Does Household Labor Supply Insure Against Income Shocks?

A loss of income by a household's primary earner puts pressure on the family budget. In response, households might try to reduce expenses, and the primary earner's spouse might take a new job or increase working hours to compensate for the loss of income. Economists call this the "added worker effect."

It is what economic theory also predicts: Households consume less when they lose income, not only in terms of physical goods but also in terms of leisure time. Yet, it's difficult to detect this effect in the data.¹ This conclusion is not necessarily worrisome: Families may have means — such as personal wealth or government assistance — to insure themselves other than the spouse's labor supply.

But the discrepancy between what theory predicts and what the data tell us may have another source: how we measure changes in household income. Most income changes are predictable, such as career advancements or retirement. If the main earner's income change is expected, the spouse may not react. In addition, when the labor market deteriorates, both partners may face worse opportunities at the same time. Thus, an interesting economic question arises: Can we find an effect on spousal labor supply consistent with economic theory once we isolate income changes that are both unexpected and uncorrelated between spouses?

Measuring Spousal Labor Supply

There are clear policy implications in this question's answer. Governments design unemployment benefits and transfers to help households smooth consumption when income falls. If households can buffer income shocks by shifting more work onto one of its members (namely, the spouse), public transfers may partially replace what families would have done on their own. But if households cannot respond through spousal work — due to, for example, the spouse having limited earning potential, facing caregiving constraints or being unable to enter employment quickly — then transfers are an important insurance mechanism.

Our recent working paper "Spousal Labor Response to Primary Income: Identification and Heterogeneity" relies on detailed administrative data from Norway. The dataset links workers to their employers, which allows us to tackle the measurement problem. The idea is that when a firm experiences a rise or fall in sales, that shock passes through to workers' earnings. This type of income change is likely unexpected and unrelated to the labor market opportunities of the spouse. Thus, our data give us the opportunity to isolate income fluctuations that are more closely related to what economists call "income shocks."

Impact of an Income Change on Spouses

Indeed, when we use our approach, we find a larger spousal response than in traditional studies that do not have our detailed data and, thus, rely on different techniques less suitable to capturing the strength of the spousal labor supply. On average, when the primary worker's income drops by 10 percent, the spouse's employment rate rises by about 1.5 percentage points, and the spouse's earnings rise by about 4.2 percent.

The response is strongest among the most vulnerable households, particularly younger and lower-wealth families. For example, among poorer households, the employment response is about 2pp, and the earnings response is 8.8 percent. This is also consistent with economic theory. Absent alternative insurance mechanisms, the household relies more heavily on spousal labor supply to support its consumption.

Designing Public Transfer Programs

Our paper also discusses how public transfers should be designed. Specifically, we build a model of dynamic household behavior with an emphasis on labor supply. In our model, families choose how much to spend on consumption, how much to save and how much each household member works. The model also incorporates several realistic features:

• Differences in skill
• The accumulation of experience while employed
• Multiple kinds of income shocks: some temporary, some persistent and some correlated across spouses

Using the model, we examine what happens when households experience an unexpected income decline. Without extra government support, families insure themselves by drawing down their financial wealth and increasing the spouse's labor supply. If the government provides additional assistance, households will rely less on spousal labor supply or household wealth. Thus, while helpful, government assistance can also reduce work incentives depending on how the program is structured.

This leads to the following policy idea: Transfers should be tied to employment status for households that are married. For example, if these households have both workers in the labor force, the amount of transfer can be relatively larger than if married households have a single worker in the labor force. The goal is to maintain incentives for spouses to be part of the labor market while always making sure that the household has sufficient support. In our model-based evaluation, this kind of conditional transfer approach can deliver nearly the same welfare but at a lower fiscal cost.

Conclusion

Our findings highlight that spousal labor supply plays an important role in smoothing income shocks, particularly for households with limited options for self-insurance (that is, younger and poorer households with relatively limited financial resources). Based on our findings, we propose a temporary relief program (transfers) that supports consumption and preserves the employment of households who suffer from negative productivity shocks to the primary worker, with additional transfers conditional on working status.

Yongsung Chang is an economics professor at Seoul National University. Elin Halvorsen is a research economist at Statistics Norway. Marios Karabarbounis is a senior economist in the Research Department at the Federal Reserve Bank of Richmond.

To cite this Economic Brief, please use the following format: Chang, Yongsung; Halvorsen, Elin; and Karabarbounis, Marios. (February 2026) "Does Household Labor Supply Insure Against Income Shocks?" Federal Reserve Bank of Richmond Economic Brief, No. 26-06.

This article may be photocopied or reprinted in its entirety. Please credit the authors, source, and the Federal Reserve Bank of Richmond and include the italicized statement below.

Views expressed in this article are those of the authors and not necessarily those of the Federal Reserve Bank of Richmond or the Federal Reserve System.