The myth of a US cost-of-living crisis

The first of two

All kinds of people are talking about the cost-of-living crisis, especially in the West and especially in America. A few minutes of internet search turned up articles or statements decrying this phenomenon from the Trump White House, Bernie Sanders, the Heritage Foundation, the Atlantic, the Washington Post, the New York Times, the BBC, numerous academic articles, etc.

This kind of talk dates back some time, actually. Paul Krugman wrote on his blog in 2014 that “wages for ordinary workers have in fact been stagnant since the 1970s.” Thomas Piketty’s Capital in the Twenty-First Century (2013) made the same argument, namely that despite steady GDP growth, technological progress, and relative social stability, life in Western society is no better, and possibly even worse, than it has been for the last several decades. All the economic gains have accumulated to a small group of the world’s wealthiest.

I think that the sentiment has gotten appreciably louder since 2020, driven by worries about inflation and home price appreciation. Arguably, this was a central issue in the 2024 US presidential election, and the 2025 Canadian general election. There is vigorous debate about the causes: insufficient taxes on the wealth, stifling government regulation, government deficit spending, corporate greed, and technological change are popular and conflicting explanations. But everyone seems to agree that a combination of slow wage growth and rising prices, especially the prices of housing, food, education, and child care, have meant that for the first time in many decades, the bulk of today’s young people are materially worse off than their parents and grandparents.

Well, I don’t agree.

Now, I cannot confidently say that every specific claim about rising costs, falling wages, and declining economic mobility are wrong. But I am fairly confident that many components of the overall argument are exaggerated or incorrect, and when this is corrected, the overall thesis is discredited.

I’m far from the first person to push back against this idea. The economist Russ Roberts has a wonderful series of animated videos on the topic, and this was an influence on my thinking. Nevertheless, I think putting out my own rebuttal is worthwhile. For one thing there are innumerable and ever-new fallacious arguments for this thesis, so no one can refute them all. For another, while some folks, including Russ, have discussed the broader principles that explain why so many have come to incorrect views on the cost-of-living-crisis idea, I think a specific, practical outworking is often more persuasive.

This is the first of two essays on the topic. In each, I’ll directly challenge a particular commentator’s argument for the cost-of-living-crisis thesis, then turn to the broader themes underlying our disagreement. In this, cost of living generally. In the next, housing costs in particular.

---

I have been following the work of the social media videographer and real estate agent Fred Smith for some time. I think his style and message perfectly capture the populist mood on this topic. His big ideas are that 1) Millennials and adult members of Generation Z (i.e. people born from ~1980 to 2005) with low and midrange incomes are unable to save money or afford the lifestyle they grew up with, and 2) members of this group have faced a completely different environment than the generations that came before, because those groups have benefited from rising home prices and have not been adversely affected by recent rises in child care, rent, and education costs. Overall, he argues that this is primarily because of macroeconomic change, and beyond the control of the typical American.

A recent short video from Mr. Smith argues against the claim that the cost-of-living crisis for young people is due to overspending. He lays out a hypothetical monthly budget for a single person making $70,000 annually. After living expenses, health insurance, retirement account contributions, and taxes, and assuming no travel, dining out, or other luxuries, he says that nothing remains for saving, making homeownership, parenthood, travel, and social mobility unreachable.

Is this a reasonable conclusion? Let’s take each part of our everyman’s budget in turn. I am paraphrasing each part of his video, but you can certainly watch it yourself.

1. Rent

   FS: Do young people overspend on rent? On the contrary. Even after the major concession of having a roommate, rent with utilities for a 2-bedroom apartment takes up $1,500/month in the budget.

   JN: This is the most egregious part of the entire video. Rental prices vary wildly with location and quality, making a global estimate difficult. I know that Mr. Smith lives in the Orlando, Florida area, so I searched for 2-bedroom apartment on Zillow in Orlando city limits. I got 811 results under $1800/month. That alone should refute his figure, but I went further. I did my best to choose an apartment in a nice building (a relatively modern-looking tower, which will be secure and amenity-rich) in a nice neighborhood (Reddit, a good source of local opinion, agrees that South Orange is a good one). A two bedroom floor plan here is currently listed for $1,630/month. For utilities, this website suggests a typical cost of $388/month in Orlando, and this is consistent with other estimates. Divided by two, the total cost is $1009/month, a far cry from Mr. Smith’s $1500.

2. Car payment

   FS: A loan on a $12,000 - $15,000 used car will run you about $400/month.

   JN: NerdWallet has current auto loan rates. I assumed the second-best credit score category (9.95% APR), a $15,000 purchase price, 25% down, and a 5-year loan term. I used an online calculator and found that the payment for such a loan will be $258/month. In fact, in order to get a payment above $400/month for a $15,000 car, you have to assume zero money down and a credit score of less than 600. I would not recommend anyone buy a car at that price under those conditions.

3. Car insurance and fuel

   FS: Combined, these will cost about $300/month.

   JN: Bankrate says that minimum car insurance coverage in Florida costs about $93/month. Without looking deeply into the situation in Florida, I’d say that it’s often wise to buy more than the minimum amount of liability insurance, as the minimum is often insufficient to repair or replace the other party’s vehicle if it’s an expensive car. I’ll add $20/month for higher liability limits.

   For fuel, I did a pretty thorough simulation. First, I decided our hypothetical Millennial works at the Trader Joe’s in Doctor Phillips. I think this represents a reasonable workplace for someone with a college degree, but without exceptional job skills. This is an 11.1 mile drive from our chosen apartment — and if you’re renting, there’s no reason to live farther than this from your workplace. I assume one round trip there every day, to account for some weekend driving. I decided that a Honda Civic is a pretty typical car, and searched AutoTrader.com for one that cost $15,000. The first result was a 2018 automatic model. These get 30 mpg city. GasBuddy showed gas prices averaging around $3.00 in Orlando on the day of the search. Converting to a monthly figure, fuel expenditures come out to $67/month. In total, we’ve got $180/month for fuel and insurance.

4. Student loan

   FS: Assuming a $40,000 student debt, the payment would be about $500/month.

   JN: This hypothetical person has finished their education, so they took out their loan some years ago. If they started college in 2018, their interest rate with an unsubsidized federal loan was 5.05%. The Federal Student Loan Simulator takes a number of inputs, but all of the important ones were already determined in this scenario. The resulting monthly payment for a 10-year, $40,000 loan was $425/month.

5. Groceries

   FS: This will be $500/month, without any eating out.

   JN: While it’s not difficult to spend that much, it’s not necessary. There are numerous Reddit threads about grocery costs in every major US city. The ones in Orlando frequently suggested that $150/week for a family of two was quite doable. I’m going to assume our subject can split some, but not all, meal costs with the roommate, and settle on $100/week for one person, or $400/month.

6. Personal care

   FS: For doctor’s appointments, haircuts, gym memberships, etc., $150/month.

   JN: First, if you’re on a tight budget, don’t have a gym membership. Second, in my opinion, medical costs for a person in their 20s with health insurance ought to be extremely low. As we’ll see later, Mr. Smith assumes a $300/month insurance premium. For that price for a single person, it’s probably not a high-deductible plan, and it’s going to be difficult to spend a huge amount. Prudential says that a low deductible is $1,650/year or less. I’m going to assume our subject, a very young person, does not hit this every year. Let’s take half that and amortize it. It comes out to $69/month. We’ll add in $15 for “personal care,” in this case a haircut every two or three months, for a total of $84/month.

7. Phone

   FS: Mr. Smith actually combines this with the next category, but I did not. I’m going to assign $50/month of his combined item to this.

   JN: Many people spend $100 or more per month on a phone, but it’s pretty well known that you don’t need to do this. If you bring your own phone, even the higher-end plans at MVNOs like Mint, Fi, Visible, etc., are in the $30-50/month range. I chose the high side: $50/month.

8. Miscellaneous

   FS: The residual of his combined phone/misc. category is $100/month.

   JN: I want my hypothetical subject to be frugal, but have some money for things that most people do. Mr. Smith says there’s nothing left over for travel, eating out, clothing, or car maintenance. I’m going to allocate $720/year for clothing, $500/year for car maintenance, $720/year for eating out (that’s $60/month), and $1000/year for travel. This comes out to $245/month.

9. Health insurance

   FS: Health insurance will be $300/month.

   JN: This varies widely depending on one’s employer (in the US), and this isn’t under one’s direct control. Rather than try to make any assumptions, I’m just going to accept his figure of $300/month.

10. Retirement account

    FS: A healthy contribution to a 401(k) is $500/month.

    JN: I quite agree with Mr. Smith here, although I’ll note that this is 8.6% of gross income, which is more than most people in this situation would actually contribute. But I think people should save more for retirement, so we’ll keep this at $500/month.

11. Taxes

    FS: Mr. Smith doesn’t account for this explicitly, but his implied monthly tax bill (given he’s arguing nothing is left over) is $1533/month. Admittedly, his visual says that the budget will have $0-500/month left over, so I think he’s aware that this is a high estimate. In the video itself, he says that $0 is left, but later he assumes this person can, in fact, save money, so I think he’s just trying to have it both ways. Let’s take his lower tax estimate of $1033/month.

    JN: An online paycheck calculator says that a person making $70,000/year in Orlando, with $800 of pre-tax deductions, owes $1049/month in taxes.

12. Saving

    FS: He says that nothing is left for savings, but later in the video, he imagines that despite this, our subject is able to save $5,000/year or, $417/month. Using his lower (and more accurate) implied tax figure would make sense of this.

    JN: In my analysis, there is $1,333/month remaining.

Here’s a summary table of his budget and mine.

---

Mr. Smith concludes by suggesting that today’s young person, even if they eschew all luxuries, either cannot save at all, or can only save a few thousand dollars per year, and will spend most of their adulthood working to afford a home. He considers a hypothetical house down payment of $50,000. Combined with their car and student loan debt, this comes to $105,000, which would take 21 years to accumulate this with $5,000 of annual savings, the implication being that this is an unreasonably long period.

Astute readers will realize that he is actually double-counting the car and student loan debt, since those debts will be satisfied in 5 and 10 years, respectively, with the money already allocated for them in the budget. That aside, he’s saying it will take our subject 10 years to save for a house.

In contrast, in my budget the subject is saving $1,333 per month, without giving up the occasional meal out, vacation, or clothing purchase. I wouldn’t recommend pouring all of their remaining income into a housing fund, but if we assume that they bank $1,000 per month for this purpose, and earn 6% interest on that money (representing some mix of stocks and bonds), they will have $70,000 after five years. If some of that was diverted to early payments on the ten year student loan, that debt could be retired and roughly $50,000 for a down payment saved in five years, half the time Mr. Smith claims.

Not to mention that Mr. Smith’s analysis assumes zero real wage growth for our subject over those 10 (or is it 20?) years. It should be intuitively obvious that this is a ridiculous assumption, but it is supported by data, such as US Census Bureau’s, which shows that earnings peak in the 5th or 6th decade of life for most people (see their report on income, Table A-1). I didn’t bother to question this assumption in the savings exercise, but obviously doing so would just make the picture that much better and his analysis that much further from reality.

In sum, he purports to show that even on a spartan budget, a single young American with a modest income is unable to meaningfully save money, and their financial security is thus largely beyond their control, because the cost of living has grown out of proportion to income. However, under a different and more careful set of assumptions, the exercise seems to show the opposite — such a person is capable of saving about a quarter of their income without sacrificing the small luxuries of American living.

---

Although his approach has the veneer of objectivity, Mr. Smith fails to explain any of his choices. As I’ve described, he has taken a number of liberties in his analysis, namely:

1. Objectively exaggerating some costs, particularly rent, car loan payment, and student loan payment

2. Equivocating on taxes by indirectly conceding that that his subject might have $500 left over each month, but then speaking as if there is nothing left over

3. Double-counting the auto and student loan debt

4. Assuming zero real wage growth over time

Once these errors are corrected, his argument falls apart.

While Mr. Smith must be aware to some degree that he’s stretching the truth, I think that he sincerely believes in his message. It’s hard to be sure what he’s thinking, because his videos are generally very short and have little explanatory detail. However, I believe that he is basing his figures on aggregate national statistics, an approach very common among decriers of the cost-of-living crisis, and an erroneous one. I think his “budget” is simply the combination of national averages obtained from a variety of sources through cursory internet searches. I do not think this is a robust method.

He appears to take this approach in many of his videos, such as those where he compares median or mean income to median or mean home and rental prices. The implicit assumption that the average homebuyer is the average income earner seems reasonable at first, but it is not. Consider a few confounding factors:

1. Wealthier people are more likely to own and/or rent multiple homes, making them a disproportionately large part of the housing market and pushing up average prices

2. Poorer people are more likely to have roommates, making them a disproportionately small part of the housing market (because two people might sign only one lease), also pushing up average prices (compared with a world where everyone lives alone)

3. The set of all wage earners includes millions of low wage earners who do not participate in the housing market at all: teenagers living with their parents, college students living in dormitories, working homeless living in shelters or on a friend’s couch, active duty military personnel living on military bases, retirees who own their house, bought it when they made a lot more money than they do now, and will stay there until they die, etc. This pushes down the average wage vis-a-vis the average home price.

The average renter or homebuyer is, in fact, quite a bit wealthier than the average wage earner, and that gap changes over time, so you cannot simply line up these two distributions.

Mr. Smith is, of course, only a representative. I think these same problems pervade most arguments about the cost of living crisis. Aggregate statistics are difficult to interpret, because they necessarily make lots of assumptions. “The average American homebuyer” is an artificial construct, and if things are going poorly for that entity, it is not clear what that means for the state of the country or anyone’s individual experience. Other, more well-supported claims about the cost of living crisis are doubtless being made, but I would encourage readers to carefully consider whether this fallacy applies there as well.

What, then? Am I some kind of national statistics nihilist? Not quite. My position can be described this way:

1. The best way to answer a complex question is to carefully define all relevant terms, then conduct a well-designed study to isolate two variables and prove whether one thing causes another thing (i.e., a scientific experiment).

2. If a prospective study is not possible, high-fidelity aggregate statistics can, to a much lesser extent, shed light on trends over time with well-defined objects of study, though even then, a great deal of caution is needed, and they usually cannot prove causality.

3. If all we have are case studies and low-quality aggregate statistics and/or poorly defined variables of interest, case studies are superior to low-quality aggregate statistics, and it is better not to even try to use the statistics.

Thus, I would argue that my simple exercise of actually estimating the cost of living for a very specific hypothetical scenario is superior to Mr. Smith’s method of breezily concatenating a bunch of aggregate macroeconomic statistics.

Perhaps the overblown worries about cost of living crisis are on the rise because these kinds of statistics have become more available and the subject of far more popular interest than they were in the past. I don’t have strong evidence on whether this is true, but it’s an interesting theory that I’ll consider and maybe write more about in the future. But even if overapplication of statistics isn’t on the rise, it’s certainly happening, and it’s certainly involved in a lot of the discussion on this topic.

Using these kinds of statistics to measure secular trends over time (e.g. “the middle class has had stagnant earnings since 1975!”) is even more fraught, and I’d advise against even attempting it. In the next post, I’ll go through some examples and show why this is.