Showing posts with label product substitution. Show all posts
Showing posts with label product substitution. Show all posts

Sunday, August 12, 2007

Are Landlines Becoming a Giffen Good?


Widespread use of VoIP tends to cause voice prices to fall. And classical economic theory would suggest that consumption of wireline calling should increase, as a result. In some cases that seems to be what happens. People call globally more often when the prices are lower.

But it just is possible,under some specific circumstances,for price declines to cause reduced consumption.

A Giffen good, for example, is an “inferior” good for which a rise in its price makes people buy even more of the product as its price rises. Conversely, there is less demand as price falls. To be sure, such Giffen goods are exceedingly rare. But one is tempted, when looking a mobile versus fixed line calling, to ask whether there are not some similarities.

Mobile calling now leads wired connections by a three-to-one margin globally and more people are shifting to "wireless only" calling. And though it is a loose analogy, perhaps we might think of mobile calling as a "superior" product and wireline calling as an "inferior" good, not in terms of intrinsic worth but in terms of the way people consume each product.

Giffen goods are named after Sir Robert Giffen, who was attributed as the author of this idea by Alfred Marshall in his book Principles of Economics. The classic example of potato consumption during a famine now is viewed as unsupported.

But in July, Robert T. Jensen, an economist at Brown University, and Nolan H. Miller, a professor of public policy at Harvard University, published an article for the National Bureau of Economic Research on Giffen goods.

The two economists say they have located a real-world Giffen good, namely rice and wheat flour in the central Hunan and Gansu provinces of China.

As Giffen suggested more than 100 years ago, goods whose price and demand move in the same direction are most likely to be essential products such as food on which households spend a large part of their incomes (and that's why neither VoIP nor landline voice service can be called Giffen goods in a formal sense).

Wheat flour and rice fit the bill in central China. When the price of the good falls, households appear to shift buying to meat. So lower prices cause less consumption.

Jensen and Miller look at poor Chinese consumers and demonstrate that they consume more rice or noodles, their staples, as prices go up.

Still, neither VoIP nor landlines strictly meet the criteria for consideration as Giffen goods. But it is an interesting notion. Might lower landline calling prices caused by VoIP actually lead to lower usage, in the presence of mobile alternatives that might be likened to “superior” goods, as compared to landline which might be thought of as an “inferior” good?

If so, lower landline calling prices will simply hasten the transition to more preferred mobile calling. I wouldn't push the loose analogy too far. But there some parallels.

As the chart suggests, consumers can buy either commodity Y or commodity X (line MN,where M = total available income divided by the price of commodity Y, and N = total available income divided by the price of commodity X).

The line MN is the consumer's budget constraint.

If there is a drop in the price of commodity X, the reduced price will alter relative prices in favour of commodity X, known as the substitution effect. This is illustrated by a movement down the indifference curve from point A to point B.

At the same time, the price reduction causes the consumers' purchasing power to increase, the income effect (line MP where P = income divided by the new price of commodity X).

The substitution effect (point A to point B) raises the quantity demanded of commodity X from Xa to Xb while the income effect lowers the quantity demanded from Xb to Xc.

The net effect is a reduction in quantity demanded from Xa to Xc making commodity X a Giffen good by definition.

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...