6 ms·

Income Inequality and the 'Superstar Effect'

79 points by unignorant 16y ago

lkrubner 16y ago
This article is a bit innocent. It does not examine how the big tax cuts of 1982 changed incentives for everyone. There is a considerable amount of research that suggests when income tax on top rates is high, workers prefer non-cash benefits, rather than money. Recall that in the USA, during the 40s and 50s, the top income rate was 91% on all income over $100,000 (roughly a $1 million in todays money). Because of this, most top stars did not seeks payment in excess of $100,000. Instead, they sought benefits in to go along with their $100,000. The big tax cuts of 1982 lowered top rates to 36%. This made it much more rational for top workers to ask for high amounts.
- TheBlack_knight 16y ago
  That's a relavent point. High marginal tax rates would create the incentive for people in the higher taxed brakets to seek non-pecuniary benefits. Things like company cars and free jets. This would suggets we have an incomplete picture form this article.
- jdp23 16y ago
  Excellent point. The article's based on a time that's been better for the rich than ever before in many different ways -- and reduced the spoils available to everybody else. You have to wonder how sustainable that is ...
patio11 16y ago
I think PG had an article about this: if you have any process with a) a distribution of skill which b) scales with addition of resources, market price at the top of the distribution should explode because talent becomes a license to print money. Banking only got to that first because they have a very simple model: decisions for many types of assets take equal difficulty to get right regardless of how much money is stake (below a ceiling in the hundreds of millions or billions). If you can reliably make the average + 2% trading $2,000 of Coca Cola at a time (and you can't, but work with me here), then you can reliably make the average + 2% trading $200 million of Coca Cola. Comparing your wages to that of someone who trades hours for dollars no longer makes sense. Quants had a similar compensation scheme for quite a while now. I think other classes programmers are going to hit on it, too: there are many things you can do which demonstrably improve revenues or cut costs and which scale to the moon. Can I again toot the horn for A/B testing, metrics, etc etc? My homework for tomorrow is writing a consulting proposal. They want to know how much purchasing my services stands to make them. Consider a hypothetical company with 9 figures of revenue a year which does no A/B testing. Pick a number for what you think you could do for a first cut of conversion optimization. 2%? 5%? 10%? Do the multiplication. You can create staggering amounts of wealth these days with programming and a few other tricks, and you can capture quite a bit of it. [Edited to add: PG link: http://www.paulgraham.com/wealth.html http://www.paulgraham.com/wealth.html My favorite essay of his, incidentally. The relevant bit: To get rich you need to get yourself in a situation with two things, measurement and leverage. You need to be in a position where your performance can be measured, or there is no way to get paid more by doing more. And you have to have leverage, in the sense that the decisions you make have a big effect.]
- lkrubner 16y ago
  There is nothing wrong with the model PG puts forward, as a start, but it is a simple model and it needs explanation. At the least, you should make some attempt to explain why things change over time. Consider radio in the USA, up until the mid 1950s. It was eventually replaced by TV, but in the 30s, 40s and 50s it was the main medium of mass communication. The nation lived by its radio. The potential audience for a radio star was greater then than it is now. So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented? Changes in the tax code are important to consider. As a thought experiment, imagine if the top income tax rate was still 91%. (During the 30s, 40s, 50s and early 60s, it was 91% on income over $100,000, about a $1 million in today's money.) How many stars would ask for income in excess of $1 million if they knew it would be taxed at 91%? They would probably instead ask for various perks, which was the trend in the US during the mid-20th century. Asking for perks instead of cash remains the trend in those nations with strongly progressive tax rates, for instance, Sweden.
  - patio11 16y ago
    So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented? I did a double-take here, because it sounds fairly obvious to me: monopsonies can get away with paying you next-to-nothing because your best alternative to "next-to-nothing" is "nothing". If there are only a few radio networks (and/or they have a gentlemen's agreement with respect to hiring talent), then regardless of skill the commercial value of your ability is zero unless you sign with them. Rush Limbaugh would be making $20k a year if there was only one radio station in town. There are now 50 and the one that gets to sign him makes a substantial amount of money. (He tapped an unlikely niche market for news and opinion: half the country.) The market now favors him rather than favoring them. Edited to add: By the way, it appears that Google et al have a gentleman's agreement with respect to hiring talent.
- dspeyer 16y ago
  I've always been rather suspicious of that essay. He says: Making wealth is not the only way to get rich. For most of human history it has not even been the most common. Until a few centuries ago, the main sources of wealth were mines, slaves and serfs, land, and cattle, and the only ways to acquire these rapidly were by inheritance, marriage, conquest, or confiscation. Naturally wealth had a bad reputation. Strongly implying that in the modern world most large fortunes come from making wealth, but presenting no actual evidence. From observation, most large fortunes come from trickery: either in markets or by being hired as a CEO and negotiating high, non-measured compensation by abusing conflicts of interest. I haven't studied the question carefully, though. Has anyone? He also strawmans the pure-pie model. The real pie model looks something like this. A very large group of people collaborate to create something of value (even in software, there are library authors, hardware makers, ISP maintainers...). Some are part of the same company. Some are tied in via contracts, shaped by negotiators. Once the good is made, the wealth is divided up among the team. (Simultaneously, the team works on the next hunk of wealth -- it's a pipeline, not an alternator.) It is meaningful to look at the distribution phase in isolation, with the caveat that how the distribution phase impacts the production phase is a factor in how good the distribution model is.
joe_the_user 16y ago
While logically explained, the process seems rather unstable. Where will it end up?
lionhearted 16y ago
Sooner or later, I'm going to get around to writing a piece and titling it, "I reject equality" where I lay out the argument that the goal of equality is fundamentally hostile to excellence. In fact, I think one of the biggest flaws in English right now is how we use the same word - equal - for concepts as diverse as law and economics. "Unbiased courts" and "taking things by force from people performing better" really, really, really shouldn't use the same word. They do currently - equality before the law, and income equality. Facts-based-unbiased-law is obviously a good thing. Take-by-force-from-best-performers, on the other hand... Anyways, I haven't written the "I reject equality" piece because of how touchy it'll be, and how I'll have to ultra-carefully craft the arguments to say exactly what I intend to say, and no more or less than that. But I do think a goal of total equality is fundamentally hostile to excellence - excellence, by definition, is better than the rest. At some point, you have to start pushing down on the best if you want everyone to be the same. I understand some people will disagree with me on this, but I personally think that's a really bad thing.
- joe_the_user 16y ago
  You know what's ironic about this reaction of yours to the article? What's ironic is that a world of fewer and fewer superstars making more and more money is a world in which there is more and more equality - in that all the others are equally in shadow of the few superstars. My Freshman philosopher professor told a parable. One despot asked advice from the another. The second despot lead the first to a field and proceeded to spin his staff, cutting all the stalks of grass above a certain level. That illustrated how to maintain the highest level stalk of all. I'm into to the rejecting equality thing. I am seriously. I just suspect you haven't thought all of it through... (for example, have you read Nietzsche?)
- j_baker 16y ago
  I don't necessarily think you mean to reject equality. I think what you mean to say is that you reject objective decision-making. Objective decision-making has a tendency to treat everyone and every decision the same, making no exceptions due to circumstance. I think that what you're saying is that decisions need to be made in the scope of the current context. In other words, they need to be made subjectively. In other words, we can all be equal under the law. I am just as important under the law as anyone else (at least in theory). But that doesn't mean that the courts need to treat everyone the same. In the context of a company, all employees can be more or less equal. But that doesn't mean they need to be treated the same. Everyone is different, and you need to take that into account when you decide on salaries and position and whatnot.
kylelibra 16y ago
Tim Ferriss also has a good post on this if anyone is interested: http://www.fourhourworkweek.com/blog/2010/07/27/the-superstar-effect/ http://www.fourhourworkweek.com/blog/2010/07/27/the-supersta...
david927 16y ago
The problem, as The Spirit Level by Kate Pickett and Richard Wilkinson states, is that the larger the inequality in a society, the worse off everyone is, even the rich. The future is more global, which means it's more and more of Taleb's Extremistan. I don't think we should be working to try to change such effects; it's natural and comes with technological progress. What I would like to see is compassion and epathy for those increasing numbers who are left behind by it. We have to figure out this or it will be detrimental to us all.
- billswift 16y ago
  I have seen quite a few discussions of that recently, but I have seen no evidence that "income inequality" itself has any negative impacts. Now as a symptom of other possible problems, yes, it can be worrisome, think of the historical divisions, lords/rajahs/mandarins versus serfs/untouchables/peasants for a real example. Today's divisions in the Western world are neither that extreme nor have I seen evidence that they are the result of some underlying pathology like in those examples.
  - maigret 16y ago
    The basic problem with inequality is that it kills the middle class, which is one of the prime engine of growth. Riches only consume a small part of their money, and invest the rest, which produces easily investment bubbles, but not much growth in terms of industry and wealth production. That's (also) why the banking sector is getting so big. At the contrary, a strong middle class also maintains well democracy, and produces continuing growth over decades. Also, take it upside down - why isn't the US currently clearly richer that Europe? Germany, Sweden provide good examples that a bit higher taxing isn't killing the economy, much more the opposite. Also India and China have such a growth because they also produce a middle class, and not only super riches. But well, it's the old discussion libertarians vs utilitarians... ;)
- barry-cotter 16y ago
  http://super-economy.blogspot.com/2010/02/spirit-level-is-junk-science.html http://super-economy.blogspot.com/2010/02/spirit-level-is-ju... 'Blogger Danne Nordling pointed out a strange fact about the book. Its measure of inequality, the most important factor in the book, is not gini, the standard inequality result. They seem to use 20/20 richest and poorest ratio. Why make this strange choice, when their source (UN Human Development Index) has gini? I smell data mining. First for fun I did a simple regression of life expectancy on income inequality and per capita GDP for all countries the UN has data for. The correlation between inequality and health is not statistically significant. '
enko 16y ago
>> the 2010 FIFA World Cup in South Africa, in which Ronaldo played for Portugal, was broadcast in more than 200 countries, to an aggregate audience of over 25 billion Something's wrong with that final number, unless the world population recently quadrupled without my noticing ..
- Mz 16y ago
  I'm guessing it is missing the decimal point: 2.5 billion
- patrickyeon 16y ago
  I think "aggregate audience" means that there were a total of 25 billion "impressions" (the equivalent of one billion people watching 25 games).
- tednaleid 16y ago
  There is more than one game during the world cup, 25B is the total number of non-unique people who watched all the games. It says that 700M people watched the final. If that many people watched every game there would have been ~35 games total (potentially watched by the same 700M people). I have no idea how many games in total are actually part of the world cup.

79 points by unignorant 16y ago

This article is a bit innocent. It does not examine how the big tax cuts of 1982 changed incentives for everyone. There is a considerable amount of research that suggests when income tax on top rates is high, workers prefer non-cash benefits, rather than money. Recall that in the USA, during the 40s and 50s, the top income rate was 91% on all income over $100,000 (roughly a $1 million in todays money). Because of this, most top stars did not seeks payment in excess of $100,000. Instead, they sought benefits in to go along with their $100,000. The big tax cuts of 1982 lowered top rates to 36%. This made it much more rational for top workers to ask for high amounts.

TheBlack_knight 16y ago
That's a relavent point. High marginal tax rates would create the incentive for people in the higher taxed brakets to seek non-pecuniary benefits. Things like company cars and free jets. This would suggets we have an incomplete picture form this article.
jdp23 16y ago
Excellent point. The article's based on a time that's been better for the rich than ever before in many different ways -- and reduced the spoils available to everybody else. You have to wonder how sustainable that is ...

patio11 16y ago

I think PG had an article about this: if you have any process with a) a distribution of skill which b) scales with addition of resources, market price at the top of the distribution should explode because talent becomes a license to print money. Banking only got to that first because they have a very simple model: decisions for many types of assets take equal difficulty to get right regardless of how much money is stake (below a ceiling in the hundreds of millions or billions). If you can reliably make the average + 2% trading $2,000 of Coca Cola at a time (and you can't, but work with me here), then you can reliably make the average + 2% trading $200 million of Coca Cola. Comparing your wages to that of someone who trades hours for dollars no longer makes sense. Quants had a similar compensation scheme for quite a while now. I think other classes programmers are going to hit on it, too: there are many things you can do which demonstrably improve revenues or cut costs and which scale to the moon. Can I again toot the horn for A/B testing, metrics, etc etc? My homework for tomorrow is writing a consulting proposal. They want to know how much purchasing my services stands to make them. Consider a hypothetical company with 9 figures of revenue a year which does no A/B testing. Pick a number for what you think you could do for a first cut of conversion optimization. 2%? 5%? 10%? Do the multiplication. You can create staggering amounts of wealth these days with programming and a few other tricks, and you can capture quite a bit of it. [Edited to add: PG link: http://www.paulgraham.com/wealth.html http://www.paulgraham.com/wealth.html My favorite essay of his, incidentally. The relevant bit: To get rich you need to get yourself in a situation with two things, measurement and leverage. You need to be in a position where your performance can be measured, or there is no way to get paid more by doing more. And you have to have leverage, in the sense that the decisions you make have a big effect.]

lkrubner 16y ago
There is nothing wrong with the model PG puts forward, as a start, but it is a simple model and it needs explanation. At the least, you should make some attempt to explain why things change over time. Consider radio in the USA, up until the mid 1950s. It was eventually replaced by TV, but in the 30s, 40s and 50s it was the main medium of mass communication. The nation lived by its radio. The potential audience for a radio star was greater then than it is now. So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented? Changes in the tax code are important to consider. As a thought experiment, imagine if the top income tax rate was still 91%. (During the 30s, 40s, 50s and early 60s, it was 91% on income over $100,000, about a $1 million in today's money.) How many stars would ask for income in excess of $1 million if they knew it would be taxed at 91%? They would probably instead ask for various perks, which was the trend in the US during the mid-20th century. Asking for perks instead of cash remains the trend in those nations with strongly progressive tax rates, for instance, Sweden.
- patio11 16y ago
  So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented? I did a double-take here, because it sounds fairly obvious to me: monopsonies can get away with paying you next-to-nothing because your best alternative to "next-to-nothing" is "nothing". If there are only a few radio networks (and/or they have a gentlemen's agreement with respect to hiring talent), then regardless of skill the commercial value of your ability is zero unless you sign with them. Rush Limbaugh would be making $20k a year if there was only one radio station in town. There are now 50 and the one that gets to sign him makes a substantial amount of money. (He tapped an unlikely niche market for news and opinion: half the country.) The market now favors him rather than favoring them. Edited to add: By the way, it appears that Google et al have a gentleman's agreement with respect to hiring talent.
dspeyer 16y ago
I've always been rather suspicious of that essay. He says: Making wealth is not the only way to get rich. For most of human history it has not even been the most common. Until a few centuries ago, the main sources of wealth were mines, slaves and serfs, land, and cattle, and the only ways to acquire these rapidly were by inheritance, marriage, conquest, or confiscation. Naturally wealth had a bad reputation. Strongly implying that in the modern world most large fortunes come from making wealth, but presenting no actual evidence. From observation, most large fortunes come from trickery: either in markets or by being hired as a CEO and negotiating high, non-measured compensation by abusing conflicts of interest. I haven't studied the question carefully, though. Has anyone? He also strawmans the pure-pie model. The real pie model looks something like this. A very large group of people collaborate to create something of value (even in software, there are library authors, hardware makers, ISP maintainers...). Some are part of the same company. Some are tied in via contracts, shaped by negotiators. Once the good is made, the wealth is divided up among the team. (Simultaneously, the team works on the next hunk of wealth -- it's a pipeline, not an alternator.) It is meaningful to look at the distribution phase in isolation, with the caveat that how the distribution phase impacts the production phase is a factor in how good the distribution model is.

joe_the_user 16y ago

While logically explained, the process seems rather unstable. Where will it end up?

lionhearted 16y ago

Sooner or later, I'm going to get around to writing a piece and titling it, "I reject equality" where I lay out the argument that the goal of equality is fundamentally hostile to excellence. In fact, I think one of the biggest flaws in English right now is how we use the same word - equal - for concepts as diverse as law and economics. "Unbiased courts" and "taking things by force from people performing better" really, really, really shouldn't use the same word. They do currently - equality before the law, and income equality. Facts-based-unbiased-law is obviously a good thing. Take-by-force-from-best-performers, on the other hand... Anyways, I haven't written the "I reject equality" piece because of how touchy it'll be, and how I'll have to ultra-carefully craft the arguments to say exactly what I intend to say, and no more or less than that. But I do think a goal of total equality is fundamentally hostile to excellence - excellence, by definition, is better than the rest. At some point, you have to start pushing down on the best if you want everyone to be the same. I understand some people will disagree with me on this, but I personally think that's a really bad thing.

joe_the_user 16y ago
You know what's ironic about this reaction of yours to the article? What's ironic is that a world of fewer and fewer superstars making more and more money is a world in which there is more and more equality - in that all the others are equally in shadow of the few superstars. My Freshman philosopher professor told a parable. One despot asked advice from the another. The second despot lead the first to a field and proceeded to spin his staff, cutting all the stalks of grass above a certain level. That illustrated how to maintain the highest level stalk of all. I'm into to the rejecting equality thing. I am seriously. I just suspect you haven't thought all of it through... (for example, have you read Nietzsche?)
j_baker 16y ago
I don't necessarily think you mean to reject equality. I think what you mean to say is that you reject objective decision-making. Objective decision-making has a tendency to treat everyone and every decision the same, making no exceptions due to circumstance. I think that what you're saying is that decisions need to be made in the scope of the current context. In other words, they need to be made subjectively. In other words, we can all be equal under the law. I am just as important under the law as anyone else (at least in theory). But that doesn't mean that the courts need to treat everyone the same. In the context of a company, all employees can be more or less equal. But that doesn't mean they need to be treated the same. Everyone is different, and you need to take that into account when you decide on salaries and position and whatnot.

kylelibra 16y ago

Tim Ferriss also has a good post on this if anyone is interested: http://www.fourhourworkweek.com/blog/2010/07/27/the-superstar-effect/ http://www.fourhourworkweek.com/blog/2010/07/27/the-supersta...

david927 16y ago

The problem, as The Spirit Level by Kate Pickett and Richard Wilkinson states, is that the larger the inequality in a society, the worse off everyone is, even the rich. The future is more global, which means it's more and more of Taleb's Extremistan. I don't think we should be working to try to change such effects; it's natural and comes with technological progress. What I would like to see is compassion and epathy for those increasing numbers who are left behind by it. We have to figure out this or it will be detrimental to us all.

billswift 16y ago
I have seen quite a few discussions of that recently, but I have seen no evidence that "income inequality" itself has any negative impacts. Now as a symptom of other possible problems, yes, it can be worrisome, think of the historical divisions, lords/rajahs/mandarins versus serfs/untouchables/peasants for a real example. Today's divisions in the Western world are neither that extreme nor have I seen evidence that they are the result of some underlying pathology like in those examples.
- maigret 16y ago
  The basic problem with inequality is that it kills the middle class, which is one of the prime engine of growth. Riches only consume a small part of their money, and invest the rest, which produces easily investment bubbles, but not much growth in terms of industry and wealth production. That's (also) why the banking sector is getting so big. At the contrary, a strong middle class also maintains well democracy, and produces continuing growth over decades. Also, take it upside down - why isn't the US currently clearly richer that Europe? Germany, Sweden provide good examples that a bit higher taxing isn't killing the economy, much more the opposite. Also India and China have such a growth because they also produce a middle class, and not only super riches. But well, it's the old discussion libertarians vs utilitarians... ;)
barry-cotter 16y ago
http://super-economy.blogspot.com/2010/02/spirit-level-is-junk-science.html http://super-economy.blogspot.com/2010/02/spirit-level-is-ju... 'Blogger Danne Nordling pointed out a strange fact about the book. Its measure of inequality, the most important factor in the book, is not gini, the standard inequality result. They seem to use 20/20 richest and poorest ratio. Why make this strange choice, when their source (UN Human Development Index) has gini? I smell data mining. First for fun I did a simple regression of life expectancy on income inequality and per capita GDP for all countries the UN has data for. The correlation between inequality and health is not statistically significant. '

enko 16y ago

>> the 2010 FIFA World Cup in South Africa, in which Ronaldo played for Portugal, was broadcast in more than 200 countries, to an aggregate audience of over 25 billion Something's wrong with that final number, unless the world population recently quadrupled without my noticing ..

Mz 16y ago
I'm guessing it is missing the decimal point: 2.5 billion
patrickyeon 16y ago
I think "aggregate audience" means that there were a total of 25 billion "impressions" (the equivalent of one billion people watching 25 games).
tednaleid 16y ago
There is more than one game during the world cup, 25B is the total number of non-unique people who watched all the games. It says that 700M people watched the final. If that many people watched every game there would have been ~35 games total (potentially watched by the same 700M people). I have no idea how many games in total are actually part of the world cup.