You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Why We Can't Measure Pain And Pleasure

In the latest issue of Fast Company, Tyler Cowen makes the intriguing claim that with the advent of the web, there is a lot of domestic product that's not being captured by a measure like GDP:

The traditional gauge of economic success is profit, but over time we'll find that such statistics as measures of GDP tell us less and less about broader efforts to improve human well-being. Much of the Web's value is experienced at the personal level and does not show up in productivity numbers. Buying $2 worth of bananas boosts GDP; having $20 worth of fun on the Web does not. And this effect is a big one. Each day more enjoyment, more social connection, and, indeed, more contemplation are produced on the Web than had been imagined even 10 years ago. But how do we measure those things?

As Felix Salmon pointed out, this is not necessarily a new argument. There've been numerous complaints in recent years that GDP doesn't capture everything we'd want to know about an economy's health -- i.e. sustainability, mental effects on the populace (happiness), etc. In fact, last fall, France commissioned Joseph Stiglitz, Amartya Sen, and others to come up with a metric for economic growth that would include non-monetary components.

Unfortunately, the prospects for a mental-health tracker still seem a ways off. Or impossible, says Yale's William Nordhaus in a new paper:

Assume for purposes of discussion that we have developed a perfect hedonimeter based on brain scanning, and further that we have accurate techniques that map how brain images correspond to reported pain, pleasure, sadness, sweetness, or ther features of reported emotions. Perhaps we can even calibrate the level of pain or frustration that would make me frown or grind my teeth. Would it make any sense to add these together, or to average these emotions?

Nordhaus says no for three reasons:

- What would be defined as a 'zero' value?

Because emotions are so contingent, the zero point will vary with mood, circumstances, genetics, context, history, and culture. Therefore, there is unlikely to be a natural zero point for happiness, misery, pain, or other emotions.

- What unit of measurement would be used?

We cannot say how the incremental pleasure that Sam experiences in eating a “delicious cheeseburger” compares with the incremental pain that Helen experiences when she has a "bad headache." Therefore, it is difficult to see how the increment of emotions can be calibrated across different individuals.

- Measures of emotion are inherently relative, not absolute:

An example will illustrate the point. Constructing an index of aggregate pain or pleasure is similar to creating an aggregate index of the blueness of the Danube River. I do not doubt that in some ideal world we can make measurements of the spatially averaged wavelength of the light coming off the water. We might be able to measure the physiological responses to particular wavelengths of light in different people. Moreover we could potentially correlate these physiological responses with how people describe their experience: whether the river is “blue” or “deep blue,” or even so pleasurable as to inspire a song about “the beautiful blue Danube.” However, it would make no sense to construct a national index of “Blueness of the Danube River” that involved adding up how individuals on a particular day report the experience of looking at the Danube River. Nor would it make sense to have an index of “Blueness” that would go up or down from day to day depending upon unemployment, inflation, or per capita income. Neither blue rivers nor blue moods constitute a meaningful index of emotions because they are not based on interpersonally cardinal [absolute] variables.

 Related: Read some criticism of Nordhaus's model for climate change at Rortybomb.

--Zubin Jelveh