The other pieces published in this series have human protagonists. This one doesn’t: The main character of this piece is not a person but a number. Like all the facts and numbers cited above, it comes from the federal government. It’s a very important number, which has for a century described economic reality, shaped political debate and determined the fate of presidents: [the consumer price index](https://www.bls.gov/cpi/?itid=lk_inline_enhanced-template).
The CPI is crucial for multiple reasons, and one of them is not because of what it is but what it represents. The gathering of data exemplifies our ambition for a stable, coherent society. The United States is an Enlightenment project based on the supremacy of reason; on the idea that things can be empirically tested; that [there are self-evident truths](https://www.archives.gov/founding-docs/declaration-transcript?itid=lk_inline_enhanced-template); that liberty, progress and constitutional government walk arm in arm and together form the recipe for the ideal state. Statistics — numbers created by the state to help it understand itself and ultimately to govern itself — are not some side effect of that project but a central part of what government is and does.
**Author John Lanchester, interviewed by Michael Lewis**
“There is something very important about that idea of a will to fact, truth, the idea that you can live better by knowing better.”
Like the democratic project itself, many of these numbers are imperfect. Gathering them — in many cases, creating them — gets more complicated over time. The CPI is a classic example of that.
Giving up on the work can seem tempting. Complicated times increase our hunger for simple answers, for easy certainties and “[alternative facts](https://www.washingtonpost.com/news/monkey-cage/wp/2017/01/31/yes-there-are-alternative-facts-thats-different-from-falsehoods/?itid=lk_inline_enhanced-template).” But if we abandon the work of creating statistical truths, we risk abandoning democracy, too.
The U.S. government does many, many things, and among the principal things it does is to categorize and measure and record and, above all, count. Wow, do the feds count a lot of stuff! Wages and illnesses, water quality and weather data, selenium levels in wild turkey dark meat (raw and cooked), the population of Ulaanbaatar, the principal exports of Mauritania, infant mortality in Chad, metrics for the Army’s next-generation sniper rifle, disparities between the industrial accident rates of U.S.-born and foreign-born people of Hispanic origins, flavonoid contents for various types of olive oil. And on and on.
The government generates a Niagara of data, and you can soon get to the point of believing that it’s harder to find things that the government doesn’t count than things it does. There might be systems that are more transparent than those of the United States: those of Norway and Sweden, for instance, which make everyone’s [tax returns publicly available](https://www.elibrary.imf.org/display/book/APTBV27N03/CH001.xml?itid=lk_inline_enhanced-template#:~:text=In%20contrast%2C%20in%20countries%20like,publicly%20available%20for%20many%20years.). Disclosing anyone’s tax return without authorization [is illegal in the United States](https://www.law.cornell.edu/uscode/text/26/7213?itid=lk_inline_enhanced-template). But for sheer volume and variety of data coming out of the state, America has no rival.
A consumer price agent checks prices for the consumer price index on Sept. 21, 1971. (Thomas J. O’Halloran/Library of Congress Prints and Photographs Division)
The production and publication of these numbers are not a byproduct of the U.S. government. They are a core aspect of the state’s identity. The United States has the oldest written and codified constitution, and, not by chance, it is also the first country to have the act of counting built into its constitution: Article I, Section 2, Clause 3: [the enumeration clause.](https://constitutioncenter.org/the-constitution/articles/article-i?gad_source=1&gclid=Cj0KCQjww5u2BhDeARIsALBuLnOe0zWk4RrZUEyq4qlvukIeFB68RChQ8D8RWZlWlmXys2W3AHDK7JQaAg-LEALw_wcB&itid=lk_inline_enhanced-template) To allocate the power, you have to count the voters. It is the rational, democratic, scientific thing to do — but as the complex, contested, ever-evolving history of the U.S. cCcensus shows, it is simpler to state the idea and the principle than it is to achieve a fully accurate count of, well, pretty much anything.
For the authors of the Constitution, some truths were self-evident. When it comes to the creation and use of statistics, that is often not true. Every statistician has a favorite story about how data can be misunderstood, and one of the most popular examples comes from World War II. Allied planes were coming back from raids over German-occupied Europe riddled with bullet holes. The damage was concentrated on their wings and fuselages. There was no evidence of damage to engines. Repair efforts were therefore directed toward wings and fuselages because that was obviously where the main damage was being sustained. [Abraham Wald, a statistician](https://www.historyofdatascience.com/abraham-wald-a-statistical-hero/?itid=lk_inline_enhanced-template), arrived from Columbia University, looked at the evidence and said, not so fast. The planes with the Swiss cheese damage to their wings and fuselages — those were the ones making it home. They were seeing no planes with damaged engines because those aircraft weren’t making it home. Conclusion: Don’t reinforce the places where you saw bullet holes, reinforce where you didn’t.
## Community corner
Statistics always need context and explanation. Consider the story of children and cars. In 1975, 1,632 pedestrians younger than 13 [were killed by motor vehicles](https://seriousaccidents.com/blog/children-traffic-fatalities-by-year/?itid=lk_inline_enhanced-template). You might expect that number to grow alongside the U.S. population, which has increased by more than 50 percent, from 216 million in 1975 to 334 million. Instead, and mercifully, the number is far, far smaller: 138 deaths (in 2019). A naive person looking at that statistic might well conclude that roads are now much safer for children. That’s the opposite of the case: The decline in deaths is because roads are now so dangerous that parents don’t allow their kids to play in the street.
Among the statistics I cite above, a number that really leaps out is the number of deaths from falls. In 2020, the most recent year for which we have the final numbers, there were [more deaths from falls](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3314071/?itid=lk_inline_enhanced-template#B9) than [from traffic accidents](https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813266?itid=lk_inline_enhanced-template#:~:text=Executive%20Summary-,There%20were%2038%2C824%20people%20killed%20in%20motor%20vehicle%20traffic%20crashes,in%20traffic%20crashes%20in%202020.). Deaths among seniors from falling have [more than doubled in 20 years](https://ssri.psu.edu/news/rate-fatal-falls-among-us-seniors-doubled-20-years?itid=lk_inline_enhanced-template). Almost all the deaths are among people older than 65, and the number rises even more steeply for Americans older than 85, who now die in falls at a rate that has almost tripled since 1999. [A 2012 public health report](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3314071/?itid=lk_inline_enhanced-template) found that the category of elder deaths from falls on the same level — i.e., not stairs or ladders — rose 698 percent between 1999 and 2007.
That is a bizarre statistic. Seven times more older Americans dying in falls? What’s going on? A number of possible reasons have been advanced. People live longer with serious conditions, and some of these accidents might be happening to survivors of serious cardiovascular or other emergencies who now live with an increased risk of falls. More people live on their own and have less help moving around. Perhaps doctors are medicating more and at the same time not taking enough trouble to consider drug interactions, leading to dizziness and fainting? Or perhaps doctors have just become worse at treating falls, and the case fatality rate has gone up? But why would that have happened? These hypotheses aren’t mutually exclusive: They could overlap, and all those things could be simultaneously true.
But there is a simpler possibility, one that needs to be borne in mind whenever a pattern in statistics looks anomalous: the possibility that a number has changed because it is being counted differently. The same 2012 paper that flagged an almost 700 percent increase in deaths from falls looked to see whether falls were leading to a concomitant increase in emergency room admissions. They weren’t. There was also no increase in fall-related hospitalizations. That means more people weren’t falling over. Instead, it looks as though the reason for the surge in recorded deaths is a change in recording practices.
Many deaths whose original inciting cause was a fall occur weeks or even months after the fall itself. To reflect this fact, when a new International Classification of Diseases framework [was introduced in 1999](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3314071/?itid=lk_inline_enhanced-template), it contained “a major change in death certification practices after implementation … followed by a gradual improvement in the quality of certification of belated deaths following falls.” In plain English, the reason the number of deaths from falls went up is because the government changed how they were counted. That’s something that always needs to be kept in mind with statistics: the question of how the numbers were made and the linked question of whether the numbers are telling the story they seem to be telling.
In all the torrent of data and statistics and numbers produced by the government, I would argue that two in particular stand out. Both are numbers that are made, not found; both involve an extraordinary amount of work, and both are, and always have been, highly controversial. The first is that very same census specified in the enumeration clause of the Constitution. The census is vitally important because any government needs to know how many citizens it has and where they live and who, broadly speaking, they are; in the case of the United States, the census also directly leads to the allocation of power in the House of Representatives. Throughout American history, there has never been a time when the census wasn’t the subject of controversy.
Employees of the graphics presentation section of the Labor Department work on consumer price index charts in October 1963. (Warren K. Leffler/Library of Congress Prints and Photographs Division)
The other number, equally important, is the main official measure of inflation, the consumer price index. At its simplest, the CPI is a vital number because it is used in multiple economic aspects of government: It is used to set levels of Social Security payments, access to payments in the Supplemental Nutrition Assistance Program (formerly food stamps), pensions and tax thresholds. It features in business contracts, in court orders and in divorce settlements. Many millions of workers have salary agreements that alter in direct relationship to the CPI. [As the Bureau of Labor Statistics says](https://www.bls.gov/opub/mlr/2014/article/the-first-hundred-years-of-the-consumer-price-index.htm?itid=lk_inline_enhanced-template#:~:text=Of%20all%20the%20economic%20statistics,of%20changes%20to%20the%20CPI.), “Of all the economic statistics produced by the U.S. federal government, none has a direct impact on the lives of everyday Americans quite like the Consumer Price Index.”
Like travel and infrastructure and domestic security, the CPI has the particular characteristics of a thing that no one notices or thinks about when it’s going well but that everybody becomes obsessed with when it’s going badly. The CPI sounds simple enough: It measures how much prices have gone up. [According to the official definition](https://www.bls.gov/cpi/questions-and-answers.htm?itid=lk_inline_enhanced-template#:~:text=The%20Consumer%20Price%20Index%20\(CPI,of%20consumer%20goods%20and%20services.), it is “a measure of the average change over time in the prices paid by consumers for a representative basket of consumer goods and services.” But there is a large basket of loudly squirming devils in the details.
The simplest way to imagine how hard it is to come up with the CPI is to ask how you would do it for yourself. You would work out all the stuff you buy, determine how much of a portion of your total budget was absorbed by each specific item and then find out the change in prices for each of them across time. The most straightforward step would probably appear to be a like-for-like comparison of item prices. Take, say, cheese. (I could choose the example [of olive oil](https://www.cnbc.com/2024/05/02/liquid-gold-a-shortage-of-olive-oil-has-fueled-a-record-price-spike.html?itid=lk_inline_enhanced-template#:~:text=%E2%80%9CThe%20world's%20biggest%20exporter%20of,to%20other%20food%20crops%20too.), but, over the past couple of years, that has become too painful to contemplate. If you know, you know.) How much has [the price of cheese](https://www.bls.gov/cpi/questions-and-answers.htm?itid=lk_inline_enhanced-template#:~:text=This%20initiation%20process%2C%20typically%20carried,where%20cheese%20will%20be%20priced.) changed over the past year? You would go through receipts from a year ago and from today, find the identical item, and do the math. But how would you find a number for, not your cheese, but all cheese? Across the entire United States? What’s the typical, representative cheese? Where would you buy it? How much would you buy?
The government agency with the responsibility of [calculating the CPI](https://www.bls.gov/cpi/questions-and-answers.htm?itid=lk_inline_enhanced-template) is the Bureau of Labor Statistics. This is how it works out the changing price of cheese:
“A particular item enters the CPI sample through a process called initiation. This initiation process, typically carried out in person by a CPI data collector, involves selecting a specific item to be priced from the category that has been designated to be priced at that store. For example, suppose a particular grocery store has an outlet where cheese will be priced. A particular type of cheese item will be chosen, with its likelihood of being selected roughly proportional to its popularity. If, for example, cheddar cheese in 8 oz. packages makes up 70 percent of the sales of cheese, and the same cheese in 6 oz. packages accounts for 10 percent of all cheese sales, and the same cheese in 12 oz. packages accounts for 20 percent of all cheese sales, then the 8 oz. package will be 7 times as likely to be chosen as the 6 oz. package. After probabilities are assigned, one type, brand, and container size of cheese is chosen by an objective selection process based on the theory of random sampling. The particular kind of cheese that is selected will continue to be priced each month in the same outlet.
“This item will be repriced, monthly or bimonthly, until it is replaced after four years through sample rotation. Repricing is usually done in person, but may be done via telephone or the internet. The process of selecting individual quotes results in the sample as a whole containing a wide variety of specific items of a category roughly corresponding to consumer purchases. So the cheese sample (or the new vehicle sample, the television sample, etc.) contains a wide variety of styles and brands of cheese, vehicles, televisions, etc.”
## Community corner
And that’s just cheese. Now scale the same process, with the same level of detail, complexity and wonkery [among all the items](https://www.bls.gov/cpi/additional-resources/entry-level-item-descriptions.htm?itid=lk_inline_enhanced-template) bought by a typical consumer — pausing for a moment to chuck a large concealing tarpaulin over all the questions prompted by the idea of a “typical consumer.” We’re talking about not dozens or hundreds but thousands of possible items — and not just in one place but all over the country. Not just canned tuna and breaded fish sticks, cornflakes and granola bars, but all types of bread and tortillas and rolls (including gluten free, obviously — it’s 2024). Cheesecakes and banana-nut breads and bacon — defined as “all types and forms (or cuts) of pork bacon, Canadian bacon and bacon substitutes such as turkey bacon, beef bacon, vegetarian bacon … slab bacon, sliced bacon, end pieces and jowl bacon … various types and forms of breakfast sausage such as, pork sausage, vegetarian based sausage, and other meats based sausage including a variety of meat combinations. Examples of meat combinations may include pork and turkey, pork and beef, etc. Forms of breakfast sausage may include loose, unlinked and linked in casings, and patty meat substitutes, formed links without casings, etc. The ingredients for breakfast sausage may include meat, poultry, cereal, soy protein, and other extenders” — and all types of pork and beef and chicken and organ meats and smoked salmon and eggs and ice cream (including nondairy, obviously; it’s still 2024) and lettuce and herbs. All types of fresh fruit, canned fruit, dried fruit, cocktail mixes, barbecue rubs and ketchups. All organ and wild meats, including liver, kidney, heart, brains, tripe, chitterlings and tongue, and, obviously, game. Examples of game tagged by the Bureau of Labor Statistics are buffalo, bison, venison, goat, rabbit, quail, rattlesnake, pheasant, grouse and quail.
But wait! The CPI, obviously, isn’t mainly about food because most household expenditure isn’t on food. In fact, food occupies [only 13.4 percent](https://www.bls.gov/news.release/cpi.t01.htm?itid=lk_inline_enhanced-template) of a typical household budget. To track a typical household’s expenditure, the BLS also monitors price changes on apparel and health and education and insurance and transport and recreation/entertainment. Within that last category, there are tents for camping, table tennis rackets, outboard motors, fish food, scuba equipment, dog grooming services, digital cameras, sewing machines, thread, needles, health club memberships, hunting knives, sheet music, every kind of recorded music, TV subscriptions (both basic and premium), dog collars, golf carts. And much, much more.
The list of tagged items is so extensive it is vertiginous. In the course of making these categories, the BLS has finally settled the question of whether professional wrestling is fixed: The category of admission to sporting events includes “football, baseball, basketball, hockey, boxing matches, horse races, and dog shows” at “all levels of competition, such as professional, collegiate, high school,” but it specifically excludes wrestling. “Flea markets, art shows, fashion shows, Wrestling” are instead in the category of “admission to movies, theaters, concerts, & other recurring events.” It is fun to imagine the meeting where that question was settled — and perhaps there is a tiny glimpse of backroom drama in the fact that Wrestling, apparently uniquely among these many thousands of CPI entry items, has a capital letter.
In the middle of all this colossal project of categorization and enumeration, the single biggest category by far is shelter, which is how the BLS defines what most of us would call housing. Shelter takes up 36.3 percent of the CPI, [a long way ahead of](https://www.bls.gov/news.release/cpi.t01.htm?itid=lk_inline_enhanced-template) food (8 percent at home, 5 percent elsewhere), energy (7 percent), transportation (6.5 percent) and medical services (6.5 percent). The number for shelter includes all rentals, from people living in trailers in West Virginia to oil workers in company housing in Anchorage to crypto bros renting Miami condos. It covers homeowners, too. In economics, there are many things that are counterintuitive, and one of them is the idea that the value of your house, in income terms, is the rent it is saving you. The “shelter” cost, for CPI purposes, is the number you would be paying for your property if you rented it. This is called “[owners’ equivalent rent](https://www.bls.gov/cpi/factsheets/owners-equivalent-rent-and-rent.htm?itid=lk_inline_enhanced-template),” and it means that even if your housing costs haven’t in fact gone up because you own the place where you live, your shelter costs, as measured by the CPI, will have increased.
There is something intellectually thrilling about this: millions of data points, from tens of thousands of sources, being recorded, categorized, quantified, analyzed and weighted, through the labor of thousands of people, and all of it to produce one single, apparently simple and self-explanatory number. It is the principle of [e pluribus unum](https://www.si.edu/object/e-pluribus-unum-gazetteer-united-states-america%3Asilgoi_68301?itid=lk_inline_enhanced-template), applied to data. All that work ends with a single number to represent all inflation, [the CPI-U,](https://www.bls.gov/news.release/cpi.t01.htm?itid=lk_inline_enhanced-template) which, at the time of writing, [stands at 2.9 percent](https://www.bls.gov/news.release/pdf/cpi.pdf?itid=lk_inline_enhanced-template). (U stands for “all urban consumers” — [about 93 percent of the U.S. population](https://www.bls.gov/cpi/overview.htm?itid=lk_inline_enhanced-template).)
Nothing about this is self-evident, though much of it, when you look underneath the hood, is the product of a rarefied form of common sense. The CPI in its modern form is the result of a continuing series of debates and arguments in the area where economics and politics overlap. The first attempt at producing a single number for inflation [began in 1921](https://www.bls.gov/opub/hom/cpi/history.htm?itid=lk_inline_enhanced-template#:~:text=Periodic%20collection%20of%20prices%20was,were%20estimated%20back%20to%201913.), using data that had begun to be collected in 1913. The data for this “cost of living index,” as it was called, was collected from a survey of White wage-earner families in 92 cities. The collection of goods used to measure inflation is known as a basket, and that first basket contained items that seem less essential today: a straw boater, for example. The category of beef cuts is wonderfully specific, and there’s [a helpful diagram of a cow](https://www.bls.gov/opub/mlr/2014/article/one-hundred-years-of-price-change-the-consumer-price-index-and-the-american-inflation-experience.htm?itid=lk_inline_enhanced-template) to assist the person compiling the data.
The inflation basket has changed over time, and so has awareness of the different rates of inflation that apply to different citizens. The older index for urban wage earners was in 1978 renamed the CPI-W, and the newer index for all urban consumers — today the standard measure of inflation — became the CPI-U. And then there’s the reality of substitution, as economists call it: the fact that as prices change, our behavior changes, too. The CPI can go up so much that it forces your spending to go down. As beef becomes more expensive, we switch to pork or chicken; if you can’t afford prime cuts to cook a steak, you use cheaper cuts to make a casserole. The BLS acknowledges this through an index that attempts to track substitution: the “chained CPI” or C-CPI-U. It was introduced in 2002. There is also a separate index for older Americans, CPI-E, introduced in 2008 after being mandated by Congress. This happened in response to political concerns that older people have different needs and spending patterns not reflected in the ordinary CPI-U. A cynic would point to older folks’ tendency to turn out and vote. All these emendations reflect the fact that inflation indexes are things that are made, created through intellectual and practical work, and are prone to give different answers when different questions are asked. It has never not been argued over, and it is often the case that people like the CPI when it’s telling them something they want to hear and level furious accusations against it when it’s saying something inconvenient.
The end product of this is a paradoxical number. For one thing, it is possible that it doesn’t often correspond to reality. As the BLS itself points out, because people’s lives are so different, [it seldom mirrors](https://www.bls.gov/cpi/questions-and-answers.htm?itid=lk_inline_enhanced-template#:~:text=The%20CPI%20is%20often%20used,to%20millions%20of%20American%20workers.) a particular consumer’s experience. In particular, the poor tend to suffer higher inflation than the rich. Better-off people have assets, which, broadly speaking, rise in value as inflation climbs. Poorer people don’t, and they spend a bigger proportion of their income on those basics of life that are particularly exposed to surges in inflation: food and fuel. These are part of what is called “noncore inflation,” a strange term that tries to separate out from the rest of the economy the part of inflation that is chronically affected by fluctuating prices. But if you’re poor, there’s nothing noncore about the cost of the food on your plate or the gasoline in your tank: They are central to your experience of living day to day. Noncore is as core as it gets.
Because of all this, inflation is always and everywhere a political phenomenon. For tens of millions of people, it has a direct impact on how they live. In the 1960s, [economist Arthur Okun](https://www.brookings.edu/wp-content/uploads/1980/01/1980a_bpea_gordon_hall.pdf?itid=lk_inline_enhanced-template), the chairman of President Lyndon B. Johnson’s Council of Economic Advisers, coined something he called the [economic discomfort index](https://www.nytimes.com/2022/09/13/opinion/inflation-misery-index-economy-prices.html?itid=lk_inline_enhanced-template), a name shortened and improved by Ronald Reagan to “misery index.” (Economics, “the dismal science,” is full of gloomy-sounding names for gloomy realities. I like the way the term “misery index” gives up any attempt to combat this tendency and just goes fully grim.) The misery index is the unemployment rate plus inflation. At the time Okun devised it, [the index stood at 8.12](http://www.miseryindex.us/indexbyPresident.aspx?itid=lk_inline_enhanced-template). In the intervening years, it has fluctuated between a high of 21.98 under Jimmy Carter in June 1980 and a low of 5.06 under Barack Obama in September 2015.
One fact stands out: When the misery index has been in double figures, during an election year, the incumbent president has lost. This was the case for Carter in 1980; for George H.W. Bush in 1992, when it hit 10.89[;](https://www.miseryindex.us/indexbymonth.aspx;?itid=lk_inline_enhanced-template) and for Donald Trump in 2020, when the pandemic’s impact took the index from 5.21 in September 2019 to 15.03 seven months later. The only president to have survived a double-digit misery index in an election year was Reagan in 1984, but he seems to have been protected by the fact that the number was declining all through the campaign, from 19.93 at his inauguration in 1981 to 11.25 by the time of his reelection in 1984.
If the misery index had retained its predictive power, there wouldn’t be much to discuss about the current presidential campaign. Both unemployment and inflation have trended downward during the Joe Biden-Kamala Harris years. The current number is 7.2, comprising 4.3 percent unemployment and 2.9 percent inflation. This number, good by historic standards, is the product both of policy and, to be frank, of luck, thanks to the ebbing of the upward pressure on prices created by the pandemic and Russia’s invasion of Ukraine. In a normal election cycle, that below-10 misery index would be a reliable indication that the incumbent could switch on cruise control and start making plans for a second term.
Close observers of U.S. politics might have noticed that — to put the situation very mildly — it isn’t working out like that. I argued earlier that some numbers need to be explained and contextualized to make their real meaning clear. You need to extract the meaning to tell the story. Inflation in general, and the misery index in particular, isn’t supposed to be like that. It’s supposed to tell you how people feel without further explanation. But for some reason, the good news about the economy isn’t translating into good vibes in the polls. Wags have taken to [calling this phenomenon](https://www.newyorker.com/news/the-financial-page/has-kamala-harris-risen-above-the-vibecession?itid=lk_inline_enhanced-template) a “vibecession.”
The question is: What’s happening? Is this something to do with the data or with the electorate or both?
I think it’s difficult to claim there is something fundamentally wrong with the CPI. Or rather, that there is something fundamentally wrong with it that has only just come to light. Since its creation, the index has been debated, contested and labored over. Before there was even a single number for the index, when the BLS was making its first calculations about wages and the cost of living, it was coming under attack for bias. [From an official history](https://www.bls.gov/opub/mlr/2014/article/the-first-hundred-years-of-the-consumer-price-index.htm?itid=lk_inline_enhanced-template#:~:text=Many%20of%20the%20same%20labor,or%20giving%20way%20to%20political): “Many of the same labor organizations that praised the Bureau for its work in providing relevant data during labor disputes — data which often led to favorable decisions for laborers, such as the decision in the Anthracite coal strike — now accused the Bureau of employing faulty methodology or giving way to political pressure.” [They’re talking about 1904](https://www.bls.gov/opub/mlr/2014/article/the-first-hundred-years-of-the-consumer-price-index.htm?itid=lk_inline_enhanced-template#)! Even the creation of the index in the first place was the result of extensive arguments around whether and why a cost-of-living index was desperately needed. All statistics are to some extent man-made, and the CPI is as man-made as it gets. But that has always been true, and the work that goes into production of this one economic statistic to rule them all is as transparent as it could possibly be. Look at the BLS’s bewilderingly complete website on the CPI — for the nerdy and data-minded, it is a thing of gloriously wonky beauty.
One big thing about inflation that has changed, however, is so simple that it is easy to overlook. It’s the fact that we were once used to inflation, and now we aren’t. In 1964, inflation was a slain dragon: [It stood at 1 percent](https://www.investopedia.com/inflation-rate-by-year-7253832?itid=lk_inline_enhanced-template). Ten years later, it was over 12 percent and had become an economic phenomenon nobody could ignore. When Reagan was first elected in 1980, inflation was running at 12.6 percent. By 1984, two decades of inflation meant that a 1964 dollar was worth 30 cents. The U.S. dollar had lost more than 70 percent of its value in 20 years.
Because we’ve forgotten this, we’ve forgotten the way in which inflation makes societies a little crazy. Most people are familiar, at least in broad terms, with the stories of countries such as [Zimbabwe](https://www.washingtonpost.com/news/worldviews/wp/2016/05/05/zimbabwe-is-planning-to-print-its-own-u-s-dollars/?itid=lk_inline_enhanced-template) or [Weimar Germany](https://www.washingtonpost.com/news/wonk/wp/2012/08/21/great-hyperinflation-episodes-in-history-and-what-they-tell-us-about-the-fed/?itid=lk_inline_enhanced-template), in which inflation got so bad that it broke daily life: People needed bags full of cash to buy a loaf of bread, and customers literally ran around stores trying to keep ahead of store clerks changing prices. We think of inflation-related derangement as something that happens to countries such as Argentina, where President Javier Milei’s supporters were in June celebrating his achievement in bringing annual inflation down to a mere [276 percent](https://buenosairesherald.com/economics/argentinas-yearly-and-monthly-inflation-rates-down-to-276-and-4-2?itid=lk_inline_enhanced-template). When inflation gets out of hand, your country turns into Argentina, and you get a head of state with cloned dogs and mutton-chop whiskers [waving a chain saw](https://apnews.com/article/milei-argentina-chainsaw-fed35a37c6137b951e4adada3d866436?itid=lk_inline_enhanced-template) over his head — this is what we know to be true.
But inflation has at times made U.S. politics pretty crazy, too. Inflation led President Richard M. Nixon to [introduce wage and price controls](https://www.federalreservehistory.org/essays/gold-convertibility-ends?itid=lk_inline_enhanced-template) in 1971. This is how out of kilter inflation can make politics become: that it can bring a Republican president to implement policies more common in North Korea. [That policy failed](https://www.pbs.org/wgbh/commandingheights/shared/minitext/ess_nixongold.html?itid=lk_inline_enhanced-template), as it always does, because the controls caused shortages that then caused further surges in prices. After Nixon was forced out of office, Gerald Ford brought in the [Whip Inflation Now](https://www.fordlibrarymuseum.gov/whip-inflation-now-win?itid=lk_inline_enhanced-template#main) program, urging citizens in the direction of voluntary thrift. Americans were encouraged to wear WIN badges; to carpool, plant vegetables and turn down their heating. “It was a failure,” the Federal Reserve’s [history of the period](https://www.federalreservehistory.org/essays/great-inflation?itid=lk_inline_enhanced-template) says crisply, while heroically refraining from the joke that LOSE would have been a more appropriate name.
An artist works on a consumer price index chart for the Bureau of Labor Statistics in September 1971. (Thomas J. O’Halloran/Library of Congress Prints and Photographs Division)
If it’s true that “the Great Inflation was the defining macroeconomic event of the second half of the twentieth century,” as the Federal Reserve argues, then it’s also true that this entire episode has been largely forgotten. Many younger than 50 spent much of their adulthood living through the “[Great Moderation](https://www.federalreservehistory.org/essays/great-moderation?itid=lk_inline_enhanced-template),” as it was hubristically dubbed. Boom and bust were a thing of the past, and inflation just wasn’t something anyone had to think about. That has changed dramatically since 2020. To take just [three examples](https://www.washingtonpost.com/business/2023/01/10/price-swings-inflation/?itid=lk_inline_enhanced-template): From November 2021 to November 2022, [butter prices](https://www.cnbc.com/2023/01/13/why-inflation-hit-these-10-items-hardest-in-2022.html?itid=lk_inline_enhanced-template) went up 41.4 percent, fuel oil by 65.7 percent and school lunches by 254 percent. For a family without much money, that was [an emergency](https://www.weforum.org/agenda/2023/01/us-inflation-prices-increased-economy/?itid=lk_inline_enhanced-template).
Nothing like this has happened in a long time, and we have quite simply forgotten what inflation feels like. In addition, when the BLS says, correctly, that the rate of inflation is falling, what people sometimes hear is that prices are falling — which they aren’t. The rate of increase of the price is falling, but the number on the sticker is continuing to go up. This is a “well duh” thought to anyone who is used to thinking about economics, but it is a more complex idea than economists realize. When people are told that inflation is falling, but they know from direct experience that prices aren’t going down, they are sometimes prone to think they are being lied to.
It doesn’t help that there are loud voices yelling at them that, yes, they are indeed being lied to. In every society with significant inflation, that inflation is used as a political football. Today, the American right has latched onto it as a political weapon. The first specific critique in the Heritage Foundation’s [Project 2025 playbook](https://www.project2025.org/?itid=lk_inline_enhanced-template) for dismantling the federal government takes a firm hold of the topic: “Look at America under the ruling and cultural elite today: Inflation is ravaging family budgets …” Inflation is repeatedly harped on in Project 2025. It says, “Congress should require that the Consumer Price Index market basket include measurable family-essential goods.” What that means isn’t clear, but it is true that families with low incomes — because so much of their expenditure is directed toward food and energy — are particularly exposed to inflation. What isn’t fair — and is in fact the opposite of the truth — is the implication that the BLS isn’t interested in the real impact of inflation. Just look at some of the voices on X: “[They’ve been lying about inflation](https://x.com/xxclusionary/status/1820732960380776506?itid=lk_inline_enhanced-template). The CPI is categorically bs.” “[The CPI basket](https://x.com/MulderSonny/status/1825952614703575415?itid=lk_inline_enhanced-template) is also heavily influenced and fake data.” “[Government produced CPI](https://x.com/NyaPfanner/status/1825993061219909976?itid=lk_inline_enhanced-template) or inflation rate is just propaganda.” “The CPI number is a fake number to fool Americans into thinking things are not as bad as they feel.” “The CPI basket is a total hoax.” “CPI is a lie.” “[Govt CPI is fake data](https://x.com/WallStreetSilv/status/1669190729095102467?itid=lk_inline_enhanced-template). They massage the data so it is lower so that cost of living adjustments for social security are reduced over time.” “The CPI is totally faked.”
The quality of debate isn’t improved by the fact that the Biden administration’s signature economic policy is called the Inflation Reduction Act, when it is [nothing of the kind](https://www.washingtonpost.com/us-policy/2022/08/07/inflation-reduction-act-false/?itid=lk_inline_enhanced-template). (The Congressional Budget Office estimated the act’s effect on inflation [would be statistically “negligible.”](https://apnews.com/article/inflation-biden-health-congress-climate-and-environment-63df07e15002c01fb560a6f0e69fcb03?itid=lk_inline_enhanced-template)) The Harris campaign understands the salience of the issue and has signaled support for a federal crackdown on price-gouging at grocery stores. This might be good politics but only marginally relevant economics, given that [37 states](https://www.nbcnews.com/politics/2024-election/trump-slams-harris-call-ban-price-gouging-37-states-already-rcna167158?itid=lk_inline_enhanced-template) already have laws against price-gouging and that inflation is low by historic standards — and, in any case, has very little to do with price-gouging. But those facts are a difficult sell. Voters don’t like to be told things aren’t as bad as they think.
According to a poll in May, a majority of Americans think that the United States [is in a recession](https://www.theguardian.com/us-news/article/2024/may/22/poll-economy-recession-biden?itid=lk_inline_enhanced-template). It isn’t: The U.S. economy has been performing remarkably well by international standards and is a long way ahead of those of the country’s peers in the Group of Seven. The same poll found that 72 percent of Americans think inflation is rising. It isn’t. The poll also found that nearly half of respondents thought the stock market was down for the year and that unemployment was at a 50-year high. The realities are that the S&P 500 is up for the year and that unemployment is at [a 50-year low](https://www.commerce.gov/news/blog/2023/02/news-unemployment-its-lowest-level-54-years?itid=lk_inline_enhanced-template#:~:text=).
What seems to be happening here is that a subjective sense that times are hard is occluding the realities. Nerds have a saying, “The plural of anecdote is not data.” In other words, your story is not a general truth. That’s fine, but it cuts both ways: On an individual level, data loses to story. Every time. Your statistics don’t trump my experience. The BLS numbers show an economy in pretty good shape. There are loud voices telling Americans how bad everything is, and not many willing to take the risk of saying the opposite. Technology increasingly makes it the case that people can choose to live inside a sealed chamber of selected opinion, curating their feeds and selecting their narratives; if they want to hear that times are hard, there is no shortage of forces eager to confirm the belief.
A big factor here, I think, is that the CPI fails to recognize the disproportionate impact of food inflation. That effect is objective for the poor, who spend more of their income on food; but it is subjectively high for everyone. It might be 10 years since you last bought a hunting knife or yoga mat, so you will notice a rise in their costs once a decade. If you pay for home or car insurance once a year, you will notice a rise annually. But food inflation is the cost you can encounter not just daily but multiple times a day, and so it can really get into your head. [An Ipsos poll in mid-August](https://www.ipsos.com/en-us/americans-increasingly-concerned-about-inflation-us-election-nears?itid=lk_inline_enhanced-template) found that 50 percent of Americans think the cost of living is a top concern for the country, even as inflation has fallen precipitately — from 9.1 percent in June 2022 to 2.9 percent in July 2024. The culprit for the disconnect is, surely, to be found in a fact cited in the poll: Food at home is up 28 percent over the past five years. As a category, food at home might make up only 8 percent of the CPI — but it is the only category that, on a daily basis, punches you in the face.
Personal truth is everything. That’s how we live. But what are the consequences of that for the Enlightenment project of statistics — the truths that we make by categorizing and counting and generalizing? The CPI is an impossible number. To try to assess every piece of economic activity in a country of nearly [340 million people](https://www.census.gov/popclock/?itid=lk_inline_enhanced-template), find a typical price for it, measure how that price changes, repeat for every urban area in the country, weight it according to how much of a typical household’s expenditure it represents, monitor how much is spent on each item or category as lifestyles change (out with the straw boater, in with the Netflix subscription), put all these millions of data points through an algorithm and come up with a single number summing up what is happening to everyone everywhere — it’s nuts, it can’t be done. Especially when you bear in mind that, as the BLS admits, the precise number “seldom” applies. And yet, it has to be done because otherwise your society is, in economic terms, flying blind. Just as a simple practical point, without the CPI, all those things which need to be adjusted for inflation — half of all federal spending, by some measures — can’t be.
### In conversation
#### Michael Lewis talks with the authors behind this series
Casey Cep on the National Cemetery Administration
Dave Eggers on NASA’s Jet Propulsion Lab
John Lanchester on the Bureau of Labor Statistics
[As the BLS itself says](https://www.bls.gov/cpi/questions-and-answers.htm?itid=lk_inline_enhanced-template#:~:text=This%20initiation%20process%2C%20typically%20carried,where%20cheese%20will%20be%20priced.), “The CPI does not necessarily measure your own experience with price change.” It captures a broader general truth. To produce it, you have to accept that what you’re doing is in some sense impossible; but at the same time, you accept that it has to be done, and it has to be done as well as it possibly can be. You have to try to achieve objectivity while knowing that it can’t be attained. Maybe this double-mindedness is something we have lost or are on the verge of losing. Carl Sagan, [writing in 1995](https://www.yahoo.com/news/carl-sagan-prediction-1995-surprisingly-160000330.html?itid=lk_inline_enhanced-template), had a dark vision of a future that looks a little like this.
“I have a foreboding of an America in my children’s or grandchildren’s time — when the United States is a service and information economy; when nearly all the key manufacturing industries have slipped away to other countries; when awesome technological powers are in the hands of a very few, and no one representing the public interest can even grasp the issues; when the people have lost the ability to set their own agendas or knowledgeably question those in authority; when, clutching our crystals and nervously consulting our horoscopes, our critical faculties in decline, unable to distinguish between what feels good and what’s true, we slide, almost without noticing, back into superstition and darkness.
“The dumbing down of America is most evident in the slow decay of substantive content in the enormously influential media, the 30-second sound bites (now down to 10 seconds or less), the lowest common denominator programming, credulous presentations on pseudoscience and superstition, but especially a kind of celebration of ignorance.”
The idea that the CPI doesn’t apply to everyone — fine. The idea that the CPI might need adjusting — also fine, and adjusting it to meet reality is a large part of the work that goes into producing it. But the idea that the CPI is a deliberate falsehood is a lie, and a dangerous one. It is a sign that Sagan’s dark prophecy is starting to be true. And if it is, we are beginning to look at a world that has moved past the Enlightenment. We can admit that the Enlightenment was a flawed project, whose ideas about objective truth and knowledge contained all sorts of coded values about hierarchy and power and race. But Enlightenment values are implicit in the kind of knowledge-making that is embodied in the CPI — in the idea of a simple-looking number that tries to sum up a complicated, messy reality. Get rid of those values, and we have to start thinking about what comes next.
The thing that comes next as yet has no name. The term “[Counter-Enlightenment](https://www.cambridge.org/core/journals/american-political-science-review/article/abs/origin-of-the-counterenlightenment-rousseau-and-the-new-religion-of-sincerity/924CE2FBFEB43A31F90B81C363868D61?itid=lk_inline_enhanced-template)” has already been used up by the religious backlash to the Enlightenment. The Counter-Enlightenment had a different idea of truth, grounded in what it saw as older, more permanent verities. The new enemy of the Enlightenment doesn’t believe in any kind of verities beyond personal feeling and experience. It has no use for statistics or numbers or data. It doesn’t hold any truths to be self-evident. It would never even bother to conceive of the CPI, let alone dedicate millions of hours of labor to create it. So, yes, the CPI is flawed. But it is based on the idea that we can work to make a number that tries to tell a society-wide general truth. Get rid of that idea, and we would be abandoning the Enlightenment and the ambitions for progress that came with it. We would enter a new philosophical anti-system. Call it the Darkening. Let’s not give in to it without a fight.
