change with time in the US and differences between cohorts.

Figure 6

Bangladeshi immigrants who arrived in the last five years had a poverty rate of 25.9%, against 10.5% for those in the US 20 years or more

Share of immigrants aged 25 to 64 below the poverty line, by years since arrival, pooled 2022–24, with 90% intervals.

• 0 to 4 years
• 5 to 9 years
• 10 to 19 years
• 20 years or more

Show data

Median household income rose for Indian, Pakistani, Bangladeshi and Nepali Americans

Pooling the 2005, 2006 and 2007 samples and comparing them with 2022–24, in 2024 dollars, the median Indian American household income rose from $122,600 to $163,100 (33%, 90% interval 31 to 35%), the Pakistani from $90,700 to $111,400 (23%, 17 to 29%) and the Bangladeshi from $66,400 to $81,700 (23%, 7 to 39%). The median for all US households went from $75,500 to $80,700. Nepali households, first coded in 2012, went from $60,600 in 2012–14 to $104,700 in 2022–24. Over those years the Nepali American population grew from about 105,000 to 241,000 and the share born in Bhutan fell from 16% to 9%, so the change also reflects who is in the group. Each period's figures describe that period's residents, not the same households followed over time. The Bangladeshi median was 54% of the Indian median in 2005–07 and 50% in 2022–24; the change, −4.0 points (90% interval −10.9 to 2.9), is within sampling error. Sri Lankan and Bhutanese samples are too small for single-year series.

Poverty fell over the same years. The Bangladeshi American rate was 23.4% in 2005–07 and 15.9% in 2022–24, a fall of 7.6 points (90% interval 2.8 to 12.3). The Pakistani rate went from 14.7% to 11.7%, the Indian from 8.2% to 6.4%, and the Nepali from 26.1% in 2012–14 to 9.8%. The national rate went from 13.2% to 12.3%.The 2020 ACS 1-year sample was released only as experimental estimates because the pandemic disrupted data collection [11]. Its points are drawn open and carry no intervals, because its replicate weights vary far less than in other years and understate sampling error. The 2024 weights use population controls with a revised estimate of net international migration, which added about 5 million people to the 2024 population total [12]. The pooled resident counts in this chapter and the changes from 2023 to 2024 in the chart partly reflect that reweighting.

Figure 7

Indian American median household income rose from $117,700 in 2005 to $166,000 in 2024, in 2024 dollars; the median for all US households rose from $74,200 to $81,600

Median household income by householder's origin, single years, 2024 dollars. Shaded bands are 90% intervals for the four South Asian series; the other Asian American and all-US series, with large samples, have none, and neither does 2020 (open markers), the experimental release.

• Indian
• Pakistani
• Bangladeshi
• Nepali
• Other Asian American
• All US households

Show data

What these data cannot show

ACS income is pre-tax money income reported by respondents. It leaves out taxes, noncash benefits and wealth, and the poverty thresholds are the same everywhere in the country, so poverty rates in New York and Ohio are measured against the same thresholds [13]. The comparison with state medians is the only geographic adjustment here, and it reflects state income levels rather than living costs.

The groups rest on how people describe their race. Residents who report more than one race, including many children of mixed marriages, are not counted, and Indo-Caribbean and Afghan Americans are counted only in part. Afghan has a code of its own only in the 2024 file; in 2022 and 2023 Afghan respondents coded as Asian carry a general "other Asian" code, so the other Asian American comparison group includes them in every year. The Sri Lankan and Bhutanese estimates rest on 2,022 and 398 person records. The ACS does not ask about visa type, caste, religion or where a degree was earned, all of which a fuller account of these gaps would need.

Appendix tables

Figure 8

Each sample's income amounts are scaled by its adjustment factor and then by the consumer price index to 2024 dollars

IPUMS adjustment factor ADJUST and the CPI-U annual average by ACS sample year, 2005 to 2024

Table of two rows, the ADJUST factor and the CPI-U annual average, with one column for each ACS sample year from 2005 to 2024. 2005: ADJUST 1.0192, CPI-U 195.3; 2006: ADJUST 1.0157, CPI-U 201.6; 2007: ADJUST 1.0168, CPI-U 207.342; 2008: ADJUST 1.0184, CPI-U 215.303; 2009: ADJUST 0.9995, CPI-U 214.537; 2010: ADJUST 1.0076, CPI-U 218.056; 2011: ADJUST 1.0182, CPI-U 224.939; 2012: ADJUST 1.0102, CPI-U 229.594; 2013: ADJUST 1.0075, CPI-U 232.957; 2014: ADJUST 1.0084, CPI-U 236.736; 2015: ADJUST 1.0013, CPI-U 237.017; 2016: ADJUST 1.0076, CPI-U 240.007; 2017: ADJUST 1.0112, CPI-U 245.12; 2018: ADJUST 1.0131, CPI-U 251.107; 2019: ADJUST 1.0101, CPI-U 255.657; 2020: ADJUST 1.0061, CPI-U 258.811; 2021: ADJUST 1.0299, CPI-U 270.97; 2022: ADJUST 1.0423, CPI-U 292.655; 2023: ADJUST 1.0195, CPI-U 304.702; 2024: ADJUST 1.0152, CPI-U 313.689.

Sample year	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
ADJUST	1.0192	1.0157	1.0168	1.0184	0.9995	1.0076	1.0182	1.0102	1.0075	1.0084	1.0013	1.0076	1.0112	1.0131	1.0101	1.0061	1.0299	1.0423	1.0195	1.0152
CPI-U	195.3	201.6	207.342	215.303	214.537	218.056	224.939	229.594	232.957	236.736	237.017	240.007	245.12	251.107	255.657	258.811	270.97	292.655	304.702	313.689

Figure 9

Estimates from this file agree with the Census Bureau's published 2024 profiles except the Sri Lankan median income

Median household income and the poverty, bachelor's degree, uninsured and owner-occupied shares in 2024: this file against table S0201

Table of eight groups in 2024 comparing this file's median household income, poverty rate, bachelor's share, uninsured share and owner-occupied share with the Census Bureau's published table S0201. Asian Indian alone: median household income $166,700 in this file against $165,721 in the Census table, poverty 6.6% in this file against 6.7% in the Census table; Pakistani alone: median household income $109,700 in this file against $111,719 in the Census table, poverty 11.3% in this file against 11.0% in the Census table; Bangladeshi alone: median household income $76,200 in this file against $76,365 in the Census table, poverty 18.1% in this file against 18.2% in the Census table; Nepalese alone: median household income $103,500 in this file against $102,046 in the Census table, poverty 11.5% in this file against 11.4% in the Census table; Sri Lankan alone: median household income $129,200 in this file against $111,484 in the Census table, poverty 9.0% in this file against 8.5% in the Census table; Bhutanese alone: median household income too few records in this file, not published in the Census table, poverty 12.2%† in this file, not published in the Census table; Asian alone: median household income $117,400 in this file against $117,289 in the Census table, poverty 9.7% in this file against 9.7% in the Census table; Total population: median household income $81,600 in this file against $81,604 in the Census table, poverty 12.0% in this file against 12.1% in the Census table.

2024, single year	Median household income, this file	Census table	Poverty, %, this file	Census table	Bachelor's, %, this file	Census table	Uninsured, %, this file	Census table	Owner-occupied, %, this file	Census table
Asian Indian alone	$166,700	$165,721	6.6	6.7	79.0	78.8	3.8	3.9	63.3	63.0
Pakistani alone	$109,700	$111,719	11.3	11.0	61.3	61.8	7.7	7.7	57.5	57.3
Bangladeshi alone	$76,200	$76,365	18.1	18.2	54.0	53.9	6.7	6.8	45.6	46.9
Nepalese alone	$103,500	$102,046	11.5	11.4	47.0	48.4	9.8	10.1	59.4	57.6
Sri Lankan alone	$129,200	$111,484	9.0	8.5	64.2	63.4	5.5	6.2	61.0	58.9
Bhutanese alone	too few	not published	12.2†	not published	29.9†	not published	13.9†	not published	too few	not published
Asian alone	$117,400	$117,289	9.7	9.7	59.1	59.1	5.2	5.2	63.3	63.2
Total population	$81,600	$81,604	12.0	12.1	36.9	36.8	8.2	8.2	65.3	65.3

---

Data sources

• American Community Survey 1-year samples (IPUMS USA)1-year samples 2005 to 2024, with the 80 replicate weights and a supplementary extract (USA extracts 13 to 16, November 2026) for the Census poverty measure [2]Years: 2005 to 2024 · Weights: household weight HHWT for household income, person weight PERWT for person measures, with the 80 replicate weights
• Consumer Price Index for All Urban ConsumersConsumer Price Index for All Urban Consumers, annual averages [1]Years: 2005 to 2024 · Weights: not applicable (price index)
• US Census Bureau published ACS tablesSelected Population Profiles, 2024, for the check against published tables [5]Years: 2020 to 2024 · Weights: published weighted estimates with 90% margins of error

Methods

Groups. IPUMS detailed race (RACED): Indian 610 (Asian Indian), 678 (Asian Indian and another Asian group) and 657 (Sikh, a write-in coded only in 2023); Pakistani 669; Bangladeshi 664; Nepali 643; Sri Lankan 670; Bhutanese 641. Nepali and Bhutanese codes start in 2012. Other Asian American is every other single-race Asian code, including combinations of Asian groups and Afghan respondents, who fall under the general other-Asian code 671 in 2022–23 and have their own code, 655, in 2024. The Afghan reference row uses 655 in 2024 only. Households take the origin of the householder; person measures use each person's own origin.

Measures. Household income is HHINCOME for households (not group quarters), weighted by the household weight on the householder's record. Income per adult-equivalent divides it by the square root of the number of people in the household. Poverty uses the Census Bureau's own calculation, CBPOVERTY (family income as a percentage of the official threshold), from a supplementary IPUMS extract that joins person by person to the main file. From 2023 the file codes zero or negative family income as 0, the same code as "not determined"; those people are counted as poor when they live in households and are not unrelated children under 15, which brings the 2024 rates within 0.5 points of the published tables (see the appendix tables). IPUMS's constructed POVERTY variable, which forms families from IPUMS relationship codes, is not used. Education (EDUCD) is for people 25 and over; no diploma means 12th grade or less without a diploma. Health insurance (HCOVANY, HINSCAID) is for the civilian population outside institutions. Limited English means speaking English "not well" or "not at all" (SPEAKENG), age 5 and over. Employment is EMPSTAT for ages 25 to 54. Adults in school are people 18 and over enrolled in school (SCHOOL). Years in the US is the survey year minus YRIMMIG.

Dollars. ACS respondents report income for the 12 months before their interview, so the amounts in a 1-year file span about two calendar years. Following Census practice, each amount is multiplied by the sample's adjustment factor (ADJUST; 2022: 1.042311, 2023: 1.019518, 2024: 1.01525) to put it in dollars of the survey year [14], then moved to 2024 dollars with the CPI-U annual average (2022: 292.655, 2023: 304.702, 2024: 313.689) [1]. Medians are computed after this conversion. The factors for every year from 2005 are in the appendix tables.

Uncertainty. Standard errors use the 80 successive-difference replicate weights, with variance 4/80 times the sum of squared deviations of the replicate estimates [6]. The pooled 2022–24 estimates pair replicate r across years. Intervals for changes between periods treat the periods as independent samples and add their variances. Medians and percentiles are interpolated within $2,500 bins (household income), $1,000 bins (income per adult-equivalent) and single years of age, the Census Bureau's practice, because reported incomes cluster at round numbers. Intervals are 90% (1.645 standard errors). Estimates on fewer than 50 records are suppressed; those on fewer than 200 records, and every Bhutanese estimate (all rest on 109 households), are marked †. Intervals for all US residents are not computed.

Regression. Weighted least squares of log household income for householders aged 25 to 64 in the six origins (Indian is the reference), with year indicators, then sex, five-year age groups and five education levels, then employed household members (0, 1, 2, 3 or more) and household size (1 to 6 or more), then state, then years in the US (citizen at birth, meaning born in the US or abroad to US-citizen parents; under 10; 10 to 19; 20 or more) and citizenship. Each model is re-estimated under every replicate weight for its interval. Households with zero or negative income (396) are excluded.

Check against published tables. For 2024 alone, the same code is compared with the Census Bureau's Selected Population Profiles, which are published only for groups of 65,000 people or more and so not for Bhutanese Americans [5]. The published tables use the full ACS sample and the public microdata a subsample, so small differences are expected. Medians agree within 2% for every group except Sri Lankan Americans (247 households in the 2024 microdata; published margin of error ±$12,958). Bachelor's, uninsured and homeownership shares agree within 2.1 points. Poverty rates agree within 0.5 points. Census tables for "Asian Indian alone" exclude code 678.

Source: ACS 2024 1-year sample via IPUMS USA, weighted; Census Bureau table S0201, 2024 1-year [5].

Reproduction. south-asian-voters/chapters/04-gaps-by-origin/build_data.py reads the IPUMS ACS parquet files, computes every number and writes data.json and claims.json; build_page.py and charts.js assemble this page.

References

1. US Bureau of Labor Statistics (2026). Consumer Price Index for All Urban Consumers (CPI-U), US city average, all items, series CUUR0000SA0, annual averages. Bureau of Labor Statistics. data.bls.gov/timeseries/CUUR0000SA0
2. Ruggles, S.; Flood, S.; Sobek, M.; Backman, D.; Cooper, G.; Rivera Drew, J. A.; Richards, S.; Rodgers, R.; Schroeder, J.; Williams, K. C. W. (2025). IPUMS USA: Version 16.0 [dataset]. Minneapolis, MN: IPUMS. doi:10.18128/D010.V16.0
3. Im, C. (2025). Facts about Indians in the U.S.. Pew Research Center fact sheet, May 1, 2025, with companion fact sheets on Bangladeshis, Bhutanese, Nepalese, Pakistanis and Sri Lankans. www.pewresearch.org/race-and-ethnicity/fact-sheet/asian-americans-indians-in-the-u-s/
4. Chakravorty, S.; Kapur, D.; Singh, N. (2017). The Other One Percent: Indians in America. New York: Oxford University Press.
5. US Census Bureau (2026). American Community Survey 2024 1-Year Estimates, Selected Population Profiles, table S0201, United States. data.census.gov, population groups 001, 012, 3782, 3783, 3786, 3787 and 3790 (read November 6, 2026). data.census.gov/table/ACSSPP1Y2024.S0201
6. US Census Bureau (2010). American Community Survey Design and Methodology, chapter 12: Variance Estimation. US Census Bureau. www.census.gov/content/dam/Census/library/publications/2010/acs/acs_design_methodology_ch12.pdf
7. US Department of State; US Department of Homeland Security; US Department of Health and Human Services (2017). Proposed Refugee Admissions for Fiscal Year 2018: Report to the Congress. US Department of State. www.rpc.state.gov/documents/FY+2018+Report+to+Congress.pdf
8. OECD (2026). Income Distribution Database: Terms of Reference. OECD Project on the Distribution of Household Incomes (equivalence elasticity of 0.5). www.oecd.org/content/dam/oecd/en/data/datasets/income-and-wealth-distribution-databases/idd-tor-2012-onwards.pdf
9. Kochhar, R.; Cilluffo, A. (2018). Income Inequality in the U.S. Is Rising Most Rapidly Among Asians. Pew Research Center, July 12, 2018. www.pewresearch.org/social-trends/2018/07/12/income-inequality-in-the-u-s-is-rising-most-rapidly-among-asians/
10. Budiman, A. (2024). Income inequality is greater among Chinese Americans than any other Asian origin group in the U.S.. Pew Research Center, May 31, 2024. www.pewresearch.org/short-reads/2024/05/31/income-inequality-is-greater-among-chinese-americans-than-any-other-asian-origin-group-in-the-us/
11. US Census Bureau (2021). Census Bureau Announces Changes for 2020 American Community Survey 1-Year Estimates. Press release CB21-123, July 29, 2021. www.census.gov/newsroom/press-releases/2021/changes-2020-acs-1-year.html
12. US Census Bureau (2025). Population Controls and Other Methodology for the 2024 ACS 1-Year Estimates. ACS user note. www.census.gov/programs-surveys/acs/technical-documentation/user-notes/2025-03.html
13. US Census Bureau (2025). American Community Survey and Puerto Rico Community Survey 2024 Subject Definitions. US Census Bureau (section "Poverty Status in the Past 12 Months"). www2.census.gov/programs-surveys/acs/tech_docs/subject_definitions/2024_ACSSubjectDefinitions.pdf
14. IPUMS USA (2026). Note on the Standardization of ACS/PRCS Income Variables and Other Dollar Amount Variables. IPUMS USA. usa.ipums.org/usa/acsincadj.shtml

How to cite

@misc{singh2026incomepovertyeducation,
  author = {Singh, Vishal},
  title = {Income, poverty and education across South Asian American groups},
  year = {2026},
  url = {https://vishalsingh.org/south-asians/income-poverty-education},
  note = {South Asians in America, chapter 3. Published 7 November 2026.}
}

Acknowledgments

ACS microdata are from IPUMS USA, University of Minnesota. IPUMS is not responsible for the analysis here.