Home › What the research says

Lifestyle Creep: What the Research on Social Comparison Shows

Spending up toward the top

Marianne Bertrand and Adair Morse published Trickle-Down Consumption in the Review of Economics and Statistics in 2016. They document that non-rich households consume a larger share of their current income when exposed to higher top income and consumption levels, and that permanent income, wealth effects, and local price pressure cannot fully explain it. What supports the status interpretation is the composition: the budget shares that non-rich households devote to more visible goods and services rise with top income levels. They estimate middle-income households would have saved between 2.6 and 3.2 percent more by the mid-2000s had top incomes grown at the median rate since the early 1980s, and report that exposure to higher top incomes predicts more self-reported financial duress and more personal bankruptcy filings.

A lottery that randomizes your neighbors

Correlational work on neighborhoods always has a selection problem, so the Dutch Postcode Lottery is unusually valuable. Each week it randomly selects a postal code and distributes cash and a new BMW to lottery participants in that code. Peter Kuhn, Peter Kooreman, Adriaan Soetevent, and Arie Kapteyn studied the results in the American Economic Review in 2011. Winners' own consumption effects were largely confined to cars and other durables, and most BMW winners liquidated the car. The finding that matters here is the social one: lottery non-participants who lived next door to winners had significantly higher levels of car consumption than other non-participants. They did not win anything. Their neighbor did.

And then the bankruptcies

Sumit Agarwal, Vyacheslav Mikhed, and Barry Scholnick took the same idea further in the Review of Financial Studies in 2020, using lottery prizes of random dollar magnitudes in very small Canadian six-digit postal codes, which contain a median of 13 households and are often smaller than a city block. They report that a larger prize won by one neighbor increases subsequent borrowing and bankruptcy filings among the other neighbors, and give a magnitude: a neighbor's win equal to the sample's median annual income, about C$29,000, raised bankruptcies among close neighbors over the following two years by roughly 6.6 percent relative to the local average. The mechanism evidence is the striking part. Prize size increased the value of visible assets such as houses, cars, and motorcycles on neighboring bankruptcy filers' balance sheets, but not invisible assets like cash and pensions.

What this does and does not mean for you

These are population-level causal estimates, not a diagnosis of any individual. Nobody has shown that you specifically bought something because a neighbor did, and the effects are averages across many households, most of whom did nothing at all. The studies also cover particular settings, a Dutch lottery, a Canadian province, and US consumption data over decades. What survives all of that is the direction and the pattern: exposure to other people's visible consumption raises spending, the increase concentrates in visible goods, and it shows up downstream as reduced savings and more distress. That is enough to justify a structural defense without any moralizing about willpower.

The envelope change to make

The trigger in this research is visible consumption by people near you, which you cannot control. The variable you can control is where new money lands by default. Set envelope amounts as percentages of take-home pay rather than as fixed dollar figures, and give savings a percentage like every other envelope. Then a raise increases the savings envelope automatically, in proportion, before you decide anything. Under fixed dollar amounts, a raise has no assigned home, so it flows into the variable envelopes where the comparison effects live. Decide the split once while nothing has changed, and let the arithmetic handle the next raise.

Sources

Every source below was retrieved and checked. Findings are reported as the source states them.

• Trickle-Down Consumption — The Review of Economics and Statistics, 2016
  Non-rich households consume a larger share of current income when exposed to higher top income levels, with budget shares of more visible goods particularly responsive; middle-income households would have saved 2.6 to 3.2 percent more by the mid-2000s had top incomes grown at median rates, and exposure to higher top incomes predicts more self-reported financial duress and more personal bankruptcy filings.
• The Effects of Lottery Prizes on Winners and Their Neighbors: Evidence from the Dutch Postcode Lottery — American Economic Review, 2011
  In a lottery that randomly selects a postal code and distributes cash and a BMW to participants there, winners' consumption effects were largely confined to cars and other durables, and lottery non-participants living next door to winners had significantly higher levels of car consumption than other non-participants.
• Peers' Income and Financial Distress: Evidence from Lottery Winners and Neighboring Bankruptcies — The Review of Financial Studies, 2020
  Using randomly sized lottery prizes in very small Canadian six-digit postal codes containing a median of 13 households, a neighbor's win equal to the sample's median annual income (about C$29,000) raised bankruptcies among close neighbors over the following two years by about 6.6 percent relative to the local average, and prize size increased the value of visible assets such as houses, cars and motorcycles on neighboring bankruptcy filers' balance sheets but not invisible assets such as cash and pensions.

Common questions

Is lifestyle creep actually a real effect or just a saying?

There is causal evidence for the social comparison part specifically. The Dutch Postcode Lottery randomly assigns which postal code wins, and non-participating neighbors of winners bought more cars than comparable non-participants elsewhere. That design rules out the usual explanation that similar people cluster together. Bertrand and Morse find the same direction in US consumption data, concentrated in visible goods. What no study measures is the version people usually mean, which is your own spending drifting up after your own raise.

Why is it always cars, houses, and visible things?

Because that is where the effect shows up in the data. Bertrand and Morse report that the budget shares non-rich households allocate to more visible goods and services rise with top income levels. Agarwal and colleagues found neighbor lottery prizes raised the value of visible assets like houses, cars, and motorcycles on nearby bankruptcy filers' balance sheets, but not cash or pensions. Invisible consumption does not respond the same way, which is the clearest sign that this is about signaling rather than income.

Does a percentage-based budget prevent lifestyle creep?

No study has tested that, so no honest answer says yes. What a percentage split does mechanically is assign a destination to money you do not have yet. Fixed dollar envelopes leave a raise unassigned, and unassigned money defaults into whichever categories are elastic. Percentages route a fixed share into savings without a new decision. That is a claim about arithmetic and defaults, not a claim about your psychology, and it is the part you can verify yourself.

Related