Public Finance, 2nd Edition

``gross-of-tax'' price) rises from q

0

to q

T

, the consumers suVer a loss of

consumer surplus (ignoring income eVects) equal to q

T

BCq

0

. Because the

producer price (the ``net-of-tax'' price) falls from p

0

to p

T

the producers

suVer a loss of producer surplus equal to p

0

CDp

T

. Using the impact, or tax

revenue, as a measure of the true burdens, then, would overstate the produ-

cers' true economic losses by (q

T

BAp

0

 CAD) and understate the con-

sumer's true economic losses by q

T

BCq

0

.

1

Even though this example is only a partial equilibrium analysis of tax

incidence, it illustrates a fundamental point: The market's reactions to a tax

are a crucial determinant of its ultimate pattern of burdens.

FIRST-BEST THEORY, SECOND-BEST THEORY,

AND TAX INCIDENCE

We have delayed discussing the general theory of tax incidence until this part

of the text since the most interesting questions in tax incidence are inherently

second best in nature, precisely because they depend on the market's response

to distorting taxation. In Wrst-best theory, all questions of distributional

equity are incorporated into the interpersonal equity conditions,

@W

@U

h

@U

h

@I

h



@W

@U

j

@U

j

@I

j

all h, j  1, ...,H 16:1

which the government satisWes through a set of lump-sum taxes and transfers

among the consumers. As a general rule, these lump-sum redistributions

aVect equilibrium market prices, contrary to the common assumption that

they do not, and these price changes in turn aVect people's utilities. The

incidence, or burden, on the consumers could presumably be measured as

the diV erence in their utility levels before and after government redistribu-

tion, but computing the change in utility for each person is not especially

interesting. Consider the various possibilities.

On the one hand, the equilibrium market prices may not change if, for

example, aggregate production technology is linear with constant marginal

opportunity costs. If producer prices remain unchanged, so too will the

vector of consumer prices with lump-sum redistributions. In this case, the

tax and transfer payments are perfect income proxies for the change in utility

in this senseÐif any one person received (paid) the tax (transfer) back from

(to) the government, his original utility level would be restored. Therefore,

the impact and incidence of the redistribution are identical, so that the

incidence question is trivial. In practice, tax theorists have often been willing

1

This example is meant only as an illustration of the distinction between tax impact and tax

incidence. As we have noted in other contexts, partial equilibrium Marshallian consumer and

producer surpluses are generally not valid measures of consumer and producer losses.

16. THE THEORY AND MEASUREMENT OF TAX INCIDENCE 525