Cornucopia: Increasing Wealth in the Twentieth Century

Copyright 1991-2000 J. Bradford DeLong

.pdf version

.doc version

 

Abstract

 

J. Bradford DeLong
Department of Economics, U.C. Berkeley, #3880
Berkeley, CA, 94720-3880
(510) 643-4027 phone; (510) 642-6615 fax
delong@econ.berkeley.edu
http://www.j-bradford-delong.net/


Cornucopia: Increasing Wealth in the Twentieth Century

Copyright 1991-2000 J. Bradford DeLong

5,844 words

 

 

A. Increasing wealth

This twentieth century has been above all the century of increasing wealth.

No previous era we know of has seen anything like the proportional growth in material, economic wealth--in the productivity of workers and the standards of living of consumers--that the twentieth century saw in the industrialized democracies that make up the core of the world economy. No previous era and no previous economy has seen anything like the level of material wealth and productive potential attained by the twentieth century's economy.

Today the average American possesses a degree of material comfort that in many ways outstrips the reach of even the richest humans of previous centuries. Perhaps a billion people living today are within striking distance of middle-class American productivity levels and living standards. Moreover, even lower middle-class households in relatively poor countries have material standards of living that in many dimensions--access to entertainment and news, public health, variety and extent of diet, potential literacy, materials with which to build shelter--would make them the envy of many of the prosperous of past centuries.

Perhaps the best indicator of the extraordinary level and rate of advance of material well-being and productive potential is that we take it for granted. If in the eighteenth century people began to think of the idea of progress, and in the nineteenth there actually began to be visible progress, in the twentieth century we expected and today we expect progress. We assume that each generation will live between half again and twice as well in material terms as its parent generation. We find it hard to imagine what it would be like to live in a society not experiencing rapid material progress. And the gulf in material standards of living between the end of the twentieth century and the beginning of the twentieth century is so large that it is hard for us to grasp how large a gulf there is between how we lived and how people lived in previous centuries.

 

1. Montgomery Ward’s prices

One place to begin to begin to measure the gulf is the 1895 Montgomery Ward catalog. At the turn of the last century, Montgomery Ward was the largest mail-order retailer in the United States. It supplied rural and small-town households around the country with goods produced in America’s factories. It was one of the few ways the forty percent of Americans living in small towns or on the farm could buy the products of modern industry. The regular arrival of the mail-order catalogues, followed by goods shipped from centralized warehouses proved a boon to rural consumers and a comfortable market niche for companies like Montgomery Ward, willing to supply goods ranging from sterling silver teaspoons to sets of the Encyclopedia Britannica to drill presses.

 

Multiplication of Productivity 1895-2000

Time Needed for an Average Worker to Earn the Purchase Price of Various Commodities

 

Commodity Time-to-Earn in 1895 (Hours) Time-to-Earn in 2000 (Hours) Productivity Multiple
Horatio Alger (6 vols.) 21 0.6 35.0
One-speed bicycle 260 7.2 36.1
Cushioned office chair 24 2.0 12.0
100-piece dinner set 44 3.6 12.2
Hair brush 16 2.0 8.0
Cane rocking chair 8 1.6 5.0
Solid gold locket 28 6.0 4.7
Encyclopedia Britannica 140 33.8 4.1
Steinway piano 2400 1107.6 2.2
Sterling silver teaspoon 26 34.0 0.8

Source: 1895 Montgomery Ward Catalogue


Comparing the prices charged in the Montgomery Ward catalog with prices today--both expressed as a multiple of the average hourly wage--provides an index of how much our productivity in making the goods consumed back in 1895 has multiplied.

Consider a one-speed bicycle, costing $65 if ordered from Montgomery Ward in 1895. The price of a bicycle measured in "nominal" dollars has more than doubled over the past century (as a result of inflation). But the bicycle today is much less expensive in terms of the only measure that truly counts, its "real" price: the work and sweat needed to earn its cost. It took perhaps 260 hours’ worth of the average American worker’s production in 1895 to mount up to enough money to buy a one-speed bicycle. Today a bicycle--the high-tech cutting-edge consumer good of 1895--costs 1/36 as much in labor time today as it did back then. On the bicycle standard–measuring wealth by counting up how many bicycles it can buy–the average American worker today is some 36 times richer than his or her counterpart was back in 1895.

Other commodities tell their--different--stories. A six-volume set of (cheaply made) novels by Horatio Alger costs 1/35 as much in this labor-wage-metric now as it did a century ago. A 100-piece dinner set from Crate and Barrel today costs 1/12 as much in labor time as a set from Montgomery Ward used to. A cushioned office chair has become 12 1/2 times cheaper.

But there are other goods with less of a productivity multiple: a multiple of less than 5 for a solid gold locket, a productivity multiple of only a little more than two for a Steinway piano, and a productivity multiple of 0.8 for a sterling silver teaspoon: it costs more hours to earn the money to buy a silver teaspoon (mail order from Ross Simons) today than it took back in 1895. There are some commodities for which our productive potential has not increased over the past century.

 

Caveats

Or has it? A farm household ordering a few silver implements back in 1895 was presumably seeking flatware that would not corrode rapidly. They did not know how to mix chromium atoms with iron and carbon atoms to make corrosion-resistant flatware. We do. Thus our everyday utensils are made of stainless steel: our silver is reserved for when (and who) wants to set a glittering table. Is the right comparison to make that of the price of a sterling silver teaspoon then with a sterling silver teaspoon now, or of a silver teaspoon then with a stainless steel teaspoon now? It matters a lot. For those who think that the important characteristic is that it is made up of silver, it is indeed 25 percent more expensive now than it was back then when you could pick the silver up off of the ground in Nevada. But for those who think that the important characteristics is that it does not rust, a teaspoon today costs only one-fiftieth as much in terms of labor time as it did a century ago.

Or consider the Encyclopaedia Britannica. Today its print version costs $1,250--in labor hours one quarter as much as it cost back in 1895. But the Encyclopaedia Britannica CD 2000 costs $49.95--a drop in labor-time price of a factor of 100, if the CD is taken as an equivalent product. And the Concise Columbia Electronic Encyclopedia can be accessed over the internet for free.

Even comparing commodities made in 1895 with the like commodities made today is not quite straightforward.

 

 

[Picture: commodities in the table above]

 

 

The Index Number Problem

The answer to the question "how much wealthier are we today than our counterparts of a century ago?" depends on which set of commodities you view as central and important. If you care only about personal services–having a butler around to answer the door and polish your silver spoons–then you would find little difference in national average wealth between 1895 and 1990: an hour of a butler's time then cost about an hour's worth of the time of an average worker; an hour of a butler's time today costs about the same; on the butler-hiring standard we are no richer off than a century ago. But suppose you care a lot, instead, about your ability to by mass-produced manufactured goods–like bicycles. Then the multiple is 36.

This divergence is the index-number problem. It is ultimately unresolvable: there is no single, unique, correct index that will tell you how much higher wealth of productivity is: it all depends on what you value, and what set of weights you choose to evaluate the different production possibilities of different eras. Thus it depends in a very real sense on just who you are. If your tastes, needs, and desires are different, the appropriate measure of economic growth will be different too.

 

2. Long-Run Estimates of GDP per Worker

Nevertheless the government arrives at a single, consensus, official or semi-official estimate of the pace of long run economic growth. It publishes the time series of real (that is, inflation-adjusted) GDP--gross domestic product--in Historical Statistics of the United States. You can construct a picture of long-run economic growth by taking the time series for real GDP from Historical Statistics, splicing the series onto contemporary estimates of current GDP, and then divide GDP by the number of workers in the American economy to arrive at an estimate of how the economy's per-worker productive potential has changed over time.

What is this "GDP" that is the standard measure of economic output? Annual GDP is a measure of all the final goods and services produced in a country in a year. It includes all those consumption goods and services produced--those bought by households for their own use. It includes all investment goods purchased by businesses to expand their stocks of productive capital. And it includes goods and services purchased by the government, and used by the government to accomplish its tasks.

As of the end of 1999, forecasts of U.S. GDP in 2000 put it at $9,300 billion, which with 142 million workers (employed plus unemployed) comes to an annual real GDP per worker (measured in dollars of the year 2000's purchasing power) of $65,540. Back in 1890, a little more than a century ago, the spliced-together time series from the Bureau of Economic Analysis tells us that GDP in 1890 (at the year 2000's prices) was $300 billion. With an 1890 labor force of 21.8 million that translates into an annual real GDP per worker in 1890 of $13,700.

The semi-official time series tells us that material standard of living and potential economic productivity at the start of the third millennium was nearly five times what it had been only 110 years before: a rate of per-worker real economic growth of 1.4 percent per year. And if we adjusted for the decline in the working year since 1890, we would find a six to sevenfold increase in measured per-work hour real GDP. According to these estimates, an American household with six times the median family income back in 1890--a household as wealthy and as high up in the relative income distribution then as a household with an income of $300,000 today--was no better off in measured material welfare than the average American household today.

Note that this broad upward sweep in measured output per worker is the central feature of the figure above. On this scale, the business cycle-centered concerns of newspaper financial pages are barely visible. The recessions and depressions and the transitory booms that make headlines are by and large hard to see. The 1970 and 1991 recessions are small ripples. The 1974 oil shock and the 1982 inflation reduction depressions are small notches. They had little effect on the volume of production. They are only minor interruptions in its upward march.

There is one exception. Between 1929 and 1950 the U.S. undergoes the Great Depression, the immense boom of World War II, and then the process of reconversion from the total war economy of 1942-1945 to a peacetime economy. %he Great Depression of 1929-1941 temporarily annihilated a generation's growth in incomes. It saw unemployment peak at a quarter of the labor force. Unemployment remained above ten percent and production per worker remained below its 1929 level. until the beginning of World War II. The Great Depression provoked fears that the run of economic growth that had commenced with the industrial revolution had played itself out, and that "secular stagnation" had set in. But the Great Depression was unique. It has not been repeated. Neither has the extraordinary military mobilization of World War II.

Moreover, this upward jump of productivity and wealth has not been confined to the world economy's industrial core. Even as early as 1987, 97 percent of households in Greece--not usually considered one of the world’s industrial leaders--owned a television set. In Mexico in that same year there was one automobile for every sixteen people, one television for every eight, one telephone for every ten.

Thus Historical Statistics produces an estimate of the multiplication of economic wealth over the past century in the range given by the prices in the century-old Montgomery Ward Catalogue. A six to seven-fold multiplication is more than our increasing power to make some of the commodities listed (Steinways, solid-gold lockets) and less than our increasing power to make other commodities (bicycles, chairs, and books).

 

 

3. Alternative Answers

But the set of calculations underlying the numbers reported in Historical Statistics are only one way of calculating increasing wealth. The time series of Historical Statistics is, mostly and roughly, what economists call a Laspeyres index. It corresponds to a certain conceptual experiment. Suppose that we took a representative sample everything produced in some year in the past, stuffed it into a time machine, moved it forward to today, and sold it; how much would it be worth? That is what the "2000 prices" in the statement "GDP per worker in 1890 was some $13,700 a year at 2000 prices" means. And the resulting estimate of long-run economic growth compares the value at today's prices--$13,700--of the commodities produced by an average worker then to the value at today's prices--$65,000--of the commodities produced by an average worker today.

A little thought will convince you that this is a significant understatement of the extent and magnitude of economic growth. When we hear that "average GDP per worker in 1890 was equal to $13,700 at 2000 prices," we think that the material standard of living then was about what we could obtain now if we had $13,700 to spend to support us for a year. But that is not the case: the material standard of living then was roughly what we could obtain now if we had $13,700 to spend, and were required to spend it all on commodities that have been around for more than a century. People then could buy wrought-iron fences, bicycles, books by Horatio Alger, cushioned chairs, flour meat, and a primitive telephone. They could not buy modern entertainment or communications or transportation technologies, no modern appliances, no modern buildings, no antibiotics, no air travel. An income of $13,700 today that must be spent exclusively on commodities already in use in the late nineteenth century is, for all of us, worth a lot less than $13,700.

The argument that our commodity-focused price indices miss most of the real action–that price indices focusing on the services provided would produce vastly greater estimates of long-run economic growth–is made most powerfully by William Nordhaus in his study of the economic cost of light. Nordhaus attempts to construct a consistent series of the real labor-time cost of illumination from the dawn of civilization until today. He concludes that the past hundred years have seen a ten thousand-fold decline in the real price of illumination. Yet commodity-based price indices have only captured a ten-fold decline in this real price.

Nordhaus guesses that the Historical Statistics estimates of economic growth have understated true economic growth since 1800 by between 0.5% and 1.4% per year–an amount that cumulates to a multiplicative factor of between 3 and 15 over the past to centuries, and to a conclusion that real wages since 1900 have multiplied by a factor between 20 and 100.

Inventions largely commercialized and diffused in this century include late-nineteenth century inventions like the monorail, the telephone, the microphone, the cash register, the phonograph, the incandescent lamp, explosives, the electric train, linotype printing, the steam turbine, the gasoline engine, the streetcar, movies, motorcycles, automobiles, concrete-and-steel construction, electric appliances of all kinds, inflatable tires, radio, aspirin, x-rays, taxicab meter. Twentieth-century inventions include the espresso machine, plastics, airplanes, helicopters, hydrofoils, the zipper, the traffic light, heat-resistant glass, television, bulldozers, antibiotics, highways, jet engines, radar, insulin, photocopiers, nylon, transistors, integrated circuits, computers, fiber-optic cables, videotape, oral contraceptives, lasers, CT scanners, catalytic converters, and genetic fingerprinting. A rough cut is that roughly 45% of the value of what middle-class consumers in rich industrial countries use at the start of the third millennium is in commodities that were not invented or that were not in widespread use at all in the last years of the nineteenth century.

[Figure: collage of commodities invented in the twentieth century]

Measuring economic growth as the Laspeyres index does means that we value goods produced in the past at their current prices, and so understates economic growth as it produces an estimate of a six to sevenfold improvement in material standards of living in the past century. What if we were to turn it around, and calculate economic growth valuing goods and services not at their prices as of 2000, but at their prices as of 1890? The figure below shows one of the standard diagrams of introductory economics. Along the horizontal axis of the diagram we plot the amount produced per worker of "old fashioned" goods that were invented and were in use in 1890. Along the vertical axis of the diagram we plot the amount produced per worker of "modern" goods. The point labeled "then" shows what annual per-worker production was back then in 1890: $13,700 worth (at today's prices) of old-fashioned goods and $0 of modern goods. The point labeled "now" shows what annual per-worker production is today: (roughly) $35,000 worth (at today's prices) of old-fashioned goods and (perhaps) $30,000 worth (at today's prices) of modern goods.

 

 

In this oversimplified and stripped-down way of looking at it, production "then" and production "now" are each two dimensional quantities: you simply cannot compare them--say that one is X times larger than the other--without imposing more structure.

 

One way to proceed--the way that Historical Statistics more-or-less proceeds--is to use the fact that $1 worth of modern goods buys $1 worth of old-fashioned goods today, and to look not at the points showing what is actually produced but at all the possible combinations of goods that could be purchased (at today's prices) with the income needed to buy per-worker production either then or now. Draw, as in the figure above, the price line showing how dollar-for-dollar one could at today's prices buy more of modern and less of old-fashioned goods. Compare the positions of the lines that show the resulting sets of available choices.

Because the two lines are parallel, the ratio of their relative distances from the bottom-left corner shows the ratio of the today’s-dollar value of production "now" to production "then." This is what the Laspeyres index does: it uses the metric provided by today’s prices to make a single numerical comparison between the value of goods produced "now" and the value of goods produced "then."

But what if we took some other set of prices? Instead of taking a representative sample of everything produced in 1890, stuffing it into a time machine, bringing it forward to today, selling it; suppose we took a representative sample of everything produced today, stuffed it into a time machine, took it back to 1890, and sold it then at the prices that then prevailed?

Then we would have a very different answer, for a large chunk of what is produced now was unavailable back in 1890. It has a very high--in many cases an infinite--price. Drawing the 1890 price line showing the terms at which old-fashioned goods were then exchanged for modern goods produces a nearly horizontal line. And the relative distance between the two lines is very large, much larger than the six to sevenfold increase in output per work hour calculated from Historical Statistics.

 

We would then conclude that, measured in the metric provided by prices in the past, that economic growth over the past hundred years has been nearly infinite: even all of the resources of the economy of the past would have been unable to produce even an infinitesimal fraction of some of the goods we make today. We are nearly infinitely better at producing the floating-point arithmetic operations, or airplane flights, and antibiotics that we take for granted--because they are so cheap--today.

 

 

4. Credible Answers

Are such large estimates of the rate of economic growth and improvement in material welfare over the past century credible? The answer is both yes and no.

A credible answer to the question of how much material prosperity has multiplied over the twentieth century might look like the answer to one of the following two thought experiments:

However, these two questions will not have identical--or even close--answers. It is clear that the answer to the first question suggests a very large increase in material prosperity over the past century. Given the absence in 1890 of modern innoculations, modern antibiotics, and other technologies of the past century, it is hard to argue that anything less than an astronomical income back in 1890 could compensate. J.P. Morgan could not go to the movies, or watch football on television. He has no VCR. To travel from New York to Italy took a week, not a night. Was he better off in a material welfare sense than an average inhabitant of the U.S. today? Perhaps, but it is not clear. The answer to the first question is very large indeed.

Personally I have no problem at all with the first answer's conclusion that Historical Statistics vastly understates growth: my household's income today is roughly $200,000 a year–about three times average GDP per worker. Suppose that you stuffed me and my family into a time machine, sent us back a century to 1890, and then gave us an income equal to eighteen times that of 1890 average GDP per worker–an income that would put us at the same place in the relative income distribution then as some $1,200,000 a year would today. We would not be among the 500 or so richest families in the country that might be invited to the most exclusive parties in the mansions of Newport, Rhode Island; but we would be among the next outer circle of 5,000 or so.

Would we be happy–or at least not unhappy–with the switch? Our power to purchase some commodities would be vastly increased: we would have at least three live-in servants, a fifteen-room house (plus a summer place). If we lived in San Francisco we would live on Russian Hill, if we lived in Boston we would live on Beacon Hill. If we lived in New York we would live on Park or Fifth Avenue.

But the answer is surely that we would not be happy. I would want, first, health insurance: the ability to go to the doctor and be treated with late-twentieth-century medicines. Franklin Delano Roosevelt was crippled by polio. Nathan Meyer Rothschild–the richest man in the world in the first half of the nineteenth century–died of an infected abscess. Without antibiotic and adrenaline shots I would now be dead of childhood pneumonia. The second thing I would want would be utility hookups: electricity and gas, central heating, and consumer appliances. The third thing I want to buy is access to information: audio and video broadcasts, recorded music, computing power, and access to databases.

None of these were available at any price back in 1890.

I could substitute other purchases for some. I could not buy a washing machine, but I could (and would) hire a live-in laundress to do the household's washing. I could not buy airplane tickets; I could make sure that when I did travel by long distance train and boat I could do so first class, so that even though travel churned up enormous amounts of time it would be time spent relatively pleasantly. But I could do nothing for medical care. And I could do nothing for access to information, communications, and entertainment technology save to leave the children home with the servants and go to the opera and the theater every other week. How much are the central heating, electric lights, fluoridated toothpaste, electric toaster ovens, clothes-washing machines, dishwashers, synthetic fiber-blend clothes, radios, intercontinental telephones, xerox machines, notebook computers, automobiles, and steel-framed skyscrapers that I have used so far today worth–and it is only 10 A.M.?

I would not be satisfied with my attempts to substitute using late nineteenth century technology. First of all, I would be dead. Second a very large chunk of my-high-material standard of living is the broad range of commodities newly-invented over the course of the past century that I can choose to purchase, and that I do use because they give me capabilities that were simply not possible a century ago.

By contrast, the answer to the second question suggests a smaller increase in material prosperity--perhaps a factor of ten. Someone with an income of $5,000 a year in the U.S. today has much better access to medical care and mass entertainment than a middle-class household of a century ago, and better transportation and clothing. He or she has better winter vegetables, but perhaps a worse overall diet. And he or she has worse housing, worse access to non mass-media forms of entertainment, and a much lower ability to purchase goods that are currently fashionable. The comparison of the poor today with the middle class of a century ago produces an estimate of the pace of economic growth not too dissimilar from that produced by Historical Statistics. The comparison of the middle class today with the rich of a century ago produces an estimate that is much much larger. So how much economic growth is worth to you depends on where you sit: for those near the bottom of the income distribution in industrial economies growth looks much less impressive, in large part because many of the new commodities invented over the past century are of no use if you cannot afford them.

So do we shrug our shoulders and accept the Historical Statistics answer that we today are eight times as rich as our counterparts of a century ago? (And that the gulf is larger if we care about manufactured goods; and smaller if we care about personal services, and some kinds of luxuries?) No, I do not think that we accept this answer, because the Historical Statistics answer is the least of the possible answers we could arrive at: we know that it provides a lower bound to "true" growth. It provides a lower bound because the calculations that underlie it leave out the many things we make today that were not made back in the 1890s.

If we must put forward one number, we will probably due the least damage to reality if we take the Boskin Commission's guesstimate that unmeasured improvements in quality and the invention of new goods and new types of goods have led standard measures to understate true economic growth (from a first-world middle-class perspective at least) by 1% per year. Such an assumption leads to a conclusion that economic output per worker has multiplied sixteen-fold since 1890.

 

 

5. The Limit of Human Felicity

However, in the end the quantitative estimates of the pace of economic growth--and the wide range of such estimates generated by the unresolvable index number problem--is in some ways less interesting than a picture of the qualitative difference twentieth century economic growth makes. We can see the magnitude of the contribution that the changing set of commodities we can produce makes to our wealth by reading Looking Backward, Edward Bellamy’s turn of the last century utopian novel. In Looking Backward the narrator–thrown forward in time from 1895 to 2000–hears the question, "Would you like to hear some music?"

He expects his host to play the piano–a social accomplishment of upper-class women around 1900. To listen to music on demand then, you had to have–in your house or nearby–an instrument, and someone trained to play it. It would have cost the average worker some 2400 hours, roughly a year at a 50-hour workweek, to earn the money to buy a high-quality piano, and then there would be the expense and the time committed to piano lessons.

But today, to listen to music-on-demand in your home, all you need is a CD or a tape player–or in a pinch, if you are willing to let others choose your music for you, a radio. The labor-time value of a Steinway piano may have only halved when measured in average worker-hours. But if what you value is not the piano itself but the capability of listening to music at home, the cost has fallen from 2400 average worker-hours a century ago to 10 hours today (240 dollars for the boom-box plus 10 dollars for the CD). So when we calculate the increase in material wealth, do we count the halving of the labor-time price of the commodity (which is what Historical Statistics does); or do we count the 240-fold decrease in the real labor-time price of the capability of listening to piano music? And whose piano playing do you really want to listen to?

After answering "yes" to the question "would you like to hear some music?" Bellamy’s protagonist is stupefied to find his host "merely touched one or two screws," and immediately the room was "filled with music; filled, not flooded, for, by some means, the volume of melody had been perfectly graduated to the size of the apartment. ‘Grand!’ I cried. ‘Bach must be at the keys of that organ; but where is the organ?’"

He learns that his host has called the orchestra on the telephone–for in Bellamy's utopia you can dial one of four orchestras, and then put it on the speakerphone. Bellamy’s protagonist then says that:

To Edward Bellamy–a self-described utopian visionary, a late-nineteenth century minister’s son from western Massachusetts–a radio that could tune into any of four stations is "the limit of human felicity..." What if someone were to take Edward Bellamy to Tower Records? Or Blockbuster Video? His heart would stop. Yet we do not think of our modern ability to cheaply listen to high-fidelity go-anywhere listen-to-anything music as a remarkable or even a notable part of our economy. We do not daily give thanks for our cassette players and genuflect in front of our CD collections. We do not reflect that they have brought us to the limit of human felicity.

The technological inventions of the past century have transformed experiences that were rare and valued luxuries–available only to a rich few at great expense at relatively rare performances of the symphony or the opera–into features of modern life that we take for granted. Bellamy’s view of us might be somewhat analogous to our view of a civilization in which everyone had several boxes of gem-quality diamonds sitting in their basement, ignored because no one could find a use for them, and in which no one thought of these boxes as in any way interesting.

If you asked Edward Bellamy--or any other nineteenth-century or earlier sketcher of utopias--whether we here today have the knowledge of technology and of productive organization needed to provide at least the material abundance needed to build a utopia, they would all say "of course." And they would in turn ask of us why we do not recognize that those of us in the middle and upper classes of the industrial economies have, in material well-being at least, reached the limit of human felicity.

 

 

 

B. Economic Growth in Long-Run Perspective

1. Before and Since the Commercial Revolution

The world as a whole has not had the sixteen-fold multiplication of its material prosperity that our third-millennium middle-class standpoint sees in the United States over the past century. It is a mistake to claim that growth and development over the course of the past century has been confined to the United States, or to the United States and western Europe, or to the United States, western Europe, and Japan. It is true that only twelve percent of the world's population lives in countries where GDP per capita at the start of the third millennium is more than $20,000 per year: the world distribution of income today is skew to a remarkable extent.

But comparing cross-national estimates of material productivity measured at exchange rates that equalize purchasing power across countries to our guesses about the pace of long-term economic growth above suggests that the average inhabitant of Thailand of Tunisia today has three times the productive potential of the average inhabitant of the United States in 1900; and the average inhabitant of Argentina, Botswana, Uruguay, or Mexico has five times the material productive potential of the average inhabitant of the U.S. in 1900. (However, unequal income distributions within nations make such estimates misleading as guides to economic welfare, even leaving aside the heroic and shaky assumption that comparisons across countries today can be related to comparisons across time within a country.)

Perhaps 36% of the world's population in 2000 lives in a country with a level of material output per capita less than that of the United States in 1900. But even in such poor countries today, most inhabitants are living much better than their predecessors. Angus Maddison estimates that world per capita GDP at the end of the twentieth century is five times what it was at the century's start--and Maddison's estimates are (in concept at least) Laspeyres estimates that make insufficient allowance for technological change and the invention of new commodities. Making half as great an allowance for the impact of new commodities and technologies produces an estimate of a ninefold increase in world GDP per capita over the twentieth century.

Thus not just in the United States, but worldwide, the twentieth century is unique in its pace of economic growth. Such rapid growth in standards of living has never been seen before, anywhere. The nineteenth century saw perhaps a doubling of measured material standards of living in the United States–perhaps a tripling once proper account is taken of the impact of new technologies like the railroad and the telegraph, and the expanded range of technological capabilities. Nineteenth century growth was itself remarkably fast: people christened the nineteenth century the "industrial revolution" because it was remarkable compared to what had happened before. And before the nineteenth century growth was even slower. The standard of living in the Netherlands, probably the richest economy in the world at the end of the eighteenth century, might (or might not) have been some fifty percent higher than it had been three centuries before, at the time of the Renaissance.

 

[Figure: the countryside: Roman Italy, circa 0; Netherlands, circa 1500; U.S. midwest, circa 1900; U.S. midwest, today]

 

And before that? Between the invention of agriculture and the commercial revolution that marked the end of the middle ages, wealth and technology developed slowly indeed. Medieval historians speak of centuries and half-millennia when they speak of the pace at which key inventions like the watermill, or the heavy plow, or the horse collar diffused across the landscape. And improvements in technology relatively quickly led to increases in population, until the human population once again reached a new Malthusian steady state in which births were held in checks by death. For most of human history before the industrial revolution, increases in technological capability led to increases in the population that could be supported on a given natural resource base, with little if any appearing as an improvement in the median standard of living.

So slow was the pace of change that people, or at least aristocratic intellectuals, could think of their predecessors of a thousand years before or more as effectively their contemporaries. And they were not far wrong. Marcus Tullius Cicero, a Roman aristocrat and politician of the generation before the Emperor Augustus, might have felt more or less at home in the company of Virginia planter Thomas Jefferson. The slaves outside grew different crops. The plows were better in Jefferson's time. Sailing ships were much improved. But these might have been insufficient to create a sense of a qualitative change in the order of life for the elite. And at bottom being a slave of Thomas Jefferson was probably a lot like being a slave of Marcus Tullius Cicero.

 

[Figure: Slaves at Monticello–Slaves working the fields of ancient Rome]

Overall, however, the differences in standards of living and in technologies used to manipulate the world were small, or at least "small" relative to the pace of change in the nineteenth and twentieth centuries. Even the first century of the industrial revolution produced more "improvements" than "revolutions" in standards of living. With the railroad and the spinning and weaving of textiles as very important exceptions, most innovations during the first century or so of the industrial revolution proper were innovations in transportation, in how goods were produced, and in new kinds of capital but not consumer goods. Improvements in productivity--in the first half of the nineteenth century at least--were concentrated in a few relatively narrow sectors rather than spread throughout the economy.

So slow was the pace of improvement that literary intellectuals in the first half of the nineteenth century debated whether this industrial revolution was worthwhile. Was it an improvement or a degeneration in the standard of living? And opinions were genuinely divided, with as optimistic a liberal as John Stuart Mill coming down on the "pessimist" side as late as the end of the 1840s. The figure above shows–approximately–the relative pace of economic growth in productivity levels and material wealth for the world as a whole over the past ten centuries. The estimates are rough and approximate only. But the figure does not do violence to the qualitative picture of relative rates of economic growth over the past ten centuries. And in the leading-edge economies of Europe (plus the European-settled North American economies) the acceleration of growth into the twentieth century was an order of magnitude faster still.

 

 

2. Massive and colossal productive forces

In 1848, in the middle of the nineteenth century, before the industrial revolution proper had spread far from its original homes in Belgium and in the British midlands, a young German philosopher-turned-political activist marveled at the extraordinary pace of economic growth in his day. He saw it as a new historical epoch that was only a century old, and yet was opening wide the door to utopia. He saw the epoch as equivalent to that of Prometheus, the mythological Greek demigod who defied the chief god Zeus, brought knowledge of fire to humanity, and transformed humanity's condition.

The young Karl Marx wrote that the economically ruling class–the capitalist class, the entrepreneurial class, the business class, the bourgeoisie–of this epoch was:

Karl Marx was dumbfounded at the pace of the economic transition he saw around him in the middle of the nineteenth century. Yet compared to the pace of economic growth in the twentieth century, all other centuries–even the nineteenth century that so impressed Karl Marx–were standing still.

 

 

 

 

MEASURED REAL GDP PER WORKER

Year Real GDP per Worker (2000$)

1890

13738.355

1891

14644.5276

1892

13669.0583

1893

12937.5563

1894

14181.8993

1895

13523.1137

1896

14490.2359

1897

14465.6027

1898

15412.8386

1899

15471.7483

1900

16861.4535

1901

16722.0454

1902

17200.3528

1903

16662.1965

1904

17600.2181

1905

19295.2626

1906

19200.0017

1907

17309.301

1908

19050.7212

1909

18896.4562

1910

19177.2424

1911

19670.9182

1912

20092.5419

1913

18204.3009

1914

18399.5885

1915

20606.4618

1916

19742.725

1917

20539.4018

1918

19930.068

1919

19878.4804

1920

19456.3087

1921

20284.3461

1922

22591.8576

1923

22875.1653

1924

23068.4126

1925

24172.0508

1926

23999.559

1927

23897.778

1928

24917.6593

1929

22198.6028

1930

20003.1108

1931

17124.5741

1932

16569.6418

1933

17637.2288

1934

18874.4947

1935

21305.2734

1936

22120.5391

1937

20900.6967

1938

22288.8411

1939

23766.9552

1940

27675.9101

1941

31956.7392

1942

35355.9812

1943

36010.2234

1944

33386.9019

1945

26779.4605

1946

29588.0454

1947

31199.5106

1948

31680.2439

1949

34863.5452

1950

36161.2036

1951

36994.8257

1952

37873.1038

1953

36792.5043

1954

38217.6335

1955

38406.0935

1956

38455.5642

1957

37566.7052

1958

39147.3771

1959

39392.4435

1960

39783.5604

1961

41197.5659

1962

42147.4061

1963

43673.3701

1964

45418.9074

1965

47243.0487

1966

47772.7888

1967

48912.4276

1968

49268.1325

1969

48298.5652

1970

48821.7682

1971

49918.6476

1972

51148.4694

1973

49538.7886

1974

47642.4325

1975

48651.992

1976

49592.9508

1977

50859.6237

1978

50754.118

1979

49349.616

1980

48979.1276

1981

46971.8951

1982

47863.0903

1983

50278.6464

1984

51120.1566

1985

51726.7783

1986

52615.0628

1987

53732.8151

1988

54874.3128

1989

55824.1065

1990

55582.0345

1991

54453.972

1992

55286.803

1993

55913.7448

1994

57211.1134

1995

57859.8766

1996

59145.2009

1997

60751.4333

1998

61814.2835

1999

63782.5085

2000

65034.965

 

 

REAL[?] GDP PER WORKER

Year Real GDP per Worker (2000$)

1890

4573.10112

13738.355

1891

4923.73171

14644.5276

1892

4641.95103

13669.0583

1893

4437.692

12937.5563

1894

4913.4014

14181.8993

1895

4732.24796

13523.1137

1896

5121.64171

14490.2359

1897

5164.32086

14465.6027

1898

5557.79163

15412.8386

1899

5635.10439

15471.7483

1900

6202.98209

16861.4535

1901

6213.5223

16722.0454

1902

6455.48329

17200.3528

1903

6316.35605

16662.1965

1904

6738.9983

17600.2181

1905

7462.26961

19295.2626

1906

7500.05512

19200.0017

1907

6829.44893

17309.301

1908

7592.07513

19050.7212

1909

7606.28136

18896.4562

1910

7796.88491

19177.2424

1911

8077.97574

19670.9182

1912

8334.04304

20092.5419

1913

7626.72007

18204.3009

1914

7786.0082

18399.5885

1915

8807.50947

20606.4618

1916

8523.14215

19742.725

1917

8956.19148

20539.4018

1918

8777.83283

19930.068

1919

8843.10236

19878.4804

1920

8742.28305

19456.3087

1921

9205.94491

20284.3461

1922

10356.2433

22591.8576

1923

10591.5005

22875.1653

1924

10788.3221

23068.4126

1925

11418.0683

24172.0508

1926

11450.5235

23999.559

1927

11516.5541

23897.778

1928

12128.7269

24917.6593

1929

10913.8142

22198.6028

1930

9933.25087

20003.1108

1931

8589.27656

17124.5741

1932

8394.46211

16569.6418

1933

9025.12127

17637.2288

1934

9755.30779

18874.4947

1935

11122.328

21305.2734

1936

11663.9927

22120.5391

1937

11131.5397

20900.6967

1938

11990.1581

22288.8411

1939

12913.7958

23766.9552

1940

15188.8615

27675.9101

1941

17714.4925

31956.7392

1942

19795.7561

35355.9812

1943

20364.6974

36010.2234

1944

19070.901

33386.9019

1945

15450.4047

26779.4605

1946

17242.3818

29588.0454

1947

18364.187

31199.5106

1948

18834.556

31680.2439

1949

20935.4048

34863.5452

1950

21932.8786

36161.2036

1951

22664.0067

36994.8257

1952

23435.2477

37873.1038

1953

22995.3986

36792.5043

1954

24126.167

38217.6335

1955

24488.8064

38406.0935

1956

24766.7839

38455.5642

1957

24437.4834

37566.7052

1958

25721.6596

39147.3771

1959

26142.805

39392.4435

1960

26667.718

39783.5604

1961

27893.0952

41197.5659

1962

28822.9845

42147.4061

1963

30166.696

43673.3701

1964

31687.6965

45418.9074

1965

33291.6137

47243.0487

1966

34003.2533

47772.7888

1967

35164.3051

48912.4276

1968

35776.0069

49268.1325

1969

35424.4356

48298.5652

1970

36168.0554

48821.7682

1971

37352.3053

49918.6476

1972

38657.1816

51148.4694

1973

37816.8954

49538.7886

1974

36734.7731

47642.4325

1975

37890.2094

48651.992

1976

39011.1968

49592.9508

1977

40409.68

50859.6237

1978

40731.1338

50754.118

1979

40002.0212

49349.616

1980

40100.7181

48979.1276

1981

38843.8377

46971.8951

1982

39978.6136

47863.0903

1983

42418.325

50278.6464

1984

43561.7239

51120.1566

1985

44521.6507

51726.7783

1986

45741.3382

52615.0628

1987

47182.5394

53732.8151

1988

48669.1495

54874.3128

1989

50009.1401

55824.1065

1990

50292.7046

55582.0345

1991

49767.1832

54453.972

1992

51036.1516

55286.803

1993

52133.6301

55913.7448

1994

53879.3975

57211.1134

1995

55038.0171

57859.8766

1996

56826.0844

59145.2009

1997

58955.9571

60751.4333

1998

60590.2787

61814.2835

1999

63147.8619

63782.5085

2000

65034.965

65034.965