Network News

X My Profile
View More Activity
Posted at 11:24 AM ET, 01/ 3/2011

Some thoughts -- and graphs -- on inequality and income

By Ezra Klein

I've been spending some time thinking about income inequality for a piece I'm writing. As part of it, I made this graph comparing trends in inequality -- as measured by the share of pre-tax income that goes to the top 1 percent -- and median household income (adjusted for inflation). This is, I think, the mental picture a lot of people have in their heads when they think about inequality:


The story this graph seems to tell goes something like this: From about 1947 to 1970, times are good for the average American and not as good for the average rich American. That's the part of the story that gets a lot of attention: It suggests that perhaps there's a connection between falling inequality and rising wages. But the timing of the next few inflection points doesn't really work for that theory: In the 1970s, median household income begins stagnating. But it's not until the mid-1980s -- and really beginning in 1987 -- that the income share of the top 1 percent begins skyrocketing. Then in the ’90s, inequality rises sharply, but so too do median household incomes. And then in the aughts, inequality rises, but median household incomes don't. There's not a consistent relationship between the two variables.

But that graph, which has informed my thinking on inequality, is, in some crucial ways, misleading: It compares changes in a percentage of something (the top 1 percent's share of income) to changes in the value of something. Here's a graph that's apples-to-apples: It tracks changes in the dollar value of median household incomes against changes in the dollar value of incomes for the top 1 percent (both data sets are adjusted for inflation). I've set the level both incomes were at in 1947 to be the baseline:


This graph tells a different story: Income inequality falls between the 1940s and 1970s because median household incomes are doing well, not because the incomes of the rich are doing poorly. As you can see, their incomes are growing too -- just not as fast.

In this graph, too, the story changes in the 1970s -- but it changes for both groups. Whatever hits median incomes in the 1970s also hits the incomes of the rich -- and harder. Where median incomes begin to stagnate, incomes of the rich begin to fall. But then, in about 1985, the incomes of the rich take off -- and never really slow down again. Stock market crashes create some volatility, but they don't keep incomes of the wealthy down for long. Whatever is driving the gains for the rich, however, isn't doing much for median household income, which never really speeds up again.

A few thoughts:

1) Income inequality and median income stagnation don't track: There's not much evidence here for the view that the top 1% are taking income gains that would've gone to the rest of us, or at least for the view that that's the primary driver of median income stagnation. Median household income begins stagnating in the early-1970s. The run-up in inequality really begins in the late-1980s. Further complicating the picture, income inequality shoots up in the 90s, and so too does median household income.

2) What happened in 1987? From about 1952 to about 1986, the top 1 percent's share of income fluctuates between 7 percent and 10 percent. But between 1987 and 1988, it jumps sharply -- rising from 10 percent to 13 percent in a single year -- and never comes back down. So what happened in 1987? There's a massive stock market crash that year, but it's not a crash that's considered to have had profound or lasting impacts on the real economy. Most explanations peg it as a market-driven, rather than economy-driven, event. And yet something that year does seem to have profoundly changed income equality in this country, and in a lasting way. But what?

3) What are the rich good for, exactly? There's almost no relationship between the fortunes of the richest Americans and average Americans. From the 40s to the 70s, the top 1 percent doesn't make big gains, but the average American does. In the 90s, they both do. In the aughts, the rich do great and median households go nowhere. There's a model of the economy in which the rich create the growth that sustains everyone else -- this is the model in which low tax rates on the wealthy are a crucial component of long-term economic growth -- but I see no evidence in either graph that that's how the economy actually works.

4) The story at the top appears to be financialization. There's evidence -- both on these graphs and elsewhere -- that the real driver of gains among the very rich is money from the financial sector. But I've never seen a really persuasive explanation for what exactly happened there. The time period roughly corresponds to deregulation, but it's also possible that deregulation roughly corresponds to the time period when the financial industry started making more money and throwing it around in Washington. So what's the best explanation of why the finance sector began playing such a different role in America's income distribution starting in the 90s?

5) The story in the middle is ... what? Tim Noah did a yeoman's job trying to answer that question in this series. One thing he didn't address is that, unlike the rise in inequality, median income stagnation has not been unique to the United States. Canada has suffered from it, for instance. That makes me skeptical of overly U.S.-centric explanations. One question that's never been answered to my satisfaction is why the link between productivity and income growth broke:


I'd love to hear other people weigh in on any or all of these.

By Ezra Klein  | January 3, 2011; 11:24 AM ET
Categories:  Economy, Inequality  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Column: America's other deficit
Next: Paragraphs that scare me


--*Income inequality falls between the 1940s and 1970s because median household incomes are doing well*--

Bzzzzzzt! That "because" is either a lie, or ignorance, or both, Klein.

Posted by: msoja | January 3, 2011 11:35 AM | Report abuse

Very interesting post, Ezra. Thanks. I do wonder, in regard to the last graph, how much of a role technology may have played in breaking the link between productivity and incomes?

Posted by: horacemann | January 3, 2011 11:36 AM | Report abuse

"And then in the aughts, inequality rises, but median household incomes don't. There's not a consistent relationship between the two variables."

When discussing "median household income" you should clarify if you just mean cash wages, or are including the dollar value of benefits such as health insurance.

One explanation I've seen previously is that cash wage growth has been stagnant because all income growth is being consumed by the increased cost of health insurance.

Posted by: jnc4p | January 3, 2011 11:39 AM | Report abuse

Well, 1987 is just after major tax reform. Also, as corporate executives' compensation switched from cash to stock awards, after their salaries above $1 million was declared be “unreasonable,” and therefore not tax deductible, the way corporations operated shifted, I think. (This happened in 82). There was a huge surge in mergers, and with it, a huge surge in downsizing. Also, the 80s is when the manufacturing sector's decline really accelerated, and that would have a major impact on median income, as men who'd been in high-paying factory jobs had to move into lower-paying service and construction jobs.

Posted by: mnlp | January 3, 2011 11:40 AM | Report abuse

Delinking income and productivity must have a lot to do with busting unions. And what about globalization and "free trade"? The standard economists' mantra that free trade and global wage competition does no harm to domestic jobs makes no sense, intuitively.

But the divergence of those lines is critical, and the biggest reason we have an upcoming problem with Social Security financing. Rising productivity should amply fund any drop in worker to retiree ratios. But not when the dividend is being preempted by employers at the expense of employees.

Behind it all is the steady drumbeat of right-wing propaganda since about 1970, which has so far succeeded in distracting the frog from the steady increase in the ambient aquatic temperature.

Posted by: jtmiller42 | January 3, 2011 11:45 AM | Report abuse

Very interesting stuff. Ezra, I was thinking about your "productivity and median income growth over time" graph. I suggest that the median family has seen most of their "income" growth thru intangibles that are very difficult to graph/measure. Hours worked per week, life satisfaction, access to new technologies.

What I'm getting to is: what if the massive increase in financial sector income and "median families" is a massive divergence of what carries value itself. Cheers

Posted by: Will_W | January 3, 2011 11:45 AM | Report abuse

Perhaps the key is that the economy has become increasingly global and that any analysis that limits itself to the USA while seeking trickle down impacts is too narrowly focused. The wealthy are certainly investing, which is generally the source of their increasing wealth, but their investments are global and therefore the impacts of those investments are global. I think the stagnation of wages in the US are related to the outsourcing of production overseas. Just a few thoughts.

Posted by: wvng | January 3, 2011 11:47 AM | Report abuse

I think we should look at what happened to the medium income as opposed to segregating the top 1%. You're looking in the microscope the wrong way.

Posted by: RisingTideLiftsAllBoats | January 3, 2011 11:55 AM | Report abuse

From the book "AN AUTISTIC WORLD (1)"

One of the usual inferences from inequality, is the recognition for the need of a more equilibrated distribution of wealth, which in turn would produce a better society. That is only partially true. Dividing wealth among individuals is not a bad idea, but isn’t a fair idea either because doesn’t value the different efforts of those individuals to acquire prosperity. What is needed is the realization of common sense. A society that has thirty or forty percent of its population under an economic struggle is a society in trouble, even if it doesn’t count correctly the number of people unemployed and disregard the health concerns of those individuals. It would be a matter of time until they will ask “why us, and not you,” in which case two options could be given: one is to publicly admitting that the current system is effective only for part of our community; the other is to transform those individuals into living zombies in order to ignore their reality.

Posted by: kanino | January 3, 2011 11:56 AM | Report abuse

It's possible for those parts where there doesn't seem to be a connection to be hidden by some other number. For example, if GDP is going way up, it's possible for rich income to jump and median income to still rise, just not as much as it would if the GDP gains were evenly distributed. Maybe a graph of median and top 1% income vs. per capita GDP, or fraction of GDP to each income group? That would capture some of that broken productivity-income link.

Posted by: _SP_ | January 3, 2011 11:58 AM | Report abuse

I'd really like to see what the graph of
median family income looks like if you
subtract one or both of median out-of-pocket
healthcare costs, and median housing costs.

Neither one will make a dent in the top-1%,
but I suspect that when you subtract out
the healthcare costs which rose rapidly
after 1980, you'll see a truer picture, that
the median family is substantially worse off
than in the 1970s. Which ought to be very

Posted by: richardcownie | January 3, 2011 11:59 AM | Report abuse

The vast changes in income inequality coincide with the introduction of 401Ks. The 401Ks pumped huge sums of money into the stock market but left the individual investor devoid of power.

Something to think about when viewing proposals to give individuals vouchers to purchase their own healthcare.

Posted by: HealthcareAnnotator | January 3, 2011 12:06 PM | Report abuse

PS: Graphing the actual trends in productivity and earnings (vs. growth) shows an even more distinct divergence, in about 1973 -- a real fork in the road. (Shown during Real News Network interview with Yves Smith, 12/26.)

Something happened at that point. It should be possible to isolate what it was.

Posted by: jtmiller42 | January 3, 2011 12:11 PM | Report abuse

You're graphing two quite incomparable figures. Median family income is in dollars; income share is in percent.

The question is that income growth mostly went to the top income groups -- not only the top 1%, but incredibley much to the top 1% -- during the Bush years.

Posted by: fpfrankpalmer | January 3, 2011 12:14 PM | Report abuse

Really, there just isn’t much of a relation between the two.

The narrative for the upper 1% is that their income is just a function of the market. The market did basically nothing for 15-20 years before 1981, and the rich were stagnant; it rose fairly consistently for almost the next two decades, and they did well. Their income sees drops when/where the market drops. End.

For the middle class/median, the “petroleum price” narrative handles post-war America fairly easily. When the price of petroleum has been steady or falling, the median has done well (up until the 70s, then again from the early 80s through ~2000 with the exception of an oil shock circa 1990). When oil prices jumped up, e.g., the 70s, the 1990ish shock, the naughts, the middle class did poorly. This is all expected in a petroleum based economy when you focus on ‘inflation adjusted income’; the ‘inflationary’ effect of a petroleum price shock will have price growth outpace earning growth, and the inflation adjustment will therefore show falling (or stagnant) earnings.

Posted by: eggnogfool | January 3, 2011 12:18 PM | Report abuse

I have my own speculations as to what is driving this, but the real question is - Can Obama reverse or slow Reaganomics? That's what we were all hoping for. We're essentially living in Ronald Reagan's eighth term as President.

Posted by: willows1 | January 3, 2011 12:21 PM | Report abuse

Neoliberalism happened, both as social and economic policy, pushed largely by a Utopian elite. I cannot recommend David Harvey's work highly enough here. "The Enigma of Capital and the Crises of Capitalism" and "A Brief History of Neoliberalism" are first rate explanations of what you're struggling with here.

Posted by: joemomma1 | January 3, 2011 12:23 PM | Report abuse

Is the increase in productivity due to more productive technology or more productive workers?

Posted by: ideallydc | January 3, 2011 12:24 PM | Report abuse

Looks like this is heavily tied to the stock market.

Many companies went public to finance founders/partners wealth during retirement. Private workers don't have the luxury of pensions anymore, so by going public (or having a similar liquidation event), it creates wealth and liquidity for them.

Look at facebook today, they just had a $2billion liquidity event from Goldman and DST. Zuckerberg is now worth $15billion, but this has nothing to do with "Policy," more on how people are now structured to make money and gain liquidity.

Posted by: marteen | January 3, 2011 12:30 PM | Report abuse

There were big changes in the tax code in 1986, perhaps that had something to do with the 1987 divergence.

I found this and I have no idea how accurate it is, but it does seem to provide some limited information.

Posted by: Skowronek | January 3, 2011 12:30 PM | Report abuse

@willows1: Amen. But the REAL question is Does Obama want to undo Reaganomics. I don't think Larry Summers does.

Posted by: jtmiller42 | January 3, 2011 12:31 PM | Report abuse

You pose some good questions here. What is not taken into account in your graphs though is how many people are in the middle to rich income gaps. That can have a effect on the numbers.
As we saw in the census data, out population has grown a lot since the 1950's. This makes it harder for people to find a job, and the world itself becomes more competitive. That too, can have a impact on families staying in poverty, and the rich who were able to invest in companies that were/are doing well and earn more from the market.

Posted by: roseonpolitics | January 3, 2011 12:34 PM | Report abuse

As others have pointed out, I doubt there was an event in 1987 which immediately cause a spike in income for the rich. My guess is that several events from the preceding few years had their first real effects felt in 1987.

As for the delinking of productivity and income growth, I think technology and the collapse of the unionized manufacturing sector are the most likely culprits. People went from good paying jobs with regular raises to being replaced by automation and getting non-union jobs with no regular raises and worse and worse benefits.

Posted by: MosBen | January 3, 2011 12:39 PM | Report abuse

"and really beginning in 1987 -- that the income share of the top 1 percent begins skyrocketing."

i worked on wall street, for a big stock brokerage firm, in 1987, and it was the perfect time for the rich to grow richer, and many things were changing in the stock market.
there were all different and new, creative offerings for the richest investors, and most people didnt understand much about them.

i worked for the head of the tax shelter department...and there was no shortage of wealthy clients, or offerings in everything you could imagine.
and then, came new, glamourous operations, catering to the wealthy clients, in brand-new "hedge funds," and "commodity funds."
no-one knew exactly what they were, but the rich folks were buying them.
and of course, came the junk bonds.
there were lots of new creative ventures for rich people, to make them richer, and maneuver with their taxes.
everything seemed to be changing, and in favor of the wealthy investors...who had money for speculation, and the required net worth, in order to participate, in the first place.
these opportunities were not open to everybody.
rich people with account executives were buying stock in computer firms, in gold, in early waste management companies, video companies....there seemed to be so many new opportunities that "small people" didnt even understand very well, and in order to get involved in these kinds of hedge funds and commodity needed to have a high net worth. so these opportunities were only for the rich, to begin with.
everything seemed to change in the market at that time, and it certainly was a chance for the rich to grow richer.
that was exactly 1987, as i remember.

Posted by: jkaren | January 3, 2011 12:47 PM | Report abuse

Some things to consider
Income tax changes (1987)
Capital gains tax rates
Hedge funds (uncommon before 1985)

The income arms race - consider two comparisons:

if Bill Gates had a $1B in 1987 why was Jack Welsh worth so little (at that point).Yes, Microsoft was a great idea and extremely successful, but Welsh was head GE, one of the greatest companies in the world which was many times larger.

In 1965 some presidents of fortune 500 companies made only $100K, so did Hank Aaron, and many investment bankers, Wall Street lawyers, doctors and small businessman. If you were president of a fortune 500 company did this seem fair?

Posted by: sd16 | January 3, 2011 12:53 PM | Report abuse

The missing element that explains why American incomes are stagnate and the top 1% is increasing is Outsourcing.

The primary beneficiary of oursourcing is not only the stockholders but also the senior management of companies which do outsource.

The incomes of the average americans have been redistributed upwards as more graduates of elite business schools learn how to shift earnings from one group upward to another. The source of all this can be found in those business schools who first invented outsourcing.

Posted by: EducatingTheFools | January 3, 2011 1:14 PM | Report abuse

First, this post made my toes curl. Now…

Your second graph appears to have correlation with S&P 500 (would need to do the math to make sure):

I thought I would find huge changes in capital gains since this often how the richest of the rich make their money (Buffet), but the historical rates don’t match up with your graph: .

It is also possible since another large percentage of the 1% are lawyers and doctors who own small businesses that there it is a change in corporate tax rates (I will look that up later).

Ezra said, "There's almost no relationship between the fortunes of the richest Americans and average Americans."

I am not sure I entirely agree with you there. There is definitely Pareto Efficiency - i.e. can have both rich and middle class incomes increase without any leaching – but when you figure in national income and taxes the picture becomes different. When we lowered rates on the 1% it meant that a greater percent of Tax revenue was coming from the middle class (see CBO paper): – see “share of federal tax liabilities” etc.


Posted by: chrisgaun | January 3, 2011 1:17 PM | Report abuse

The "income" in this chart is only taxable income; it doesn't include unrealized economic income.

I bet the spike in 1987 is attributable to the reduction in marginal tax rates. High net worth taxpayers recognized their previously unrealized gains in 1987 - they were betting (correctly) that marginal rates would increase again.

The omission of unrealized income from this chart means that most of the wealth of the Forbes 400 isn't reflected at all (despite the fact that they can borrow against that wealth and spend it).

Posted by: DavidSMiller1 | January 3, 2011 1:17 PM | Report abuse

ugh, writing too fast... I said "greater percent of Tax revenue," which should say something like "Middle class now shares greater percent of tax liability burden."


Posted by: chrisgaun | January 3, 2011 1:22 PM | Report abuse

"One question that's never been answered to my satisfaction is why the link between productivity and income growth broke"

I think part of this discrepancy is that for many workers, there really hasn't much in the way of productivity gains over the past generation, and lots of jobs were created in these sectors.

A McDonald's worker in 2005 was probably more or less as productive as a McDonald's worker in 1975. The same probably goes for many types of jobs - retail store clerk, construction, restaurant waitstaff, etc.

Consider an economy with a manufacturing company and a service company. The manufacturing company produced 1,000 units in 1975 with 100 workers, and due to automation it produced 1,500 units in 2005 with 15 workers. In 1975, the service company had 100 workers who produced 1,000 units, and it had 185 workers producing 1,850 units in 2005.

So in both years there are 200 workers (100/100 in 1975, 15/185 in 2005). Total output rises from 2,000 (1,000 and 1,000) to 3,350 (1,850 + 1,500). The government statiticians will divide GDP by the number of workers, and find output per worker grew from 20 to 33.5 from 1975 to 2005, a 67.5% increase.

However, the service workers are only producing 10 units of output (the same in 1975 and 2005), and the remaining manufacturing workers are executives and highly skilled technicians with six figure salaries, and are producing 100 units of output each.

So you won't see the wages of the median worker (who works at the service company) rising with measured productivity, and you'll see the people who are left running the manufacturing company doing extremely well, leading to increased income inequality.

Posted by: justin84 | January 3, 2011 1:24 PM | Report abuse

I think it is very simple. The top 1% wealth creation model reached tipping point. Next will come some sort of disruption as a natural consequence. That will be painful for the middle class and devastating for the wealthy.

Posted by: BertEisenstein | January 3, 2011 1:46 PM | Report abuse

Along with DavidSmiller and others, I'd suspect the tax reform act of 86. One thing I didn't notice being mentioned: it's possible that the lower tax rates brought wealth and income in from the cold, so to speak. In other words the rich had more incentive to evade taxes before 1987 so the jump in the 1 percent share is more of a statistical artifact than reality. If inflation, such as we had in the 70's is bad for wealth-owners, the 1 percent share may have really started its upward curve in the early 80's.

Posted by: bharshaw | January 3, 2011 1:47 PM | Report abuse

One factor in the 70s could be inflation. Top 1%ers probably had portfolios with bonds that lost value relative to inflation, while the average American probably saw their wages roughly keep up with inflation.

Posted by: doctored | January 3, 2011 1:54 PM | Report abuse

During the period of 1947 to 1980 most "income" of the wealthy was reinvested into business to shelter it from taxes. What you need to look at more likely is how the net worth of the top 1% grew in comparability to the middle class net worth during the same period.

It will be more difficult but far more meaningful. The 50'ies and 60'ies were great years for the upper classes. Maybe you can look at Cadillac sales compared to Chevy's growth in sales.

Posted by: AprilC | January 3, 2011 1:56 PM | Report abuse

Ezra asks: "why did the link between productivity and income growth break?" The answer is computers and automation. It used to be that an increase in productivity was tied to humans working harder and smarter, so when productivity increased the workers were compensated for their effort, But then computers and automation came along and machines were now responsible for increases in productivity, so there was no need to pay workers more. In fact you could pay them less, because they did not have to work as hard or smart.

Posted by: cummije5 | January 3, 2011 2:04 PM | Report abuse

This is obviously the question of the day, without doubt. (And by day I mean quarter century). Thank you for focusing on it.

I wonder, however, what your graphs would look like if you were to look at individuals rather than families. Since one of the largest changes in society over the past hundred years has been the requirement that the average family have two income earners.

I also second the notion that life's primary costs be included -- petroleum, food, housing and health care.

Posted by: comma1 | January 3, 2011 2:07 PM | Report abuse

Ezrak: You should read the full analysis, not just the well known graphs, by Emmanuel Saez and colleagues (see at: and then scroll down). But basically the story is a combination of what others here have said: a series of changes in taxes during the 12 years of Reagan/BushI that decreased income tax for wealthy, increased the difference between earned and unearned/capital gains tax, financializtion, increased ability of rich to set their own incomes (salry, bonues, stock options, etc), increased total taxes (payroll deduction/SS/Medicare, real estate) on working and middle class (including earned income salaried upper middle class which is where the resentment of the 100-250,000 earners come in), rise in cost of health care and insurance, etc.

Posted by: drstevea | January 3, 2011 2:30 PM | Report abuse

Several factors changed dramatically during the later time period that could all have contributed: massive increases in the speed of technology transfer and world-bank investment to low and middle-income countries, combined with expansion in global trade and tax policies that rewarded "off-shoring" of jobs and punished the Unions, combined with an inflated dollar and persistently low currency values in fast-growing markets---all of these could have benefited people in the United States at the top of the heap, while simultaneously disadvantaging people who occupied the middle.

Posted by: devrex | January 3, 2011 2:41 PM | Report abuse

Financialization: most of the "innovation" consisted of ways that big money could bet "safely" (CDOs etc.). Money won through wagering represents exactly nothing in the underlying economy. Except you can use it like it's the real thing.

So the rich grew super-rich on side bets and then held the economy hostage when the bets came due.

Posted by: GBMcM | January 3, 2011 2:45 PM | Report abuse

It definitely seems like a combination of tax code changes combined with the effects of health care, education and other rising costs.

Maybe a way to graph would be a baseline for income changes or income within selected bands (middle income quintiles, maybe and top 1% or top 5%) as well as lines that track health care, higher education and tax burdens.

If the super-wealthy were accruing more money, the rising costs of healthcare and education and wouldn't have affected them as much as those in median income bands where the costs were rising faster than the wealth they were accumulating.

Posted by: dubrowg | January 3, 2011 2:49 PM | Report abuse

I think I can help answer question #5. Since modern economics defines productivity simply as ratio of output (GDP) per labor hours worked, the fewer labor inputs you need would lower the aggregate standard of living over time. Put another way, you simply dont need as many people to do the same tasks as before.

Just take the microeconomic view of a modern small business. They can now use quickbooks to replace an accountant, salesforce to replace a secretary, turbotax to do their taxes, gmail to manage their correspondence, etc, etc. It makes sense that software companies have done so well in the period you are questioning - they have helped replace thousands of workers for better or worse. And look at the fortunes made in that space - I think it would account for a lot of that gap you are looking at.

Posted by: staypuftman | January 3, 2011 2:51 PM | Report abuse

The link between productivity and median income got broke because big corporations and equity owners don't rely on domestic labor to generate profits anymore. The global American corps are increasingly distancing themselves from American workers and the American national economy. It's really not all that difficult to understand if you quit watching TV news and read between the lines of the print media - even for a bonehead accountant like me. Corporations don't pay for productivity, by the way. They never pay more than they have to because their single objective is profit accumulation. Your second graph is a stark depiction of Reaganomics. The first graph is meaningless, but I'm guessing it may reappear somewhere to support someone's argument that the top 1% is simply catching up to where it should be and/or that all the chatter from the Left about inequality is a myth.

Posted by: craigemoen | January 3, 2011 3:07 PM | Report abuse

If Ezra can read some books, then in addition to the data and analyis previousl cited at perhaps he should read and think about Pierson & Hacker's Winner Take All Bartel's Unequal Democracy and Krugman's Conscience of Liberal for some history and economics and poli sci. By the way, the first graph in Kleins essay is sooo stoopid, one can only presume it is a straw dog meant to be knocked down by himself in the second graphic. Again, Saez's graphs pretty much tell one what you need to know.

Posted by: drstevea | January 3, 2011 3:54 PM | Report abuse

Ezra: great work. Please follow this where it leads (it may take a year or two). Consider what a Hans Rosling presentation addressing this would look like. That's where you want to get.

- Productivity gains from computers didn't really kick in until the 90's (after lots of spend in the 80's) after technology starts getting integrated into and changing workflows. Sorry, no cite.

- Re the productivity/income growth link: I'd wonder if growth of the equities markets (including did the democratization of markets by 401k's) meant a larger ownership class and increased demands for earnings to be delivered to the shareholders rather than reapportioned to labor.

- Generally, financialization & computerization both enable a very low friction environment for money movement. In these environments, the peak performers are rewarded in outsize ways.

Sorry no time to comment further.

Posted by: BHeffernan1 | January 3, 2011 4:25 PM | Report abuse


Please add two more lines to your graph -- the percentage of top 1%'s income that was taxed, and the balance of trade.

That tax (which was high in the 40's-80's) was spent by the government on infrastructure and American goods and services, which meant rising incomes for American workers.

The balance of trade is your inflection point for the early 70's, when US income for both the rich and the median was exiting the country.

Posted by: grhabyt | January 3, 2011 4:35 PM | Report abuse

Regarding the final graphic, union membership peaked in 1975 and declined thereafter. Union political power waned during the 1970 and fell continuously afterwards. Lots of bad things to say about unions, but they put a priority on the link between wages and productivity.

BTW, median household income hides some interesting statistics. It increased after 1973 only because women joined the workforce and also gradually reduced the gender pay gap. Male workers' hourly wages haven't grown from the 1973 peak. (Elizabeth Warren once co-authored a book about this.)

In interpreting the orange lines, I'll note that slopes seem to have changed circa 1981. And we know what happened then.

Posted by: pjro | January 3, 2011 4:40 PM | Report abuse

OH! Here is part of the picture. Although you are considering pre-tax income it is still important to look at taxes because if the 1% had more disposable income then that certainly would go into investments (increasing future income). Here from 1986 - 1987 the top marginal income tax rate went from 50% to 38.5%:


Posted by: chrisgaun | January 3, 2011 5:04 PM | Report abuse

HI Ezra,
Mark Perry, who runs the Carpe Diem blog, may have a lot to offer to assist you in understanding the discrepancies you are wrestling with. Fair warning: you and he are not on the same side of the inequality argument, but I think his work is generally useful in framing the argument (albeit from a more conservative, markets-based view).

Posted by: bzod9999 | January 3, 2011 5:56 PM | Report abuse

As others have suggested, it might be worth looking at the tax rates, particularly to explain the boom for the top 1%.

See this chart ( ) for a graph of your data alongside the top income and capital gains rates. Rather than 1987, the key moment for the top 1% seems to be in the late 70s or early 80s, when the top tax rates plummeted.

Posted by: Ulium | January 3, 2011 9:51 PM | Report abuse


Between 1980 and 1990, the number of credit cards more than doubled, credit card spending increased more than five-fold and the average household credit card balance rose from $518 to nearly $2,700. With the cost of money sinking and average balances climbing, profits soared.


Posted by: Mazzi455 | January 3, 2011 9:59 PM | Report abuse


Between 1980 and 1990, the number of credit cards more than doubled, credit card spending increased more than five-fold and the average household credit card balance rose from $518 to nearly $2,700. With the cost of money sinking and average balances climbing, profits soared.


Posted by: Mazzi455 | January 3, 2011 10:00 PM | Report abuse


Between 1980 and 1990, the number of credit cards more than doubled, credit card spending increased more than five-fold and the average household credit card balance rose from $518 to nearly $2,700. With the cost of money sinking and average balances climbing, profits soared.


Posted by: Mazzi455 | January 3, 2011 10:01 PM | Report abuse


Between 1980 and 1990, the number of credit cards more than doubled, credit card spending increased more than five-fold and the average household credit card balance rose from $518 to nearly $2,700. With the cost of money sinking and average balances climbing, profits soared.


Posted by: Mazzi455 | January 3, 2011 10:02 PM | Report abuse

Another factor the might be affecting the productivity vs wage graph is the constant change to the way GDP and inflation are calculated. Every time a change is made, such as ignoring energy and food increases in "official" inflation, inflation looks a little lower, and GDP seems a little higher. That's why the Fed is scrambling to create (official) inflation when we know that prices are already going up.

Posted by: ProfElwood | January 3, 2011 10:29 PM | Report abuse

Ezra your racism is showing. Is it based on empiricism or ideology?

Race, class, gender: Trifecta!

Posted by: soma_king | January 3, 2011 11:17 PM | Report abuse

I see the economic equality issue as follows.
#1 Historically, growth in incomes amongst the rich is correlated with economic volatility and crashes.

#2 Economic volatility hurts the income growth amongst the poor and middle class. This is a historical relation. That said, I imagine that it is because economic volatility seriously undermines workers' ability to negotiate.

#3 Economic volatility is not the only cause for poor wage growth. In the 1970s, economic performance was poor because of poor monetary policy. This hurt workers for a different reason.

#4 If the 1970s is an aberration, then wage growth anti-correlation between poor/middle class and the rich is much stronger.

#1 This sort of comparison can be very sensitive to

Posted by: zosima | January 3, 2011 11:53 PM | Report abuse

Whoops left over bit of an edit got clipped at the bottom. I was going to say that this sort of comparison can be very sensitive to how the wage categories can be defined. I'm not sure if this would change the results, but it is worth keeping in mind.

Posted by: zosima | January 3, 2011 11:55 PM | Report abuse

Please include health and pension benefits in your numbers.

Posted by: staticvars | January 4, 2011 12:05 AM | Report abuse

Why don't you correlate the top 1% income with the S&P 500 average? It would probably be more meaningful as that is a closer approximation of where the money is.

Posted by: staticvars | January 4, 2011 12:07 AM | Report abuse

I just heard it on the radio that "Wise Health Insurance" can offer health insurance for just $1 a day any one aware of this ? have anyone purchased insurance through them. I did search for them and found them online.

Posted by: tracystewart03 | January 4, 2011 2:01 AM | Report abuse

Three major reasons middle class stagnated, as the rich have soared:

1) Obviously, tax rates on upper income, for two reasons: first, they pay less. Second, there was less incentive to roll earnings back into the business (which creates jobs and tightenes labor markets). When top tax rate was95%, there as a real incentive to do so.

2) "Free" trade policies: This is a huge weapon for employers to drive down wages. (Anyone with a family member in manufacturing knows this.) Even when jobs don't leave, the threat alone is enough to keep the workers in check. As one ivy leaguer once explained, "The perception of risk alone is enough..." In addition, it has led to manufacturing jobs (which pay well) to be replaced by service jobs (which don't). A terrible trend.

3) A successful anti-union campaign. Leverage determines wages, not the value of labor. By limiting workers leverage, you limit their wages.

The 90's were 'good times' that broke the trends of the past 30 years, but only because the bubble had unemployment so low (sub-4%) that some workers managed to have leverage on their own. (Though most still did not -- some had enough gains to offset the continued decline for HS educated workers.) But, policy makers are insane (like a fox?) if they think our labor market will normally have a tight enough labor market to allow such gains for workers. It's only happened once in our history, and it won't be the norm in the future. There is a reason they call it a 'natural rate' of unemployment.

Here's the miserable reality, since right-wing economics took hold:
1) Most families are contributing another worker to the workforce,
2) working longer hours,
3) have less generous health benefits,
4) lost their pensions,
5) while the average worker has a much higher level of education today, and
5) they haven't seen real income gains --

All while productivity and profits have soared.

I think it's time to end these awful policies. But, money controls the game. And, money likes them.

It's about power, and power has been taken from workers systematically -- both in the workplace and in our political system (see Citizens United). It's a shame that workers continue to be so docile.

Posted by: rat-raceparent | January 4, 2011 11:17 AM | Report abuse

Should have added another factor... Defense spending.

Look at the defense budget: $1-1.5 trillion a year. That's a tax of $3,333-5,000 per head (not per taxpayer) every year, and a large portion of it runs to the large firms whose profits are distributed to stock owners. Figures based upon

Now, there's a lot of items in the defense budget that are... defensible.

But, our congress keeps ordering equipment that the military doesn't even want! There's the fat, but we're not allowed to talk about that, huh? Why? Money rules that game, and those beneficiaries have money alright.

Posted by: rat-raceparent | January 4, 2011 11:36 AM | Report abuse

"Obviously, tax rates on upper income, for two reasons: first, they pay less. Second, there was less incentive to roll earnings back into the business (which creates jobs and tightenes labor markets). When top tax rate was95%, there as a real incentive to do so."

So that you could have more income to be taxed at 95%? No. A tax rate of 95% provides no incentive to invest in a business.

Empirical evidence bears this out.

From 1950-1962, business equipment investment was 5.0% of GDP on average.

After the Kennedy tax cuts, business equipment investment was 6.9% of GDP on average (1963-1980).

After Reagan and the era of lower tax rates, business equipment investment was 7.9% of GDP on average (1981-2009).

So investment in business equipment represented nearly 40% more of GDP after Reagan's election than during the Eisenhower/early Kennedy era.

Of course, you can't ascribe all of the increase in equipment investment to tax policy, but the numbers don't seem to support the notion that very high taxes lead to more business investment.

""Free" trade policies: This is a huge weapon for employers to drive down wages."

Free trade also drives down prices for the (vast) majority of Americans who don't work in manufacturing. It also drives wages sharply up for workers in other countries who we trade with - whom we should care more about because they are poorer and more limited in opportunities than average Americans.

To the extent free trade drives down wages, it is because Americans workers are paid more than the value of their output.

Almost no one advocates for trade barriers between California and Illinois or San Fransisco and Sacramento, yet they do between California and Japan and Detroit and Seoul.

"In addition, it has led to manufacturing jobs (which pay well) to be replaced by service jobs (which don't). A terrible trend."

Most of that change is due to productivity growth rather than outsourcing. The proportion of labor in manufacturing has long been declining in the industrial world, even as output has been on a long term rise.

"A successful anti-union campaign. Leverage determines wages, not the value of labor. By limiting workers leverage, you limit their wages."

Why are North Koreans, Cubans and Venezuelans paid so poorly, particularly when compared to American workers?

Why do managers and engineers earn far more than unionized grocery store workers?

When unions try to push wages above the value of labor, you get GM. Why did GM invest so much in SUVs and neglect small cars? Look at the margins that GM could get on small cars with UAW labor.

Posted by: justin84 | January 4, 2011 12:51 PM | Report abuse

Great analysis - a lot of unanswered questions..

Two Points:

1.Although median income stagnation has not been unique to the US - the argument in Hacker & Pierson's book (Winner-Take-All-Politics) is that it was much less in other countries and the driving force behind what stagnation occurred in those countries was that executive pay had to compete with American companies. As further evidence they linked the greatest percentage of wage stagnation to English speaking countries - where they had to compete more directly with America for executive talent.

2. What happened in 1987? I'm not sure, but my age is showing here: I graduated from college in 1987. I remember a phrase from that period - and it went something like "You can't go to heaven until '87." I don't remember specifically what was being referenced, but certain changes in the inheritance tax laws are implied. There may have been other tax breaks and changes in the tax code that kicked in at that time that I am not aware of.

Just my $0.02....

Posted by: RMGHicks | January 4, 2011 4:12 PM | Report abuse

Drop the outliers. Add a line for the top 1 percent minus the top .01 or .001 percent.
This will give you a better look at true income inequality. The multi-multi-millionaires should be considered but they may be distorting the graphs. I'm sure the Post can come up with the data.

Posted by: RabelRabel | January 4, 2011 5:49 PM | Report abuse

Post a Comment

We encourage users to analyze, comment on and even challenge's articles, blogs, reviews and multimedia features.

User reviews and comments that include profanity or personal attacks or other inappropriate comments or material will be removed from the site. Additionally, entries that are unsigned or contain "signatures" by someone other than the actual author will be removed. Finally, we will take steps to block users who violate any of our posting standards, terms of use or privacy policies or any other policies governing this site. Please review the full rules governing commentaries and discussions.

characters remaining

RSS Feed
Subscribe to The Post

© 2011 The Washington Post Company