We often hear that the American dream is no longer achievable for a large fraction of Americans. Some of the people who make that claim go on to advocate more government regulation and spending to help restore the dream.
But what if the American dream is alive and well, and what if current government intervention is making it less well than it could be? In his new book, Crushing Capitalism: How Populist Policies Are Threatening the American Dream, Cato Institute economist Norbert J. Michel raises those questions. To answer them, he lays out massive amounts of data that show things are getting better for most Americans, many government interventions slow that improvement, and further government intervention would slow it further. Michel makes his case by thoroughly examining data on wages, household incomes, and poverty, and along the way notes the problems with welfare benefits, minimum wages, and tariffs.
Manufacturing vs. the service sector
One of the most frequently stated claims from observers who want the United States to have more manufacturing jobs is that such jobs pay more than service-sector jobs. I admit that even I believed it. Michel shows that in 2000 the average wages of nonsupervisory workers in manufacturing were 5.1 percent higher than the average wages of nonsupervisory workers in the service sector. That alone surprised me because I had thought the differential was much higher. But he also shows that by 2017, the ranking had reversed: average wages in manufacturing were 2.5 percent below those in the service sector.
A factor that biases the comparison is the relative ages of manufacturing and service-sector workers. Economists have long known that a given worker’s wages grow as he or she goes from youth to middle age. Michel points out that 15 percent of workers in the leisure and hospitality industries and 7 percent of workers in retail are under age twenty while only 1 percent of workers in manufacturing are under age twenty. That means that wages in the service sector relative to the manufacturing sector are even better than they look.
What if you ignored those facts and tried to use government policy to create lots of manufacturing jobs? Michel argues persuasively that you couldn’t. Although manufacturing jobs as a percent of all jobs fell by about half between 1987 and 2022, that was mainly due to increases in productivity. One striking fact he reports is about the steel industry. In 1980, producing one ton of steel required 12 workers; by 2018, that was down to 0.6 workers. Michel puts the point nicely, writing, “Manufacturing now relies so heavily on automation for this higher productivity that, unless machines are strictly outlawed, no amount of industrial policy can restore the number of US manufacturing jobs to its pre-1980s peak.”
Have real wages stagnated?
We often hear that there has been almost no growth in real wages in the past few decades. For example, in a 2018 book, Oren Cass, a lawyer who is chief economist at American Compass, stated that between 1975 and 2015, “the median worker’s wages have barely budged.”
Michel notes several problems with Cass’s claim. I’ll highlight two. First, to adjust wages for inflation so that he can compare real wages over time, Cass uses the Consumer Price Index (CPI), which notoriously overstates inflation. A better index, which also overstates inflation but less so, is the Personal Consumption Expenditure (PCE) index. Using this index, Michel shows that between 1975 and 2015, real wages grew by 22 percent, compared to the 1 percent that Cass computed using the CPI. Second, an important component of wages is employer-provided benefits. Between 1973 and 2019, which was the approximate time period that Cass discussed, these nonwage benefits grew from 13 percent of total compensation to 30 percent. In short, real wages, properly measured, have grown by a large percent since 1973.
The China shock
In a now-famous 2013 article in the American Economic Review, MIT economist David Autor and his co-authors, David Dorn of the University of Zurich and Gordon Hanson of UC-San Diego, presented evidence on what is now called the “China shock.” They showed that areas of the United States into which imports from China had most penetrated, after trade with China was opened in 2000, lost jobs. More important, they showed, many of the workers who lost those jobs failed to find other work quickly.
But other economists, including Nicholas Bloom of Stanford and his three co-authors, found offsetting positive effects of the China shock. In areas of the country that started with “high initial levels of human capital,” the China shock led to increases in non-manufacturing jobs. A dynamic economy, which ours is, will tend to have increases in jobs in one area and decreases in another.
Some pundits have suggested that the federal government subsidize those who lose jobs due to increases in trade. Autor et al. noted that we have such a program. It’s called Trade Adjustment Assistance (TAA). According to Autor et al., TAA, Social Security Disability Insurance, and Supplemental Security Income all rose substantially in areas with high exposure to imports. Disappointingly, Michel does not come out and point to a reasonable hypothesis, namely that these increases in transfer payments caused many workers to stay where they had lost jobs instead of moving to new jobs in other areas.
Although many of the facts Michel reports were relatively well known to economists who followed the “China shock” discussion, he does present one striking fact that I had not known. We often think that Chinese imports displaced American production. That’s true to some extent. But the main production that Chinese imports displaced was other imports. Pacific Rim countries, including China, he notes, “accounted for 47.1 percent of all US manufactured imports in 1990.” What changed was China’s share of Pacific Rim imports. In 1990, it was only 7.6 percent, but by 2017, it was up to 55.4 percent.
Household incomes
Another claim we often hear is that household incomes have stagnated. For instance, Brookings Institution economists Isabel Sawhill and Eleanor Krause stated in 2018 that “American households in the middle of the distribution have experienced very little income growth in recent decades.” But Michel points out that between 1967 and 2015, median real household income rose from $44,895 to $57,230. That 27 percent increase is not huge, but it’s better than “very little.” Moreover, Michel points out two “not-little” problems with the Sawhill/Krause data. First, they didn’t adjust for household size, which has declined substantially. Per person, household income rose over that time by 64 percent. And, as with Cass’s comparison of wages, Sawhill and Krause adjusted for inflation by using the CPI. Using the PCE, Michel concludes that between 1967 and 2015, real household income per person rose by 140 percent. That’s a lot.
What about a related claim that the middle class is disappearing? Michel shows that’s true but in a good way. According to a 2018 Census study, the share of households with incomes between $35,000 and $100,000 (in 2017 dollars) fell from over 53 percent in 1967 to 42 percent in 2018. There’s that disappearing middle class. But wait, there’s more. That same study, Michel notes, quoting a famous post by economist Mark J. Perry, shows that high-income families, those with incomes of $100,000 or more in 2018 dollars, rose from 9.0 percent of all families in 1967 to a whopping 30.4 percent in 2018. Moreover, the percent of households with incomes of $35,000 or less, in 2018 dollars, fell from 37.2 percent to 27.9 percent. Like the Jeffersons of 1975–85 TV fame, we’re “movin’ on up.”
Would a higher minimum wage help or hinder?
The federal minimum wage has been stuck at $7.25 per hour since July 2009. Meanwhile, between July 2009 and now, the CPI has increased by 49 percent and the PCE index has increased by 42 percent. Because a growing economy tends to raise real wages, fewer and fewer people are employed at the federal minimum. Michel points out that only about 1 percent of the labor force works at or below this minimum. (How could it be below? There are exceptions in the federal law for youths and for full-time students in some industries.) The federal minimum hardly matters.
Naturally, the low federal minimum has led many people to advocate higher minimum wages. Michel points out the obvious problem, one that economists have been aware of for over eighty years. The higher the minimum, the higher is the probability that it will cause some workers, particularly young unskilled workers, to lose their jobs or have their work hours reduced.
What about the argument that you can’t raise a child while being a single parent making the minimum? Based on 2016 data, Michel notes, out of 163 million people in the US labor force, only 222,000 were single parents supporting children while earning the minimum wage or less.
Michel refers to “the good intentions behind” the effort to raise the minimum wage. But on the very same page, he points out that unions “have historically supported minimum-wage laws” because such laws “make higher-skilled and more experienced workers relatively more attractive to employers.” That doesn’t sound to me like good intentions. Some supporters of raising the minimum wage, especially those who don’t understand basic economics, have good intentions. Other supporters have more sinister motives.
Poverty is falling
Michel notes that the US poverty rate, officially defined, was 12.3 percent in 2017, the same as in 2006. He states that there has been no clear trend since the 1960s. Here would have been the place for him to cite The Myth of American Inequality: How Government Biases Policy Debate, by Phil Gramm, Robert Ekelund, and John Early. As I noted in my review of that book in November 2022, Gramm et al. point out that in computing the poverty rate, federal statisticians leave out huge parts of the welfare state. The three authors show that including those parts implies that the poverty rate is in the low single digits.
To his credit, though, Michel gets at the same point by a different route: by looking at what various households are consuming. The consumption data show that the poverty rate declined from 30 percent in 1960 to 6.2 percent in 2000, and to 2.8 percent in 2017.
Although even low-income people are doing better, one way not to help them is to impose high tariffs. Even before President Donald Trump’s first term in office, Michel notes, “tariffs on imported clothing were nine times as high as the average tariff for all imported goods.” Tariffs on many food items are also high. These high tariffs are regressive. In 2019, he notes, households in the lowest income quintile spent 36 percent of their after-tax income on food and 7 percent on clothing, while households in the highest quintile spent 8 percent of their income on food and only 2 percent on clothing.
Immigration
The economist most associated with the view that immigration drives down the wages of low-skilled workers is Harvard economist George Borjas, who himself emigrated from Cuba. Even Borjas, though, finds the effect to be small, on the order of 3 percent. Michel quotes this finding but goes on to quote a study by economists at the International Food Policy Research Institute, based on more recent data than Borjas’s, that finds the effect to be about 1 percent. In return for our immigration, we get major benefits. One is that other workers, especially women, join the labor force because they can hire nannies.
Many people argue that our immigration policies should discriminate in favor of people with high education levels. That view is superficially plausible. But Michel points out that many of our most productive people came here with little education or are the children of immigrants with little education. Jeff Bezos’s adoptive father, for example, emigrated from Cuba when he was sixteen and couldn’t speak English. Without Bezos, we might not have Amazon. Unfortunately, Michel doesn’t quote a finding by Yale economist and Nobel laureate Willam D. Nordhaus that innovators capture only about 2.2 percent of the social surplus they create; the rest goes to consumers. If that estimate applies to Bezos, then we consumers gained over $4 trillion because Bezos’s father was allowed in. And that’s just one immigrant.
Conclusion
The American dream is alive and well. Almost everyone is becoming better off. Michel is right to argue that policies based on a misconception of the state of the US economy are dangerous and could slow the growth in economic well-being.
