Mental accounting and the fungibility problem

Fungibility is one of the quieter assumptions in economics. A unit of currency is a unit of currency: its source, its label and the mental category it was filed under should have no bearing on how it is spent. Richard Thaler's work on mental accounting, developed across papers in 1985 and 1999, argued that households routinely violate this assumption, sorting money into notional accounts — housing, food, entertainment, windfall — and treating the boundaries between those accounts as real.

The original evidence was largely experimental, and experimental violations of fungibility are easy to produce and easy to dismiss. The more interesting question was whether the pattern survives outside the laboratory, where the money is real and the stakes recur.

Two field studies

Justine Hastings and Jesse Shapiro examined household responses to petrol price changes in a 2013 paper in the Quarterly Journal of Economics. When petrol prices rose, households switched to lower-octane grades far more than a standard model would predict. The magnitude matters here: for a typical household, a fuel price increase is a small change in total wealth, and the wealth effect alone should produce almost no substitution within the fuel category. The observed switching was an order of magnitude larger than that prediction. The behaviour is consistent with households operating a category budget for fuel and adjusting within it, rather than reallocating across the whole budget.

The same authors returned to the question in 2018, in the American Economic Review, using data on the Supplemental Nutrition Assistance Program. The relevant comparison is among households already spending more on food than their benefit amount. For those households, the benefit is inframarginal — economically equivalent to cash — and fungibility predicts that a dollar of benefit and a dollar of cash should have the same effect on food spending. The measured propensity to spend on food out of benefits was substantially higher than out of cash income.

Johannes Abeler and Felix Marklein reached a similar conclusion in a 2017 Journal of the European Economic Association paper, combining a field experiment with labelled vouchers and a survey-based test, and finding that the labelling of money affected how it was spent.

The objections

Three arguments run against a purely psychological reading.

The first is liquidity. A household that cannot borrow behaves in ways that resemble mental accounting without any category boundaries being involved. Distinguishing the two requires knowing the household's actual access to credit, which is rarely observed cleanly.

The second is programme design. Benefit rules, restricted-use cards and transfer timing impose real constraints. Some behaviour attributed to mental categories may be the trace of an administrative rule.

The third is generalisation. Effect sizes in this literature vary widely across contexts, and the conditions under which labels bind — how large the labelled sum is relative to the budget, how salient the label, how long it persists — are not well characterised. A pattern reliably observed in fuel purchasing and food benefits has not been shown to hold uniformly across categories.

Reading the literature

What the field evidence supports is narrow and reasonably durable: money is not always treated as interchangeable, and the deviations are large enough to matter for policy design. What it does not support is the broader claim, common in popular summaries, that mental accounting is a general error to be corrected. In several documented cases the category boundary functions as a commitment device, and households that maintain it are not obviously worse off than those that do not.

Sources

• Thaler, R. H. (1985). Mental Accounting and Consumer Choice. Marketing Science, 4(3).
• Thaler, R. H. (1999). Mental Accounting Matters. Journal of Behavioral Decision Making, 12(3).
• Hastings, J., & Shapiro, J. M. (2013). Fungibility and Consumer Choice: Evidence from Commodity Price Shocks. Quarterly Journal of Economics, 128(4).
• Hastings, J., & Shapiro, J. M. (2018). How Are SNAP Benefits Spent? Evidence from a Retail Panel. American Economic Review, 108(12).
• Abeler, J., & Marklein, F. (2017). Fungibility, Labels, and Consumption. Journal of the European Economic Association, 15(1).