I’ve been wondering about a few questions.
Internalizing externalities is a common theme in economics,and it’s also called capturing the value you create. Don’t economists believe this happens — and happens increasingly — all the time? Karl Smith writes (and you can find his caveat here):
TFP growth depends on the returns to innovation not being captured by the innovator. Otherwise it becomes a return to the factor of production rather than total factor productivity.
Does TFP tend to fall once it has been high for a while? Is falling TFP, following a technological breakthrough, a sign of the market’s ability to capture value and internalize externalities? And is this another reason why we might prefer imperfectly defined intellectual property rights?
The second question concerns the Industrial Revolution. There is a large cottage industry about the origins of “the rise of the West,” and so on. I am not disputing the particular causal claims made in this literature. Still, I wonder what is being explained. Arguably the potency of the technological platform of “powerful machines plus fossil fuels” was not well understood in advance. Ex post, that it led to the “rise of the modern world” was somewhat of a technological accident. In this sense, studies of the origins of the Industrial Revolution, analytically speaking, are explaining “the Industrial Revolution” (to some extent). But the “sense-reference distinction” matters here. These studies are not so much explaining “the rise of the modern world,” which is more of a technological accident than we might wish to think.
Third, there remains the issue of unmeasured gains in real wages. Let’s try a simple thought experiment. Say I’ve been at George Mason twenty years (much less since 1973) and my real wage had never gone up (not the case). But my Dean were to say to me: “Tyler, U.S. health care has some new procedures, when you’re 73 you’ll have stents, and now can surf the internet and watch reruns of Battlestar Galactica. We’ve treated you very well!” Such a claim would not pass the laugh test and few people would accept it as applied to their own employment relation. Yet many of those same people make this same argument in the aggregate. I still think that if measured real wages for a group (or individual) have not gone up very much, over a long period of time, something is wrong. Wrong with the Dean, wrong with me, whatever, but something is wrong. Who would have predicted in 1972 that measured male median wages were going to stagnate and even possibly fall? You should be shocked by this result and indeed I am.
“which is more of a technological accident than we might wish to think.”
Isn’t that one of the big points though? It took us centuries to get to the point where we could actually think logically about technology (development of science) and then have the ability to experiment (capital + freedom and competition).
What I have never understood about TGS is why real wages should stagnate? Even in times where progress is tough we should achieve 2-3% productivity growth a year, which should be reflected in real wages.
“What I have never understood about TGS is why real wages should stagnate? Even in times where progress is tough we should achieve 2-3% productivity growth a year, which should be reflected in real wages.”
- not if all the gains are captured by the innovator through intellectual property.
The point of the unmeasured gains issue is that it is not possible to compare a basket of goods in 1973 to today, at all. It is ridiculous to have a term ‘real wage’ comparison, because it’s completely meaningless. In nominal terms wages have gone up. If you’re willing to compromise that fact with an exception for price levels, you have to also be prepared to admit that simply adjusting for price levels will not give you an accurate picture. That further adjustments might reveal just as important a comparison as it is to say that a dollar today is not worth the same as it was in 1973.
You can’t have your cake and eat it here Tyler.
In the “Myth of Rich and Poor”, the authors entirely explained away wage stagnation since 1973:
1. The CPI was over-reporting inflation for about 20 – 25 years. See Boskin Commission.
2. household size shrinkage by 20%. As a result, median household income is lower.
During the last 50 years, compensation has follwed a more positive trend than wages. I believe compensation should be used whenever labour income is analyzed.
agree; but you cant measure median total compensation
Say I’ve been at George Mason twenty years (much less since 1973) and my real wage had never gone up (not the case). But my Dean were to say to me: “Tyler, U.S. health care has some new procedures, when you’re 73 you’ll have stents, and now can surf the internet and watch reruns of Battlestar Galactica.
The argument is not that real wages haven’t gone up but it’s OK because people now have a few shiny new electronic trinkets to distract them. The argument is that our measure of inflation (and of changes in real wages over time) is badly broken because it fails to account for technological advances — advances whose benefits are profound rather than trivial. I now live a life that the richest person in the world in the 1970s could not have imagined. This explosion of new wealth has come in the form of almost too-cheap-to-meter communication, information and cultural goods. I try to think of what other plausible form of wealth explosion I would have personally preferred and valued more, and there aren’t many alternatives that come to mind. Faster medical advances that would have saved loved ones and life-extension technology that would have kept them around past 100? OK, that I’d trade for. But bigger houses, faster cars, hand-crafted this and that…big heaps of 1970s luxury goods? No way would I trade.
Goolsbee and Klenow estimate the unpriced value of the internet at about two percent of gdp; that was from 2006, arguably the value has gone up since then by some amount but keep in mind that MR readers are especially likely to be “infovores” with high internet valuations.
How much of one’s time does the average consumer of internet in the US, use? How does that translate per capita?
Not that I have anything better to propose, but even for data from 2004, this was — as Goolsbee and Klenow say — a ‘highly uncertain’ measure:
Only about 0.2% of consumer spending in the U.S., for example, went
for Internet access in 2004 yet time use data indicates that people spent around 10% of
their entire leisure time going online. For goods like that, estimating price elasticities
with expenditure data can be difficult and, therefore, estimated welfare gains highly
uncertain. We show that for time-intensive goods like the Internet, a simple model in
which both expenditure and time contribute to consumption can be used to estimate the
consumer gains to a good using just the data on time use and the opportunity cost of
people’s time (i.e., the wage).
At this point, I’m not sure ‘time spent on the internet’ is even a coherent construct. If I listen to Pandora, watch a Netflix movie, or talk to a friend on Skype, am I spending time ‘on the internet’? What about navigating while driving using google maps on my smartphone? Amazon is working on a browser-hosted version of their Kindle software. If I read a book in kindle-in-a-browser, would that count?
Also, I think we overeducated informavores often underestimate the dramatic democratization of these technologies. Would you have guessed that the country with the second-most FaceBook users is…Indonesia and that Mexico, Turkey, India, and the Philippines are all in the top 10?
Slocum and ok have it right.
Sorry, I meant patrick L and slocum.
Of course, if they said we’re raising your (nominal) wages enough so that you can afford stents and internet access, people in 1972 would have been thrilled.
A stagnant median isn’t the same thing as stagnation for each person.
Your little thought experiment pertains to income mobility, which is increasing.
Exactly! How can you compare one person to a median? Every person can be way better off, but the median stagnant.
It is NOT the same set of people you are tracking. A large portion of the lowest male earners today are recent immigrants: janitors, taxi drivers, farm labor, kitchen workers, construction etc. Their personal contribution would have been close to zero in 1970.
Just ask around yourself and tell us how many people you find that say their own wages were higher in the US in 1970 than they are today?
PS. Are there any studies that track lifetime wages for a set of people over time rather than a population aggregate?
“Arguably the potency of the technological platform of “powerful machines plus fossil fuels” was not well understood in advance.” An excellent example of the dangers of the passive voice. It tends to permit muddled thought. Understood by whom?
The record is clear that critical innovative entrepreneurs like Bolton and Watt knew very well what an impact their devices would make. Their customers, regulators, financiers may not have understood. The successful innovators all had rivals who were attempting (and partially succeeding) in going in the same direction. I don’t see much accident in the process, which was driven initially by growing shortages of wood as a fuel source.
“Say I’ve been at George Mason twenty years (much less since 1973) and my real wage had never gone up (not the case).” – Tyler Cowen
This is rarely seen in the real world, because most people advance in their careers as they age, and therefore they see their real and nominal wages increasing. They aren’t aware of the fact that perhaps the previous generation of workers at that same place in their careers had better standards of living.
You could have every person’s individual real wages constantly increasing while still lowering the overall real wage via several mechanisms:
- Newer workers may start at a lower wage.
- As they progress through their careers, younger workers may increase their wage at a slower rate than older workers did.
- Older workers with higher real wages leave the workforce while leaving workers at earlier points in their careers, with lower wage growth prospects.
- Inflation makes their nominal wages higher, making real wage comparison to previous generations difficult.
Of course, all of these possibilities are reversible as well. I raise this because we tend to compare others against our current selves. We see a new hire eating Ramen for lunch and assume he’s deprived. We forget that we also acquired a taste for Ramen while in school, and didn’t drop it the minute we graduated.
This works in the other ways as well. We see someone >65 working as a greeter and assume that they don’t have the money to retire because we just *know* that we will retire as soon as we can. We forget the articles about people who don’t want to stop working completely, and want a low-hour job to just be around people.
Expected lifespan continues to increase, and the “quality” of those years is also rising. That is real progress.
Glad that someone sees the brighter (or more balanced) side of the picture. I find Tyler’s “mood affiliation” too pessimistic about these issues. It is folly to viscerally believe your sepia tinted views of bygone years tinged with nostalgia.
Technically I retired at 40. And RWN’s right, it’s highly over-rated. But in practical terms a) I agree that I never really stopped working… or even stopped working because the extra income doesn’t come in handy.
But also b) despite all that I’d never become a “greeter” at WalMart unless my survival was on the line and there was literally no other work.
You might argue that I might change my mind after age 65. But based on my experience both in stay-at-home-parent culture and, peripherally, in pending-geriatric culture the path for non-dependent people tends towards volunteering. Opportunities, and rewards, for which line and office workers appear to be largely unaware. And I agree with Tyler that, going back to the old dial-up “Senior-net” internet precursor in the 1980s, retired people have wanted to be productive but as long as they can afford it they’ve their preferred outlets have tended towards non-WalMart-greeter sorts of productivity.
figleaf
Does slower population growth negatively effect per capita income growth in that division of labor grows slower? Not all products can be internationally traded.
I believe so, but I think we’ve seen a one-for-one swap — an expanding international division of labor and a slowing expansion of the domestic division of labor
No, it improves it by increasing the returns to human capital, but not so much as to fully compensate for the slower growing population. Division of labor has more to do with technology than population. At any given population many are doing the same work.
Does TFP tend to fall once it has been high for a while? Is falling TFP, following a technological breakthrough, a sign of the market’s ability to capture value and internalize externalities? And is this another reason why we might prefer imperfectly defined intellectual property rights?
Did you read this short story also?
But median wages of the same people have not stagnated. Instead, millions of poor people have migrated to America and most are far better off. Is there wage data that excludes immigrants since 1970? Or do the smaller number of high-skills, high-wage immigrants counter the larger numbers of low-skill, low wage immigrants since 1970?
yes. The Myth of Rich and Poor had a time series study that showed wage growth didn’t slow down from the mid-70s to the mid-90s — and if you factor in a cpi that overestimated inflation, wage growth improved for that person over time.
The problem is that if you add a bunch of data points on the left, the median HAS to go down…and that’s what we’ve done with low-skill immigration. It doesn’t lower the living standard of the median american, but it does lower the median data point.
Example:
Imagine if we annexed Mexico tomorrow and the population of the US grew by 100 million people as a result. The median american workers wouldn’t see his income drop AT ALL, but the “median wage” in this country would drop by 30%. That’s essentially what has happened with immigration since the 1965 immigration act. Our economy is better off for it, but our income data set us skewed-left.
Also keep in mind that a lot of women have entered the workforce as well, which has (1) reduced the scarcity of labor and (2) skewed the income curve left — women tend to enter lower-paying career fields that offer family flexibility, like teaching, health care, retail, etc.
Also, work that in the 1950s and 1960s was unpaid (like cleaning houses) is now compensated. A housewife in 1961 didn’t have an independent income of her own, but a cleaner in 2011 does.
However, as you suggest, the big factor is the monstrosity that is the CPI. One problem in economics is that we think that we have some objective way of measuring price inflation. From this false belief, we derive that “real” statistics tell us what is actually happening and nominal statistics are something we can deflate away.
“Real wage” means nothing more than “Wages in a hypothetical world where prices did X, on that basis of subjective valuations Y”. Putting faith in them is pretty naive.
(There are also confusions between income and wealth in these discussions, but that mistake is so ubiquitous that it would be tiresome to point it out everytime someone makes it. There are entire (widely read) books, like The Spirit Level, which fundamentally make this mistake.)
When one accounts for the deficiences of CPI, changes in the labour structure and changes in the labour force, the US real wage situation doesn’t seem bad at all. As for household real income data, whenever someone uses those figures at all I assume them guilty of ignorance or deception until they’re proven innocent.
Sorry I cannot understand your questions. I’m confused because I suspect that you are asking the questions that you attempted to answer in TGS. Or are you looking for the micro-foundations of your idea about macro-stagnation?
Perhaps you can clarify my confusion and your doubts by referring your vague questions to the history of batteries as analyzed by Seth Fletcher in his Bottled Lightning, reviewed today by R. Bailey in
http://online.wsj.com/article/SB10001424052748703730804576317481276537422.html?mod=WSJ_Opinion_LEFTTopOpinion
1. How can you pose questions about changes in TFP from just the history of an invention? TFP is a macro concept that makes sense only if you rely on Solow’s accounting approach to growth. How do you suggest to use TFP in the analysis of big firms’ performance in producing batteries and changing inputs overtime to produce better batteries (that is, to analyze the performance of car makers)? I don’t remember that D. McFadden ever suggested to use TFP for that purpose but I may be wrong and someone else have suggested some way to do it.
2. What sort of conclusions would you expect to draw from a better understanding of the potency of the technological platform of batteries and lithium? Would you say that any discovery is a technological accident? Would you dismiss the new platform because there is not discernible impact on the world economy’s “aggregate” output? How many innovations prompted by new batteries and lithium would be required for a new industrial revolution?
3. Under what conditions would you agree that a significant reduction in the total costs of producing cars thanks to lower environmental costs not reported in car makers’ accounting books implies an increase in real wages even if nominal wages remain unchanged? To what extent do you think those conditions are met?
But real wages is the return to the labor factor. If we expanded income to include capital income, then all the returns to innovation should show up.
I suppose that this would only imply increasing mean wages, not increasing median wages, however. We know that returns to capital have not been the source of income inequality, so perhaps the dynamic here is (a small % of) labor capturing more and more of the returns to innovation and economic organization.
Total personal income as a percentage of GDP has gone up from 80% in 1970 to 86% today. The argument is that all the newly-created personal income is going directly to the innovators, not TFP.
As we become wealthier & more equal, more workers can handle more variance in their earnings, which is manifest partly in earlier retirement and more time as a student. Ironically, all these people in school or early retirement make the wage landscape look very unequal, and they lower the median wage. Your thought experiment can’t account for this.
People in a stagnating society don’t take on more debt and more educational investments. Yet, these trends have been in place for the last 40 years, at least.
Is it possible that median wages are stagnant because an awful lot of people don’t have the knowledge needed to do productive work in an increasingly information based economy? It seems like there’s a case to be made for automation of manufacturing outpacing education quality and consistency as a cause of stagnation. Sooner or later we’re probably going to end up with an economy where a lot of work is going to be in research and development, entertainment, medicine and other ultra high skilled professions.
There’s a greater scarcity of high-end skills than low-end skills and, as a result, we’ve had growth in income inequality.
The low-hanging fruit Tyler talks about in education were the middle and working-class white kids we turned into college material. We have utterly failed to educate low-income and inner-city kids — educational attainment plateaud as a result.
Median male wages are anything but stagnant. Globally they’re skyrocketing.
Median male white American wages are stagnant. Global convergence isn’t terribly surprising, it was supposed to happen by any economic theory.
“Globally they’re skyrocketing.”
- they’re currently harvesting their low-hanging fruit: turning rural kids with a grammar school education into high school and college graduates.
It’s not so low-hanging if it was hanging there for thousands of years and nobody picked it up earlier.
“Who would have predicted in 1972 that measured male median wages were going to stagnate and even possibly fall? ”
Who did predict this? The only people I know of were environmentalists who thought we were causing damage and using up resources in a way that would have by now caused a reduction in lifespan and quality of life.
Dynamics of Growth in a Finite World and the Population Bomb both warned of stagnation and worse. We now think these books predictions of resource and food costs were wrong. But could they have had something right?
No, not particularly. For most of us in the world, fears about stagnation in the early 1970s seem so hilariously misplaced that these books are actually very fun to read even today.
In fact, I’m astonished at the “God of the gaps” motions of people pushing the income stagnation myth. First it was household data that we should worry about. Then real median incomes. Now it’s white male median workers? Perhaps, in a few years, the income stagnation crowd will have found the gap of “white male heterosexual median industrial workers in Delaware between 2005 and 2007″. And then write books trying to explain this astonishing decline of the world economy.
This shows which they last very much lengthier and thus saving you income which could otherwise are actually utilized to purchase new ones.vg
Comments on this entry are closed.