Great data, disappointing economics.

Henderson, David R.

The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War

By Robert J. Gordon

762 pp.; Princeton

University Press, 2016

[ILLUSTRATION OMITTED]

Northwestern University economist Robert J. Gordon's latest book is on U.S. economic growth from 1870 to 2014. With his careful sifting of the data, he shows how far we've come over that time span, and especially between 1870 and 1970. Examining the economy sector by sector, he finds progress in every area: life expectancy, food, housing, transportation, workplace The book's strengths are its data and conditions, clothing, entertainment, and Gordon's common-sense interpretation health care, to name some of the most important areas analyzed. In almost all cases, his data are impeccable. For those who fondly wish to return to the "good old days," he has important news: the "good old days," compared to now, were really bad.

The book's strengths are its data and Gordon's common-sense interpretation of that data. Unfortunately, he does not do nearly as good a job of using basic economics to explain the causes of economic growth and the ways to offset the lower future growth that he anticipates. At times, he shows an awareness of the harm done by many government regulations: policies on drug regulation, housing, entry into occupations, and airline security, to name four. However, he actually claims positive effects of a now-defunct regulatory agency, the Interstate Commerce Commission, that economists have definitively shown was almost wholly destructive. Also, he never explains why we've had the kind of innovation and capital investment that has driven our standard of living. When he discusses tax policy, he seems unaware that one of his proposals would likely reduce that investment and innovation.

Getting better / First the good news, and there's a lot of it. We are massively better off than people were in 1870. I know that's not earth-shattering news, but the detail that Gordon gives makes this case overwhelmingly.

Consider life expectancy. In 1870, life expectancy at birth for U.S. white males was 45.2 years. By 1940, it had risen to 64.2 years, and by 2010 to 77.9 years. The main reason for this increase was a huge improvement in health, and major contributors to good health were better food, better working conditions, sewers, running water, vaccinations, penicillin, and the replacement of horse-drawn public transit with motorized vehicles.

Between 1870 and 1940, the kinds of food people ate and the way they bought it changed dramatically. In the 1920s, there was a huge shift from local merchants to chain stores such as A&P. Even in chain stores, though, "customers lined up and waited while clerks fetched items from shelves behind them." Later, of course, there was a large saving in labor as people started picking their own items off consumer-accessible shelves. Disappointingly for a fact-filled book, Gordon does not discuss that transition. He does, however, lay out the tremendous increase in food variety resulting from food processing, canning, and refrigeration, all of which happened in that 70-year period.

Housing also improved dramatically, mainly because most houses became "networked." That is, between 1870 and 1940, most houses obtained access to clean running water, sewage, electricity, and telephone lines. All of these were enormously beneficial: running water saved carrying pails of water from an outside pump (which my mother did in our small town in Manitoba until 1958); sewage allowed for much safer disposal of human waste; electricity led to laborsaving devices in the home, plus lighting, which made evenings much more enjoyable; and telephones dramatically increased people's access to the bigger world.

A bumper sticker I used to see in the early 1970s in Los Angeles was "Cut Pollution: Ride a Horse." My impression, talking to fellow Angelenos, was that they saw no irony in this. People who think that horses were non-polluting would do well to read Gordon's book. He writes, "The horse was not only inefficient, eating up one-quarter of the nation's grain output, but also a source of urban pollution, disease, and occupational misery for the workers unlucky enough to have jobs in horse waste removal." And later: "Horses dropped thousands of tons of manure and gallons of urine on city streets; died in service, leaving 7,000 horse carcasses to be carried away in Chicago alone; and carried diseases transmissible to humans." On New York's Liberty Street, at one point, a manure heap measured seven feet high.

The advent of electric streetcars and then the internal combustion engine changed all that, making transportation easier, cheaper, and healthier, and extending people's range dramatically. One of the best parts of Gordon's book is his description of technological improvements in cars between 1906 and 1940, along with large reductions in price. I was pleased that he didn't bring the modern intellectual's disdain for cars to this book; Gordon well recognizes their huge value.

Entertainment also became incredibly more varied, easier to access, and cheaper. The three main improvements between 1870 and 1940 were the motion picture, the phonograph, and the radio. All became widespread during that time. By 1940, for instance, 80 percent of American households had a radio. Gordon also documents the huge improvement in entertainment between 1940 and 1970 (mainly TV) and between 1970 and now (smart phones that double as computers and music and video players).

If you aren't already convinced that health care in 1870 was primitive, you will be when you read Gordon's book. In a subsection titled "What Did Doctors Do?" he shows that they didn't do much good. For one thing, they still hadn't bought into germ theory, which was relatively new. A few decades later, they had. Between 1870 and 1940, not only their knowledge, but also their tools, improved. These included the stethoscope, the ophthalmoscope, the laryngoscope, the X-ray machine, and the electrocardiogram.

Gordon understates medical progress since 1970, mainly, it seems, because the big increases in life expectancy had already happened. But there have been significant improvements since 1970 in quality-of-life medicine. If I need a knee replacement now, I can have one; in 1970 or even in 1980,1 couldn't.

To his credit, he points out that regulation by the Food and Drug Administration has now slowed innovation to a crawl. Referencing work by Jan Vijg, a molecular geneticist at the Albert Einstein College of Medicine in New York, Gordon writes, "Had the regulatory norms of today existed in the 1940s, Vijg argues, innovations such as kidney dialysis and antibiotics might never have come to fruition." Who knows how many diseases we would have a better handle on now if not for the FDA's heavy regulatory hand?

The kinds of jobs we have and our pay have also improved dramatically. Although, surprisingly, Gordon presents no aggregate data showing the huge drop in workplace fatalities and injuries since 1870, he does give enough of a narrative to make you feel lucky not to be working under the conditions of the late 19th century. Hours were long, jobs were tedious and dangerous, and pay was low. The nature of work changed also. He writes, "The big shift over the century after 1870 was from truly disagreeable jobs mainly to repetitive occupations, leaving room for a small shift to nonroutine cognitive employment." After 1970, there was another shift. "The ratio of disagreeable to non-routine cognitive jobs shifted from 7.9 in 1870 to 2.1 in 1940 to 0.1 in 2010, one of the great achievements of American economic growth over the past fourteen decades." On pay, Gordon shows that between 1870 and 2010, real hourly compensation for production workers roughly twelve-tupled.

He shows that the first big improvement in clothing was the shift from home production to factory production, which helped women to pursue higher-value work outside the home. A much later shift was to imports. "The post-1980 years," he writes, "observed a near total replacement of domestic-produced clothing by imports." Because of increased imports, he notes, the rate of decline of clothing prices "more than quadrupled." Their decline averaged 0.6 percent per year from 1940 to 1980, and then 2.6 percent per year from 1980 to 2013. Gordon cautions that the downside was the loss of 650,000 apparel jobs between 1997 and 2007, when Chinese imports increased rapidly. But given his earlier emphasis on the shift to "nonroutine cognitive jobs," isn't this decline largely a good thing? How, after all, are we Americans to get more interesting jobs if we don't give up the less-cognitive jobs that can be done more cheaply in other countries?

Moreover, the main way we get progress, as Gordon shows, is with increased productivity, and increased productivity means doing more and more with fewer and fewer people. Consider U.S. agriculture, which is very productive. A table in his book shows that farmers were 46 percent of the work force in 1870 and only 1.1 percent in 2009.

But he seems at times to forget how progress comes about. He thinks that Amazon. com has "weakened the economy" because he predicts that "there will be fewer jobs for construction workers building new shopping malls and fewer jobs for clerks, stocking staff, and managers at retail stores." This is shocking coming from an economist who, elsewhere in the book, shows his understanding that producing greater output with fewer people is good, not bad.

Bad news? / Other reviews of Gordon's book have given considerable attention to two of his claims: economic growth has slowed since 1970, and it will slow even more in the future.

His data do show lower growth since 1970. The good news is that economic growth since 1970 is higher than the data suggest. The reason is that the Gross Domestic Product Deflator and the Consumer Price Index (CPI), both of which are used to adjust for inflation, overstate inflation, and therefore understate economic growth. Gordon would almost certainly agree with this point. In 1996, he was a member of the Boskin Commission, appointed by the Senate Finance Committee and headed by Stanford University and Hoover Institution economist Michael Boskin. The commission estimated that the CPI overstated inflation by 1.3 percentage points annually before 1996. Although Gordon published some of his differences with the commission's report, they were small. In a 2006 study published by the National Bureau of Economic Research, he estimates the bias to be at least 1.0 percentage point per year. If he's right, official U.S. government data substantially understate current growth. Of course, that doesn't undercut his point that growth since 1970 has been lower than before 1970 because the data before 1970 also overstated inflation and, therefore, understated growth.

But then the big question is "Why has growth fallen?" And here, Gordon disappoints. There seems to be a big elephant in the room, namely the increase in government scope and power, yet he doesn't mention that. In 1966, Medicare and Medicaid, two huge government programs, began and grew exponentially. In 1970, President Richard Nixon signed an executive order and congressional legislation creating two large agencies: the Environmental Protection Agency and the Occupational Safety and Health Administration. Since then, regulation has exploded even further, at the federal, state, and local levels. So there's a good case to be made that the drop in growth is due, in part, to a large growth in government.

Gordon's more controversial claim is that future growth will fall even further. He forecasts 1.2% annual growth in output per hour from 2015 to 2040. That compares to annual growth of 2.71% between 1948 and 1970. He then makes three adjustments, all of which drive the number lower. The 1.2%, he argues, falls to a 0.8% increase per person because of the retirement of the baby boom generation. The 0.8% then falls to 0.4% growth in median output per person because, he argues somewhat plausibly, a disproportionate share of the gains from growth will go to higher-income people. Finally, he reduces the 0.4% to 0.3% annual growth in disposable median income per person because the government will find itself raising taxes to bail out Social Security and Medicare.

This does sound grim. The good news, as noted earlier, is that both the CPI and the GDP Deflator overstate inflation and understate growth. So, even in Gordon's pessimistic scenario, growth of median disposable income per person could be, say, 1% annually.

But his most questionable number is the 1.2% annual growth in output per hour. How does he reach this? By taking the average growth from 1970 to 2015 and reducing it somewhat for a slowdown in the growth of educational attainment. In other words, he extrapolates from the last 45 years. That's pretty iffy. He quotes Joel Mokyr, an economic historian and Gordon's Northwestern University colleague, who observes, "History is always a bad guide to the future and economic historians should avoid making predictions." Gordon doesn't heed that advice.

He does realize that growth depends on economic policy. Unfortunately, with some important exceptions, the policies he proposes would reduce growth. Because he worries so much about income inequality, he advocates raising the minimum wage and substantially increasing tax rates on dividends and capital income.

On the minimum wage, he writes: "Standard economic theory implies that an increase in the minimum wage would raise the unemployment rate of the low-wage workers. However, a substantial body of economic research indicates little or no effect." Really? Is he unaware that another, much more substantial body of economic research indicates a bigger effect? So we have theory indicating an effect, some evidence of little or no effect, and more evidence indicating a bigger effect. Doesn't it make sense, then, if one is worried about growth, to go with the theory?

And surely Gordon must be aware that taxing capital more heavily, as increased tax rates on capital gains and dividends would do, would reduce the incentive to invest in capital. With a lower growth of capital, there would be lower growth of productivity and, therefore, output. Yet he mentions not a word about this large downside to his proposal. At that point in the book, I got the impression that one of his agenda items was to push for his preferred policies regardless of their effects on growth.

To his credit, he does call for large doses of deregulation. He would reduce the imprisonment rate of Americans, which, he points out, is a large multiple of the rate in Europe. That would save taxpayer money and increase the real incomes of people who would otherwise be in prison. He also calls for drug legalization, pointing out the savings to taxpayers and the gains in tax revenues from taxing drugs, although not mentioning the gains to people who, as a result of legalization, don't end up in prison. Possibly he thought that this gain was obvious. He also calls for rolling back regulations on land use, which he notes, citing Harvard economist Edward Glaeser, transfer wealth from the less affluent to the more affluent and also reduce productivity. Finally, he would get rid of many regulations that restrict occupational choice and thereby "restrict upward mobility for lower-income individuals."

If you want to see how far we have come and how tough life was a century and a half ago, read Gordon's book. If you want that progress to continue, eschew his tax and wage policy recommendations and go for large cuts in regulation.

DAVID R. HENDERSON is a research fellow with the Hoover Institution and professor of economics in the Graduate School of Business and Public Policy at the Naval Postgraduate School in Monterey, Calif. He is the editor of The Concise Encyclopedia of Economics (Liberty Fund, 2008).