Thursday, April 7, 2011

Happiness and Power

One of the common assertions around is that no combination of renewables can power a consumerist society. This may be true. Its a testable hypothesis. But is that the point? Surely the question is are we happy? And in fact Economics is (or should be) about the study of happiness, not just the study of GDP growth (consumption), its just that among some punters GDP/capita is equated with happiness. The "invisible hand" will "lift all ships on a rising tide" (or some such). In the past few years research suggests that above a certain threshold, increases in GDP/capita have a negligible effect on perceptions or feelings of happiness.

The Economist had a little graph late last year displaying this very phenomena, although they couldn't help but change it to a form more aesthetically pleasing to them.

Money and happiness

The Economist, Nov 25th 2010

DISMAL scientists who look at happiness often contend that, beyond a GDP per capita of just $15,000 (measured at purchasing-power parity), money does not buy happiness. Up to that point the correlation between the two is strong, but thereafter it falls away. If this is true then some heretical conclusions follow: rich America is no happier than poorer Brazil, so what is the point in people who live in rich countries working harder to get ever richer? Politicians should concentrate on maximising the mental health of their voters, rather than the size of their pay cheques. But plot the data another way, on a logarithmic scale where each increment represents a 100% increase in income per head, and the relationship between wealth and happiness looks more robust.

Personally I don't mind showing both of these curves together, but claiming that the second version on the right is superior is misleading. Unless you read the last sentence the strong impression given by this graph (especially with the addition of the line – notably absent on the left) is that the relationship is linear-ish (even though the axis is labeled as  log scale).

GDP is correlated with energy so we can also view this data using energy as our predictor of “happiness”. For the image below I show electricity consumption per capita (pertinent in the current debate) vs “happiness”. The data was obtained from Nation Master.  The Happiness data is the rank of the countries happiness score  based on the “quite  happy” category result (and appears to be fairly similar to that used by The Economist).


The same plateau is seen in this graph only more pronounced. The line at top shows probable empty watts or negawatts as no additional increment of happiness accrues for this excess expenditure of energy. Either some populations are very hard to make happy (or conversely easily entertained?) or a lot of electricity is potentially being wasted.

Removing the obvious outliers and concentrating on the first part of the curve (ie upto 10kWh/cap) shows that a linear relationship (while actually significant: p = tiny)  explains about 30% of the variance (r2 = 0.29) strongly suggesting that other factors are also important.


I am not saying that electricity is unimportant (or the least important), merely suggesting that it should be possible for those in affluent western countries to adjust to a lower level of consumption and still remain quiet happy.

There are of course many potential confounding factors in such a simplistic analysis; what is happiness being among them.

But are we right to merely keep projecting ever greater energy consumption claiming ‘this is what we need’, when maybe we don’t?


 Grist covers a similar point in the article accompanying these flawed US projections from the past.

See also,

The Two Most Important Graphs in the Energy Biz

No comments: