Research / Who Holds the Purse Strings
The Hypothesis
Women play a major role in household financial decisions, and specifically drive gold purchases.
The second of our three founding questions. Deliberately worded to separate two things that usually get collapsed: authority over household money in general, and authority over gold in particular.
Ask a household who manages the money and you'll usually get a shrug and some version of "we decide together." Ask who decided to buy, sell, or pledge the gold, and the answer sharpens immediately. That difference is the finding, and it's why the hypothesis was written the way it was.
It also matters at scale. Widely cited estimates put Indian women's collective gold holdings above the official reserves of several of the world's largest sovereign holders. If women are driving most of those purchases, then a decision made at a kitchen table in Chennai is a small piece of something that registers nationally. That is what makes the gold question and this one inseparable.
Economists stopped treating households as a single decision-maker decades ago. This hypothesis sits squarely in what replaced that.
Older models treated a household as one agent with one set of preferences. Convenient, and wrong. It can't explain why the same income produces different spending depending on who earns or controls it.
Manser and Brown (1980) and McElroy and Horney (1981) recast the household as two agents negotiating, each with their own preferences and fallback position. Influence follows bargaining power, not just income. (Manser & Brown; McElroy & Horney)
Chiappori (1992) went further: households reach efficient outcomes, but through a sharing rule between members rather than one unified will. Our data fits this better than either alternative, because authority turns out to be domain-specific rather than general. (Chiappori, 1992)
If a household splits decisions by category rather than deciding everything jointly, then asking "who controls the money" is the wrong question. You have to ask which money.
We asked about everyday money and gold money separately, on purpose. Verbatim examples:
"In your household, who normally manages everyday expenses such as groceries, bills and school fees?"
"When gold is bought, sold or pledged in your household, who usually makes or influences that decision?"
"When the family makes a major financial decision, who is usually involved and who has the final say?"
"In your household, how much influence do women have over major money decisions? (Usually decide / Joint / Opinion only / Rarely involved)"
"Are there particular financial areas where the women in the family usually take the lead?"
Full instruments on the questionnaires page.
"Women in the family lead strongly on wedding planning, jewelry, gifting... men generally follow the women's lead in these specific areas."
Sathya, high-income segment
Describes her husband deferring to her "without question" on charity and on wedding or event spending.
Padma, high-income segment
Names women as the ones who lead on wedding and event planning, and on gifting decisions.
Ananth, high-income segment
The most strongly evidenced of our three hypotheses. 90 respondents, Chennai, May to June 2026.
88%
of households give women equal or greater say over gold and investment decisions.
43%
decide general household money jointly, the plurality answer and far less decisive.
All
middle-income households with a clear answer credit the wife with driving or sharing the gold decision.
0
women in our middle-income sample. The clearest gap in this study.
The same households answer very differently depending on which money is being discussed.
These turned out to be empirically distinct, and that is the core result. On gold and investment decisions, women hold equal or greater say in 88% of households with a clear answer, split almost evenly between joint and wife-led, with husband-or-elder-led a small minority. On general household money, the picture flattens: joint decision-making is the plurality at 43%, with no comparable concentration. Same households, two different distributions.
This isn't a thin statistical read. Across the wealthier segments, male respondents volunteer it: women lead on jewellery, gifting, weddings, and events, and the men follow. One describes deferring to his wife without question. Another delegates gold buying to her entirely while naming real estate as his own preferred asset. The pattern holds where the money is abundant and the choice is free.
If influence tracked bargaining power alone, we'd expect women's say to rise and fall with earnings across segments. It doesn't move that way. It concentrates by domain, consistently, regardless of income. That is closer to a sharing rule than a negotiation, and it means the honest version of this hypothesis is narrower than "women control household finances." Within gold specifically, though, the evidence is consistent in every segment, with no meaningful contradiction to flag.
The conclusion
Women hold outsized and specific authority over gold decisions, even in households where general finances are more evenly joint.
This is not a separate finding from the other two. The clearest thread running through the dataset is the gold set aside for a daughter's wedding: a single behaviour that is simultaneously contingency planning, a decision women are disproportionately likely to drive, and evidence of a desire to celebrate the occasion well that households save toward years in advance.
The main constraint is size. Three samples of 30, from in and around Chennai across two months, are not statistically representative of South India or of any income group. These findings describe patterns in this dataset.
The sharpest limitation is specific to this hypothesis. Our middle-income sample was entirely male, so the finding that every household there credits the wife with the gold decision is husbands describing wives, not women describing themselves. That is a real gap, it runs in a direction we can't verify from this data, and it's the first thing the next round should fix.
The middle-income instrument also used its own scale for measuring women's financial influence, with no direct equivalent in the other two, so cross-segment comparison here is directional rather than like-for-like.
If you'd be willing to talk to us about how money gets decided in your household, we'd like to hear from you.
Participate in the study