Television viewing vs. sex

Consider two goods (A and B) that are substitutes - consumers tend to switch consumption towards the good that has become relatively cheaper when the relative price between the two goods changes. So, if the price of A falls, consumers will tend to buy more of A, and less of B. And if the price of A rises, consumers will tend to buy less of A, and more of B.

We can extend the idea of the relative price to consider the full cost of obtaining each good. So, if A becomes more difficult to find, the full cost of consuming A rises, and consumers will buy less of A, and more of B. And so on. Now consider two specific goods: television viewing and sex. Based on the basic model we just outlined, if the full cost of television viewing falls, then consumers will watch more television, and have less sex.

We can also illustrate this using the diagram below. Let's assume that the 'consumer' doesn't own a television. The 'consumer' has a choice of two goods: television viewing (on the x-axis) and sex (on the y-axis). They have limited time (say 20 hours per week), that they can spend on either activity, and this constraint is represented by the straight line (from 15 hours television viewing and no sex, to 20 hours of sex and no television viewing). They can only get to 15 hours of television watching rather than 20, because they have to find someone else who owns a television and will let them watch, and that takes up some time. The consumer will choose the point on their time constraint that is on the highest indifference curve (I0) which touches the time constraint at one point (E0), where they spend T0 hours watching television, and S0 hours having sex.

If the consumer buys their own television, they no longer have to spend time trying to find a television to watch somewhere else, so the full cost of television watching decreases. The time constraint pivots upwards (which represents a decrease in the relative price of television watching). The consumer can now reach a higher indifference curve (I1), and will maximise their utility by consuming at the point E1, where they now spend T1 hours watching television, and S1 hours having sex.

Notice that television watching has increased greatly (it is now cheaper and much more convenient), but time spend having sex has decreased. [*] So, television really might kill your sex life.

Is there evidence for this trade-off between television viewing and sex? In a new NBER Working Paper, Adrienne Lucas (University of Delaware) and Nicholas Wilson (Reed College) find suggestive evidence for a negative relationship between television ownership and coital frequency, using data from nearly 4 million individuals from 80 countries. Specifically, they find that:

...the results of the analysis are consistent with a small amount of substitutability between television viewing and sexual activity. We find that television ownership is associated with approximately a 5% reduction in sexual activity, a statistically significant yet not particularly large association.

The results are correlations rather than causal, but Lucas and Wilson at least eliminate some of the main potential confounders such as wealth or reproductive health knowledge. Interestingly, the effect seems to be concentrated among women (the coefficient of television ownership is not statistically significant for men). It's robust to the inclusion of other durable goods (refrigerator, radio, car, motorcycle, and bicycle). Interestingly though, owning a refrigerator is also negatively associated with coital frequency for women (but not men). Does that imply that night-time snacking is also a substitute for sex?

[HT: Marginal Revolution]

*****

[*] With the price change here, there are actually two effects: (1) a substitution effect (television watching has become relatively cheaper, and sex has become relatively more expensive, so the consumer should watch more television and have less sex); and (2) an income effect (the consumer now has more time because they don't have to waste time trying to find a television, so they both watch more television and spend more time having sex, because both activities are 'normal goods'). In my diagram, I've assumed that the income effect on time spent having sex is smaller than the substitution effect.