How households are studied when income does not arrive evenly

Bills tend to arrive on a schedule. Research on uneven income asks how the two calendars are reconciled, using methods that range from year-long diaries to formal saving models.

A year of diaries

Jonathan Morduch and Rachel Schneider ran the United States Financial Diaries, following a few hundred households across roughly a year and recording money coming in and going out in detail rather than asking for an annual summary. The published account describes months well above and well below a household average within the same year, and describes informal arrangements built around that pattern: timing bills against expected inflows, lending and borrowing inside families, and holding money in forms that are not counted as savings. The design trades breadth for depth. The sample was purposive rather than representative, was concentrated in a small number of sites, and is too small to support national estimates, all of which the authors state.

A model of the buffer

Christopher Carroll developed the buffer-stock account of saving, in which a household holds a target ratio of liquid assets to income and spends down or rebuilds that buffer as income moves. One implication is that consumption should follow income closely when the buffer is thin and much less closely when it is not. The model has been influential because it produces behaviour resembling what is observed, but it rests on assumptions about preferences and about expectations of future income that cannot be observed directly and are argued about on their own terms.

Liquidity is not the same as wealth

Greg Kaplan and Giovanni Violante described a group of households holding substantial illiquid assets alongside very little accessible money, and showed that these households respond to income changes in ways that resemble households with no assets at all. The argument matters for this subject because a measure of net worth can classify a household as comfortable while the money available in a given week behaves quite differently. Estimates of how large this group is vary with the definition of liquidity used, which is an active point of disagreement.

The same distinction affects how a year of records is read. An annual total conceals the order in which money arrived, and the order is what determines whether an obligation falling due in a particular week could be met. Two households with identical annual figures can face quite different sequences, which is the reason diary and account-level methods were developed alongside annual survey measures rather than as replacements for them.

Survey measures and the dispute about them

The Federal Reserve's Survey of Household Economics and Decisionmaking asks how an unexpected expense of a few hundred dollars would be handled, and the answers are regularly cited. Annamaria Lusardi, Daniel Schneider and Peter Tufano developed a related measure of financial fragility. Both are hypothetical questions, and the critique attached to them is that stated answers about an imagined expense need not match behaviour when a real one arrives. Comparisons with administrative and account data have been used on both sides of that argument without closing it.

What remains unsettled

The direction of influence between uneven income and thin buffers is not established. Income that moves sharply makes a buffer harder to accumulate, and a thin buffer makes the same movement more consequential. Separating the two requires either a long panel or a change imposed from outside, and the research base contains few of either. The studies described here are best read as complementary partial views rather than as a settled account.

Back to the library