Talk:Compensating variation

wrong sign?

I believe this entry is completely incorrect. Compensating variation refers to the definition given at the botton. It is a measure of welfare change. The definition given at the top is actually equalizing wage differentials

• The two definitions are more or less the same, if you think about it, and the term is used in a lot of different contexts (e.g. look through the search results on google) to mean, more or less: the change in $ that is needed to offset a change in [some other thing]. Christopher Parham (talk) 03:39, 18 January 2006 (UTC)

I disagree. While they vaguely sound the same, compensating variation is very specifically defined in economics. See equivalent variation for a slightly different measure of welfare that is related. These two are almost always defined together in textbooks. Note that all the definitions on Google use the definition given at the very bottom. Furthermore, the arguement made in the body of the article ignores the important problem of self-selection with respect to job attributes. For example, if there are two jobs, one is safe and one is dangerous, the difference in pay cannot be considered what you would need to pay the person in the safe job to take the dangerous job. This is because workers who prefer safety will sort into the safe jobs in equilibrium. This results in a major and well known problem for empirical work on equalizing wage differentials.

I suggest that all of the top content is moved to Equalizing Wage Differentials and the bottom definition is expanded and linked to equivalent variation.

• You are welcome to do whatever you want, but see for instance this definition: "estimate in money terms of the amount households would require as compensation in order to remain as well off after an exogenous shock as they were before the shock." e.g., shock could include change in prices, or a change of jobs, or a change in the location of your house, etc. See for instance, this paper (search compensating variation) for an instance where it is used to describe the differential for a change in job hours. Indeed, compensating variation seems to be at least as common a term for this as "equalizing wage differentials." The point is that the exogenous change is not limited to change in price. The article does point out that the market prices will be determined by the marginal actors, but does not fully explain why self-selection accounts for this. Christopher Parham (talk) 00:10, 19 January 2006 (UTC)

• I agree it doesn't have to be a price change. But the difference is NOT what you observe in the market. In fact it is impossible to observe the compensating variation difference because they are based on compensated demand functions not Marshallian demand. Anyway, I'll rework this if I get a chance. For now the discussion page provides some reference to the debate.

I agree that the sign is wrong. As I know it, CV is:

$CV=e(p_{1},u_{1})-e(p_{1},u_{0})$

so that it's positive if they are better off, just like EV.

Old content

I got rid of this. Should be equilizing differentials, or hedonic models. Not clear where this is from or if it is actually correct. Tristan47 23:17, 3 April 2006 (UTC)

In economics, the principle of compensating variation says that when two apparently similar things are differently priced by the market, there must be some difference between them that compensates for this price difference. If this were not the case, then there would be an opportunity for arbitrage.

Compensating variation is especially applied to labor markets. For instance, suppose that there are two jobs available which require no special education and thus appeal to mostly the same group of workers, but one pays twice as much. If there were no compensating variation between the two jobs, all workers would go into the higher-paying field, which would drive down wages in that field. Thus, if there is a sustained wage difference, we know that the higher-paying job has some factor that makes it less appealing than the lower-paying job; it may be more dangerous, dirtier, more socially degrading, and so on. The factors that make the difference in wages possible are the compensating variation and the difference in wages itself is called the equalizing difference.

A compensating variation can also be used to explain the difference in wages among different groups of workers. A wage gap may be seen if employers are able to set different prices for labor among different groups. For example, if allowed, employers may value equivalent work by one group less if they tend to quit sooner, forcing the employer to train replacements more often. Furthermore, there may still appear to be an unfair wage gap for seemingly equal labor even when individual employers don't discriminate wages. The principle of compensating variation suggests that there must be some hidden difference in the labor between the groups. For example, the motherhood wage gap can be partially explained as a compensating variation in the preferences of mothers and non-mothers. Family responsibilities creates a compensating variation in work preferences; a mother may be less likely to work long hours or relocate her family to take advantage of the most lucrative jobs available for her skill level.

Compensating variation can also be applied in real estate markets. For instance, two similar houses may have very different costs depending on their proximity to good schools or centers of employment.

In some cases, the price difference between the two goods can be used to measure the value of the compensating variable. For instance, suppose there was a dangerous job and a safe job requiring equal qualifications; the difference in wage between the two could shed light on how much people value their life.

The price difference reflects the preferences of the marginal consumer. For instance, suppose that there was a demand for ten air show pilots, a "dangerous job", and ninety jumbo jet pilots, a "safe job". Suppose also that of the 100 available workers, 9 are indifferent about the danger, 2 require a salary $1000 higher to take a dangerous job, and the other 89 require a salary $100000 higher to take a dangerous job. In this market, the dangerous job will pay $1000 dollars more than the safe job – the preference of the marginal actor determines the difference.

This appears to regard Compensating_differential. I will paste it in that talk in case some of it is useful. PDBailey (talk) 01:30, 5 January 2009 (UTC)