Lifestyle creep. Keeping up with the Joneses. Raising your spending after an increase in your income.
No matter what you call it, lifestyle creep is one of the most talked about topics in personal finance. And you won’t find a shortage of financial gurus (myself included) warning you against the dangers of it.
But is lifestyle creep a real concern for most people? Or has this problem been grossly exaggerated based on a small number of super-spenders?
For most of my writing career, I couldn’t tell you the answer with any certainty. But, now I can. I got my hands on a data source that records the income and spending of the same set of households over time. So instead of looking at income/spending snapshots at one point in time, I can see how households change their income/spending over time. I can see how much U.S. households increase (or decrease) their spending after an increase in income.
And what I found surprised me. Let’s dig in.
How Much Does Spending Change After an Increase in Income?
For this analysis, I’ll be using the Panel Study of Income Dynamics (PSID), which follows the same set of households over time. Not only do I have household income data, but I also have spending data from 1999–2021.
To analyze how much spending changes after an increase in income, I’m going to look at all households that increased their income over a 10-year period and drop any that are 65 or older by the end of the period. After all, I don’t want to pick up spending changes for those entering retirement. The goal is to see what happens with spending for working households that raise their income over a given 10-year period.
One such period is plotted below. This shows the absolute change in inflation-adjusted income (on the x-axis) and the absolute change in inflation-adjusted spending (on the y-axis) for all the households with positive income changes over the period 1999–2009. I’ve also colored each household based on whether their spending decreased (“No Lifestyle Creep”), their spending increased less than 75% of their income (“Some Lifestyle Creep”), or their spending increased by more than 75% of their income increase (“Serious Lifestyle Creep”):
As you can see, there are a number of households that had “Serious Lifestyle Creep” over this period. Though more households seem to have “Some” or “No” lifestyle creep, the number of red dots is not trivial. You might think that this proves that lifestyle creep is a serious problem, but this is only one 10-year period in the data.
For example, if I were to plot the same chart but for the period 2005–2015, you’d come to a different conclusion:
Now, you can see that there are far fewer households with “Serious” lifestyle creep. But neither of these plots is more accurate or true than the other. They each have their own biases based on the time periods in question.
Either way, these plots aren’t fully representative of the financial behavior of U.S. households. Why? Because the PSID data has weights for each households which adjust for sampling bias. Some households are overrepresented in the data and some are underrepresented. The weights adjust for this.
Unfortunately, we can’t visualize these weights that easily, but we can use the weights to calculate summary statistics across all time periods in the data. Therefore, instead of focusing on one period or worrying about sampling bias, we can analyze all 10-year periods while using the household weights. This will adjust for time period and sampling bias.
For example, if we were to summarize the 25th, 50th, and 75th percentile changes in spending based on different changes in income across these households over all 10-year periods, we would see the following:
Income Increase | Spending Change (25th Pct.) | Spending Change (50th Pct.) | Spending Change (75th Pct.) |
---|---|---|---|
<$10k | -$12,060 | $1,022 | $12,165 |
$10k-$25k | -$7,402 | $4,254 | $20,981 |
$25k-$50k | -$4,327 | $8,864 | $21,248 |
$50k-$100k | -$9,890 | $14,225 | $32,977 |
$100k+ | -$101 | $24,139 | $48,194 |
This shows for each 10-year income change bucket (e.g. <$10k, $10k-$25k, etc.) how much spending changed for the 25th percentile household, the median household (50th percentile), and the 75th percentile household.
For example, for those households that saw their income increase by $10k-$25k over a 10-year period, 25% of them decreased their spending by $7,402 or more. Half of these households increased their spending by more than $4,254. And 25% of them increased their spending by $20,981 or more.
Overall, you can see that spending drops or only increases slightly for most of the income change buckets in the data. Only at the 75th percentile for some of the income brackets does spending keep up with the overall change in income. This demonstrates that only a smaller percentage of households significantly increase their spending following an increase in income. This is the kind of lifestyle creep that I (and many others) have warned against. Thankfully, it appears that the fear of lifestyle creep is worse than the lifestyle creep itself.
Why Lifestyle Creep Shouldn’t Be Feared
After reviewing the data, I can say that lifestyle creep doesn’t seem to be a major issue for most people. In general, as households grow their income, they either spend less overall or increase their spending slightly over time. It’s a small minority of households that truly “creep” their spending upward. This doesn’t mean that lifestyle creep doesn’t exist, but that the fears of it are greatly exaggerated.
Of course, if you define lifestyle creep as any increase in spending following an increase in income, then, yes, lifestyle creep is quite common. However, I don’t believe that’s the intent behind the lifestyle creep warnings. Getting a big raise and then adding guacamole to your regular Chipotle order is technically lifestyle creep. But I don’t think that’s the kind of thing that is going to send you into a financial tailspin.
No, the intent behind the warning against lifestyle creep is fear. It’s the fear that you’ll endlessly increase your spending as your income goes up. Fortunately, as the data illustrates, very few behave like this. Just make sure you aren’t one of them.
Trust me, I didn’t expect this result when I first saw it either. I’ve previously warned against the dangers of too much lifestyle creep. Thankfully, now I know that these dangers are more based in fantasy than reality.
You might disagree with me based on some people you know or what you’ve seen in your line of work and that’s fine. Unfortunately, the aggregate data doesn’t agree with you. And, unless you have another data source that shows otherwise, your anecdotes simply don’t compare to the best panel data we have on the topic.
By nature we are wired to remember the exceptions in our world. If we didn’t, then we might end up as another animal’s lunch. Fortunately, we don’t live in that world anymore, but our brains have yet to evolve to see things differently.
Until then, enjoy a little lifestyle creep. It’s not as big of a problem as you think.
Thank you for reading.
If you liked this post, consider signing up for my newsletter.
This is post 408. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data