Wolfers on Happiness

Excellent talk on happiness by Justin Wolfers. Robert Frank and discussion which follow are also good but Wolfers is outstanding.

A couple of things I learned. The flat happiness line in the United States over time is often contrasted with rising GDP per capita to assert a paradox. The paradox goes away once you take into account that median earnings haven’t risen (ala TGS).

Wolfers also shows that income matters not just for happiness but for a large number of correlates, inputs and outputs of happiness. Compared to people in poorer countries, people in richer countries, for example, more often say their food tastes good, they report less pain and they smile more.



I really like the new measures that Wolfers presents. The traditional 0-10 happiness scale assures a flat-lining at some point (or at least diminishing improvements) because there's no headroom. The new measures (e.g. "were you in pain?" "were you treated with respect?") are more informative and don't suffer from that problem in the same way.

The food bit is interesting, given widespread variation in (subjective) food quality independent of GDP. Put another way, I'd rather live in England, but I'd rather eat in Vietnam. Clearly preferences here are in large part driven by culture, but I'm not sure what that says more generally about the link between income and happiness.

If you'd really rather eat in Vietnam (as opposed to making an ill-informed joke about English cuisine), I strongly suspect you are conflating Western purchasing power in South East Asia with that of the median citizen. There is no sense in which the median English person is denied food opportunities denied to the median Vietnamese.

I'm not sure if there measure of happiness is self reported but I would suspect that income and over-reporting of happiness are positively correlated. .

Assuming median incomes really have been stagnant. the median income hypothesis is interesting. But if you look at places like Japan I think it's easily disproved...

If I had a face like Brad Pitt and a gig at the Wharton School I'd be happy too.

One out of two ain't bad!

Well, you nailed it

....median earnings haven’t risen .....

What if you they have risen and you're just counting wrong?

There's an old saying that you're entitled to your own opinion, but you're not entitled to your own facts.

But as you demonstrate, lots of people believe they are indeed entitled to their own facts.

There's another saying, a map (accounting measure) is just a map (accounting measure).

That's true.

So, like, median income has fallen.

But we can use household income instead! So everybody who moves back into his parents' basement and gets a low-paying job, will cause household income to rise! Joy! Median household income has risen!

We were just looking at the wrong statistic, that's all.

OK, pull the other one. It's got bells.

Read the paper, then type.

The fact that migration tends to happen from poorer to richer places should be seen as an argument in favour of the Easterlin story and not against it. The people these people compare themselves to are the ones staying at home. Their mothers back home take pride because they now earn more than their cousins, they don t feel sorry for them. I don't know, maybe it's because I am German with lots of friends whose parents were gastarbeiters, hmmm...isn't this obvious?

Similary, I don't move to the Czech Republic because my reference group stays put, namely in Bavaria.

In fact, rich people and social elites from poor countries do not migrate, even if their prospects would be much better in a richer country. The Easterlin story is pretty solid.

Every single time I go to get a drink at the Ritz in London I find that roughly 25% of the patrons are speaking Russian. I guess that Russian day laborers have expensive tastes.

I am not sure about that... everytime someone mentions social status as the main reason for any sort of behavior you can look at the other side of the coin and disprove his point because status is so subjective. I know a lot of middle class people in Brazil who moved to the US just so they could be considered rich in Brazil (even though their actual status in the US was lower). Now, the real question is: what percentage of middle class people moving had that as their main goal? I don't think we can just assume 100% on any of those options.

Wolfers' claims at 10:25 that he is yet to find an example of an unhappy millionaire in the data. What about Kurt Cobain?

He couldn't find him.

It must be better to be rich and anonymous than rich and famous.

Even better to be rich, and not be alcoholic / junkie / mentally ill.

Yes! If we collect our statistics on rich happy people, we will find out that they are happy!

How is someone sure that no social scientist has done a study of something? I have 3000 papers in my ref database related to a subject some of you wouldn't even know exists and I've just begun.

How is someone sure that no social scientist has done a study of something?

When I'm in a big hurry to find information about X, something I don't really know about, sometimes I look for a place where people are likely to know about it, and I announce authoritatively that there is no solid information anywhere on X, that nobody knows anything except me, and I know that X=Y.

Usually then the locals immediately prove me wrong, and provide citations, and pretty quickly they're so busy arguing with each other about minor details that I can slink off with my links.

Of course I don't do this where people know me, because I don't want them to think I'm some dummkopf. But this method generally works better than a polite request for information.

Also it pays off better for them. When I politely ask for info then I thank them afterward. For a whole lot of people that isn't nearly as good as proving that some dummkopf was wrong wrong wrong. That's a whole lot more reinforcing than being nice and getting thanked for it.

I find it strange that Robert Frank suggests that the Federal Government should implement a progressive consumption tax when he also believes that individuals use local (city, county, place of employment) information to judge their relative welfare.

"The paradox goes away once you take into account that median earnings haven’t risen (ala TGS)."

Not only that but if you accept as fact that happiness increases with the log of wealth increases, that flat-line of happiness in recent years is good evidence of TGS and a strong argument against those who argue that material gains from the internet and smartphones haven't been properly accounted for. Which is also to say that measuring happiness may be a good new tool for economists indeed, as it offers a reality check against whether we are measuring the value of new technology accurately.

Well, but the internet has impacted many countries at once... if the level of hapiness grows to many countries at once many of the indicators he uses like immigration don't make sense, right?

Youtube link.

While I am in broad agreement with Wolfers on the primacy of absolute income, I disagree with the example he uses against the relative income hypothesis. You can choose your country but you can't change your reference peer group. The immigrants are happier with their absolute increase in income because their psychological reference group is still their peers in their native country. However their children are less happy -- perhaps less happy than they would've been back in their parent's less wealthy homeland -- because their reference group for a respectable "normal" income is white American natives.

To put some data to this I just checked the General Social Survey. The happiness of Mexican Americans steadily decreases as they become less immigrant. The happiness of Mexican born immigrants is about 1/6 of a standard deviation lower than American whites. But the difference in happiness between Mexican-Americans with 8 American-born grandparents and whites is 1/3 of a standard deviation. Since income remains similar across the generations, one likely explanation is that their status reference group gradually shifts from other Mexicans to general Americans. An American minimum wage job is good money for an immigrant who instinctually compares himself with his peers back home, but a symbol of ethnic disenfranchisement for a Mexican-American who instinctually compares himself with the average American white person.

White, US born
Mexican, Mexico born
Mexican, US born, parents Mexico born
Mexican, US born, parents US born
Mexican, US born, all grandparents US born

Very interesting numbers Jason.

I ran a quick and dirty regression several years ago checking Frank's house hypothesis. Contrary to the house hypothesis people do not want the largest house in the neighborhood.


"A couple of things I learned. The flat happiness line in the United States over time is often contrasted with rising GDP per capita to assert a paradox. The paradox goes away once you take into account that median earnings haven’t risen"

I don't really understand how it makes sense to compare a median measure ("median earnings") to an average one ("average hapiness"), which is what Wolfers seems to imply. In a statistical sense, you either compare real gdp per capita to average hapiness, or median earnings to median hapiness.

I suspect the scale of the wealth sample is much larger than than happiness one and more likely to be misrepresented by an upper crust of billionaires.

As in, Me, you, a lottery winner and the rest of our high school class are measured for happiness. The lottery winner would only count as one sample on a yes/no happiness questionnaire, but would greatly increase our GDP. We would not appear particularly happier than the average class. We would be richer than the average class. Our median student wouldn't be richer.

Thanks for sharing.

I agree that Justin presented some novel measures of happiness, which go beyond the happiness scale. But some of the same problems remain when it comes to aggregating number across people and interpreting the results for policy decisions. Is 2+2 the same as 1+3? Maybe we would apply the Pareto optimality to happiness:no decrease for any individual.

I found Robert Frank's point about escalation effects interesting. But thinking more about it is unclear where is the waste. It seems that people motivating each others and raising standards would be a good thing. On top of that, waste, just like value, is subjective.

On top of that, waste, just like value, is subjective.

Yes, there is more of a whiff of 'waste' being whatever academics find tacky and vulgar. So, for example, a $5,000 Viking Range in a kitchen of person who rarely cooks -- pure wasteful status competition, right? The best defense I've seen is, "But to me it is a beautifully designed object -- it gives me pleasure every time I see it. Why is it that my $5,000 range is a 'waste' while a $5,000 original work of art is not?"

A good reason to believe that the hockey helmet analogy does NOT apply to luxury goods is that hockey players are willing to vote for helmets whereas Viking Range owners are generally not willing to vote for confiscatory taxes to relieve themselves (and all their supposed zero-sum status competitors/peers) of the heavy burdens acquiring and owning too many lovely things. Probably in addition to 'The Millionaire Next Door' a copy of The Substance of Style would be good addition to Frank's library as well.

I don't want to pick on him too much, though -- on the topic of education (the negligible effects of traditional university courses and more effective alternatives), I think he's brilliant, and more work there is sorely needed.

I find it ironic, given the title of his latest book, that Robert Frank does not recognize the desire to compete as a basic part of human nature, and indeed a universal part of biological organisms more broadly. It was perhaps THE central concept to Darwin!

Limiting competition along specific dimensions could of course be desirable - as in RF’s hockey helmet example. Rules exist in all sports in fact, even “extreme” ones.

From an aggregate perspective, however, taxing specific “positional” goods (as Bob proposed in his earlier work) will merely change the composition of the “overall basket” by moving relative prices around. A shift from goods X to goods Y does not equal an aggregate “decrease in wasteful competition” – i.e. there are no “freed up resources” to be deployed for alternative uses.

RF has more recently moved away from advocating a tax on specific goods and is now suggesting an overall progressive consumption tax. This idea once again ignores the fact that the basic desire to compete cannot be suppressed by outside intervention. RF’s proposal is in fact potentially harmful, as it seeks to suppress or greatly dampen a major advantage human societies have over what we see in biology more broadly: a system of institutions, property rights and prices to allocate resources and endogenously coordinate the actions of many separate individuals. Unlike the herds of elk growing ever larger antlers, human societies constantly redefine the dimensions along which we “compete”: for better or worse, it is a status signal to drive a Toyota Prius, start a charity or “buy organic”.

While it is true that in Stalinist Russia your neighbor would not show you off with their fancy new car or expensive watch, they could certainly achieve higher status by denouncing you as an “enemy of the people” so the local authorities could meet their yearly Gulag quotas. “Deadweight” losses were quite a bit higher – as we know from people like Solzhenitsyn.

I think economists get wrong when they assume only consumption causes happiness, and thus the only purpose of work and production is consumption. In reality people enjoy baking cakes at least as much as eating them.

So you have the left-wing economist who wants to redistribute cakes for consumption, and the right-wing economist who wants to established the most efficient structure for making a lot of cakes (which is the free market), and both have this bias that the absolute ideal state of things is when everybody has plenty of cakes to consume, and thus they ignore the pleasure taken from cake-baking.

Now. If we assume that the perfectly happy person is not simply a hole into which endless consumer goods disappear, but a person who is also doing creative work in enjoyable ways, either as a job, or self-employed, or at home then there is a lot to rethink... I think it makes the case for a specific kind of free market: that of small indie family businesses and freelancing.


Yes, I think when people are not allowed to compete for status in peaceful and objective ways, they either get it more aggressive, or by more obscure rules. For example, when I was in high school, what infuriated me that nobody could tell how exactly to get popular - should I get stronger, or buy more expensive clothes, or help dumber classmates with the homework, or what. There was just a general of air, hint, gist that doing, saying, wearing some things is considered "cool", but nobody could tell what and how exactly to do so or what measure separates uncool things from cool things. It was very infuriating because I did not know even how does one begin competing, what are the rules of the game. They money-based status-competition of adult life is a lot more preferable because the rules of the game are so simple - acquire BMW: get respected.

But, didn't Arrow show you couldn't all of the people happy all of the time?

More seriously, what is the debate here? That increasing GDP is not a way to ensure happiness?

Can I just say that happiness is a rich man's concept?

I'm not sure if maybe I missed it, but what about perception?

There was a lot of talk about absolutes and relatives, but as far as I noticed there was no mention of anything being perceived. Rory Sutherland talks very well about perception at his talks on TED (www.ted.com) and is one of many good speakers there who is worth looking at (also Dan Gilbert of course). Human behaviour and aspiration is targeting what we perceive as the best. When people are standing at the border of USA and Mexico the perception will rule the path. What we perceive as the most happy scenario is the one we will choose.

A poor alcoholic who has just found one hundred dollars will be looking south toward the cheaper drink and the spoiled little mothers boy who is standing at the same border with one hundred dollars will be spending that money to get home to Minnesota. That is a huge part of what they were talking about but it never really surfaced.

People see a big house so therefore they also want a big house? What if what they see is a huge electrical bill with a pool? A poor country is rising fast in wealth and the population is finding out about the internet. Suddenly there little hut of straw can be compared to houses in other countries where everyone really have there own room. Maybe that is perceived as great, or lonely, or maybe they have also learned about a thing called an electric bill?

Regardless of which, the perception is without doubt my main frame of reference when looking at these kind of comparisons. While I enjoyed the talk and the information, I miss the reference to that particular aspect in the analysis of the results.


"I don't have theories, I have facts"
People who think they don't have theories generally just have implicit ones.

I'm not an economist, but I may resemble his strawman of those who favor GDP over happiness.

I wish Frank relied more on data and less on narrative.

I didn't know anybody growing up who had a "coming of age party". A birthday was a birthday, during certain years your friends would be happy that you could get a license to drive them around or buy them nudie mags/cigs/booze.

I approve of shifting taxes from income to consumption. Stop taxing investments.

"In every society access to a good school depends on where you live". Aren't there some places with school choice? Sweden?

I'd like to give Kudos to Alex on empirically examining "mcmansion pollution". Frank mentions within-firm inequality, so I'll refer to Mungowitz and David Friedman (I'm afraid of using too many links, so search for "Frank Folds" at Overcoming Bias).

Robert Frank points out that relatively poor people save less. Could it be that the kind of people likely to be poor are also the kind of people less likely to save? Do immigrants from Mexico start saving less?

Comments for this post are closed