Arpit Gupta has a piece on taxes and labor supply.

Here are a few things to remember

Marshallian labor elasticity is what we normally think of as labor elasticity. It measures how people will respond to an increase in their tax rate if the taxes are used to say, pay down the debt.

You have two effects here. One, taxes make working less worthwhile so people work less. Two, taxes make you poorer so you tend to work more

Hicksian labor elasticity is what you get when you take away the income effect. So you pretend that taxes make working less worthwhile but you ignore that taxes make you poorer.

This seems strange but it is important to economists because welfare measures are properly done using Hicksian elasticity not Marshallian.

Frisch Elasticity is a little more complicated but in essence it measures what happens if we tax a consumer and then turn around and give the money right back to her.

This is important because if you think of an economy with a government budget constraint then something like this is happening. Usually the money doesn’t just disappear, you use it for something that presumably benefits the tax payer.


All of that having been said I think when we communicate the effect of “taxes” to lay people we are really talking about Marshallian effects. This is because the pubic thinks about welfare in its own terms and because most taxes fund what are effectively transfers and so the tax payer doesn’t get the money back.

What you should then do is separately talk about the effect of benefits. So, you can see that taxes may have very little effect on behavior as Marshallian elasticities are low. However, benefits have stronger effects as Hicksian and Frisch elasticities are higher.

This makes the whole thing clearer in the non-economists mind. Taxes do not discourage people from working. Benefits do discourage people from working. If we taxed Americans to fund a war in Iran labor supply would probably rise. If China gave a huge gift to the United States to fund a reduction of the SS eligibility age to 55 then labor supply would fall.

This makes sense no?

And, it fits with the differing elasticities that we measure.