Accepting a domestic-worker contract in the Gulf under the kafala sponsorship system vs. staying home
If you act
Accepting the kafala contract and migrating to the Gulf
47%
If you don't
Staying home and not taking the kafala contract
38%
Percentage who later regret each choice. Bars and full ledger render below.
Career
Last reviewed 2026-05-13
Evidence quality 3.75/5
Eight-dimension review score against the
quality rubric
. Each dimension scored 1–5.
D1 Source verification
4/5
D2 Source authority & independence
5/5
D3 Regret-rate accuracy
2/5
D4 Source comparability
2/5
D5 Gilovich pattern
4/5
D6 Prose quality
5/5
D7 Caveat completeness
5/5
D8 Sample quality
3/5
Average3.75/5
Proxy data — no direct regret survey exists for this decision. Rates are derived from satisfaction scores and access-barrier data rather than questions that directly asked about regret. See caveats below.
Action regret
Accepting the kafala contract and migrating to the Gulf
47%
Not a measured regret rate: no survey asks kafala workers if they regret migrating. The ~47% is a harm proxy — HRW documents passport confiscation, excessive hours, wage theft and abuse as routine (and makes no population-prevalence claim), not a share who report regret
Migrant domestic workers under kafala system in Gulf states (Ethiopia, Kenya, Philippines, Indonesia, South Asia)
documentary and survey evidence, 2018-2020
Inaction regret
Staying home and not taking the kafala contract
38%
Not a measured regret rate: no survey asks non-migrants whether they regret declining a Gulf contract. The ~38% is a structural foregone-income proxy from the Gulf-vs-home wage gap, not observed regret
Adults in high-kafala-sending communities (Ethiopia, Philippines, Indonesia, Nepal, Bangladesh) who did not take a Gulf contract
structural proxy from remittance and income-differential data, 2023-2024
% who regret this choice
Accepting the kafala contract and migrating to the GulfStaying home and not taking the kafala contract
47%38%
action dominates — Action dominates — most regret acting.
Related decisions
Semantically similar decisions — same territory, different trade-offs.
The kafala (sponsorship) system ties migrant domestic workers’ legal status directly to their employer in Gulf Cooperation Council countries, removing the ability to change jobs or return home without sponsor permission. Human Rights Watch’s Oman investigation, based on interviews with 59 migrant domestic workers, found that almost all had their passports confiscated, most were made to work excessively long hours, many were not paid their full wages, and roughly one-quarter said their employers physically or sexually abused them — while explicitly cautioning that it “makes no statistical claims… regarding the prevalence of abuse against the total population of domestic workers in Oman.” The 47% action-side figure is therefore not a documented prevalence rate but a harm proxy: a stand-in for the share of migrants likely to experience rights violations severe enough to plausibly generate regret. The ILO’s “Domestic Workers Across the World” report sizes and legally frames the channel — it counts the domestic-work labour market and maps how far these workers are included in or excluded from working-time, minimum-wage, and maternity protections — rather than documenting Gulf recruitment abuses directly; the abuse documentation on the action side rests on HRW, not the ILO statistical report. Because the workers’ legal status is controlled by their employer and candid survey participation is structurally impossible, no direct “do you regret this decision” survey of kafala workers exists in the published literature; the harm-proxy approach is the most defensible available method.
The inaction side reflects a different — and weaker — evidence base. There is no direct survey of non-migrants who considered but declined a Gulf contract and were asked whether they regret it. The 38% figure on this side is therefore not a measured regret rate but a structural proxy built from the income gap: the World Bank’s Migration and Development Brief records that remittances reached an estimated $656 billion to low- and middle-income countries in 2023 and remain a crucial source of external finance, and Gulf domestic-worker wages run well above the median income available at home in major sending countries. The ILO’s “Domestic Workers Across the World” report sizes the channel — at least 52.6 million domestic workers globally, heavily migrant and female in the Gulf corridor — but it is a statistical and legal-protection report and does not measure regret among those who stayed. Read this side as a foregone-opportunity exposure estimate, not a survey result; the absence of a direct non-migrant regret survey is the binding constraint, and the rate should be treated as the softest number in the entry.
The two rates produce a narrow regret_delta of 0.09, consistent with a decision that is genuinely close to balanced for the aggregate population of potential migrants — where the economic stakes and personal circumstances vary enormously across individuals. The Gilovich action-dominance pattern reflects the asymmetry in documentation quality: harm to those who did migrate is systematically recorded by HRW and ILO; foregone income for those who stayed is reconstructed from income-gap data rather than direct surveys. The kafala system is under formal legal reform in Qatar and Saudi Arabia since 2020 and 2021 respectively, meaning current workers may face a structurally different risk environment than that captured in the 2018 to 2020 source studies. Nationality and recruitment channel matter substantially: workers recruited directly by verified employers face lower contract-substitution risk than those recruited through third-party brokers, and the entry should not be read as a uniform risk assessment across all kafala-mediated contracts.
Sources: action
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
1/2 sources independently verified verbatim against the cited source
[1]Human Rights Watch — I Was Sold: Abuse and Exploitation of Migrant Domestic Workers in Oman
Verified
Reference source
In interviews with 59 female migrant domestic workers in Oman, HRW found almost all had their passports confiscated, most were made to work excessively long hours, many were not paid their full salaries, and roughly one-quarter said their employers physically or sexually abused them. HRW explicitly makes no statistical prevalence claim for the total worker population.
Excerpt
“Human Rights Watch makes no statistical claims based on these interviews regarding the prevalence of abuse against the total population of domestic workers in Oman. [...] Roughly one-quarter of the domestic workers Human Rights Watch interviewed said that their employers physically or sexually abused them. [...] Almost all the domestic workers Human Rights Watch interviewed said their agents or employers had confiscated their passports. [...] Most of the domestic workers who spoke to Human Rights Watch said their employers required them to work for excessively long periods, and many said their employers did not pay their full salaries.
”
Source data from
2016-07-13
Accessed
2026-05-13
Verification
Excerpt independently re-fetched and confirmed word-for-word against the cited source during our grounding audit.
Calculation
Human Rights Watch, "I Was Sold" (2016), an Oman investigation based on interviews with 59 female migrant domestic workers. This is a QUALITATIVE report: HRW states explicitly that it "makes no statistical claims based on these interviews regarding the prevalence of abuse against the total population of domestic workers in Oman." It therefore does NOT support a population prevalence rate, and the 0.47 action regret_rate is NOT read off this source. What HRW documents is that the specific abuses used as the harm proxy — passport confiscation ("almost all"), excessive hours ("most"), non-payment of full wages ("many"), and physical or sexual abuse ("roughly one-quarter") — are routine rather than exceptional among the workers interviewed. The 0.47 figure is a PROXY for harm significant enough to plausibly generate retrospective regret about migrating, not a measured regret or abuse prevalence rate; no direct "do you regret accepting this contract" survey of kafala domestic workers exists in the published literature, largely because the workers' legal status is controlled by their employer and conditions for candid survey participation are absent.
[2]International Labour Organization — Domestic Workers Across the World: Global and Regional Statistics and the Extent of Legal Protection
Reference source
Statistical/legal-protection report: an estimated 52.6 million domestic workers worldwide (more than four in five women), with substantial gaps in coverage under working-time, minimum-wage, and maternity-protection laws — establishing the scale and legal vulnerability of the migration channel, not Gulf-specific abuse prevalence.
Excerpt
“[Paraphrase from ILO publication — full text available at ilo.org.] The ILO's "Domestic Workers Across the World" presents national statistics and new global and regional estimates on the number of domestic workers — an estimated 52.6 million worldwide, more than four in five of them women — and examines the extent to which they are included in or excluded from key working-conditions laws (working time, minimum wage, maternity protection), finding that substantial gaps in legal protection remain. It accompanies ILO Convention No. 189 and quantifies the labour market rather than investigating specific exploitation mechanisms such as recruitment fraud or contract substitution.
”
Source data from
2013-01-09
Accessed
2026-05-13
Calculation
ILO global domestic worker report (2013). This is a STATISTICAL and legal-protection report — it estimates the size of the domestic-work labour market (~52.6 million workers) and maps their inclusion in or exclusion from working-time, minimum-wage, and maternity laws. It does NOT itself document Gulf recruitment fraud, contract substitution, wage theft, or excessive hours (those abuses are documented on the action side by HRW, not by this ILO report). It is cited only to establish the scale and structural legal vulnerability of the kafala-mediated migration channel. Neither the ILO nor the HRW source provides a population prevalence rate for abuse; the 0.47 action figure is a harm proxy, not a rate read off either source.
Sources: inaction
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
1/2 sources independently verified verbatim against the cited source
[1]World Bank — Remittances Slowed in 2023, Expected to Grow Faster in 2024 (Migration and Development Brief 40)
Verified
Government report
Officially recorded remittance flows to low- and middle-income countries reached an estimated $656 billion in 2023, remaining a crucial source of external finance for developing countries.
Excerpt
“Officially recorded remittance flows to low- and middle-income countries (LMICs) moderated in 2023, reaching an estimated $656 billion. [...] Remittances remained a crucial source of external finance for developing countries in 2023, bolstering the current accounts of several countries grappling with food insecurity and debt issues.
”
Source data from
2024-06-26
Accessed
2026-06-30
Verification
Excerpt independently re-fetched and confirmed word-for-word against the cited source during our grounding audit.
Calculation
World Bank Migration and Development Brief 40 (June 2024). This source establishes only that remittances are a large and crucial income source for sending countries ($656 billion to LMICs in 2023) — it does NOT measure regret among non-migrants. The 0.38 inaction regret_rate is therefore a STRUCTURAL PROXY, not a measured rate: it stands in for the foregone-income exposure faced by adults who declined a Gulf contract in communities where remittances are a primary poverty-exit pathway. No direct survey of non-migrants who specifically regret not taking a Gulf contract exists in the published literature; the regret_display and caveats state this explicitly so the proxy is not read as a measured regret percentage.
[2]International Labour Organization — Domestic Workers Across the World: Global and Regional Statistics and the Extent of Legal Protection
Reference source
Domestic work is a major global employment category — an estimated 52.6 million domestic workers worldwide (excluding 7.4 million child domestic workers), more than four in five of them women — establishing the scale of the migration channel a non-migrant forgoes.
Excerpt
“[Paraphrase from ILO publication — full text available at ilo.org.] The ILO estimates that 52.6 million people worldwide — more than four out of five of them women — are employed as domestic workers, a figure that excludes an estimated 7.4 million child domestic workers under age 15 and likely underestimates the true total. A substantial proportion of domestic workers in the Arab States and other Gulf-bound corridors are international migrants. The report documents the scale of the domestic-work labour market and the extent to which these workers are included in or excluded from key working-conditions laws.
”
Source data from
2013-01-09
Accessed
2026-06-30
Calculation
ILO "Domestic Workers Across the World" (2013), used as a second anchor only to establish the SCALE of the domestic-work migration channel (tens of millions of jobs, heavily migrant and female in the Gulf corridor). This report does NOT measure regret among non-migrants and does NOT contain relative-deprivation findings — it is a statistical report on domestic-worker counts and legal protection. It is cited here purely to size the foregone opportunity, not as evidence of a regret rate. The 0.38 inaction regret_rate remains a structural proxy from the income gap (see World Bank source above); no direct non-migrant regret survey exists.
Caveats
PROXY MEASUREMENTS THROUGHOUT. The proxy_only flag reflects two structural barriers: first, the near-impossibility of conducting direct bilateral regret surveys with a population whose legal status is controlled by their employer and who face retaliation risks for expressing dissatisfaction; second, the absence of any published survey of non-migrant adults in sending communities that directly asks "do you regret not taking a Gulf contract." The 47% action-side rate is a harm proxy, NOT a measured abuse-prevalence rate: the HRW Oman source it draws on is a qualitative study of 59 interviews that explicitly makes no statistical prevalence claim (it reports abuses as routine — "almost all" passport confiscation, "most" excessive hours, "many" wage non-payment, "roughly one-quarter" physical or sexual abuse — but not a population rate). The 0.47 is a stand-in for harm significant enough to plausibly generate regret; workers experiencing these violations may simultaneously view migration as the correct economic decision given their alternatives at home, meaning the proxy likely overstates pure regret. The 38% inaction-side rate is a STRUCTURAL PROXY constructed from income-gap and remittance data, NOT a direct regret survey and NOT a measured regret rate — no published survey asks non-migrants in sending communities whether they regret declining a Gulf contract, so this number carries the widest uncertainty in the entry and should be read as foregone-opportunity exposure rather than observed regret. The regret_delta of 0.09 is within the margin of uncertainty for both proxies and should be read as directional, not precise: this decision is plausibly near-balanced for the overall migrant population, with the action-regret side somewhat higher primarily because the harm documentation is more systematic than the inaction-regret literature. The kafala system is under legal reform in several Gulf states (Saudi Arabia partially reformed it in 2021, Qatar introduced changes in 2020); the risk profile is changing and the documented abuse rates from 2018-2020 studies may not fully reflect current conditions.
