.,t!!

tl,r

The Theory of Regulation 319

' the regulated price P-, then regulation is unlikely. The interest group rhat
pdlllim,u'. : P eltzman Mode i
benefit from regulation will not expect to gain a large amount because
,Lrld
. :ce would be relatively unaffected. Hence it would not warrant the investment
- resources to get the industry regulated. Since the regulated price lies in
-:rweer-r P'and P*, this argument also suggests that the industries most likely
- te regulated are those that are either relatively competirive (so rhat the
-:egulated equilibrium price is near P') or relatively monopolistic (so thar the
- - :egulated equilibrium price is ne ar P'") . In both cases, some interest group lvi11

-,:r.r considerably from regulation. Firms will gain in the case of a competitive
iustry, u,,hlle consumers will gain in the case of a monopc'rlistic industry.
Casual observation suggests that it is indeed these two extremes that
,..J to be subject to economic regr,rlation. Monopolistic industries include
-:1 telephone, electric and gas utilities, and railroads. Rclatively comperitive
'rrstries include agriculture (regulatlon takes place in the form of price
:ports), trucking, taxicabs, crude oil and natural gas production, and
:: -lritieS.

:-,rker Model The Stigler/Peltzman modeling of the economic theory of

..rlation is based upon a legislator/regulator choosing regulatorv policy so as
naximize political support. In contrast, the formulatlon of Gary Becker
.lses on competition between interest groups.20 He suppresses the role of
legislator/regulator by assuming that "Politicians, pohtical parties, and
lnCrea:c :rs transmit the pressure of active groups."2l True to the economic
..reasing in y of regulatio'n, Becker assumes that regulation is used to increase the
' -: depends on :are of more influential interest groups.
, n(P) is i For simplicity, suppose that there are two interest groups denoted group 1
.. decreasing ::- . group 2. An interest group can raise its welfare by influencing regulatory
: . -:e 10.2. For I ;v. The wealth transfer that group i gets depends on both the pressure it
: rlpposition on legislators and regulators (denoted pr) and the pressure exerted by
,- :.)rt [because 7 2 (denoted p2). The amount of pressure is determined by the number of
hers in the group and the amount of resources used. GreAter pressure hv
rcal suPPort 1 as well as less pressure by group 2 implies that [Link] t h:rs more
e on thl: political process. Greater influence translates into groLrp
-riterence ctlr\ .:
- 1

and profit that

:r'ing a bigger wealrh transfer. In particultrr, if T is gror-rp 1's incre ase in
Jecreasing in h due to regulation, then T : i'(P,,p,), where I'(p,, p,) i,s the lnillrence
' -:-L\Yesterly dir
:rron of group 1. It is assumed that 11(p,, P,) is increa,sing in rl-re pressure
. denoted P' -:.rup 1 and decreasing in the pressure of group 2. In or,'ler to transfer wealth
,]ect to the .:nount T to group 1, group 2's wealth must be reduced. The difTerence
drat P- lies betw .en the amount of u'ealth taken from group 2 ancl the amount of wealth
ly price, P to group 1 is the deadweight welfare loss from regulation.
-::lved the restr--
\ property of the Becker model is that aggregate int'luence is fixed. Tl'ie
industry protir cation is that what is important for determining the amount of regulatory
e provides imPort
rtv (as measured by the wealth transfer) is tfic influcnce of one group
t from regulation- re to the influence of another group.
ofregulati, ' :ach group chooses a level of pressure so as to m:rximize its welfarc givr'n
320 Chapter 10 Introduction to Economic Regulation

lllSigtrre.1O'3 P oliticat E quilibrium: B ecker Model

Pz

the pressure chosen by the other group. Since greatet pressut€ usi:
the group's resources, each group will not want to apply too much pre
On the other hand, the less pressure a group applies, the greater the infl
of the other group. Hence, by reducing pl, the relative influence'of
declines, so that the wealth transfer it gets will be smaller. Taking into
the benefits and costs of pressure, one can derive the optimal value of p,, :
any value for pr. This optimal level of pressure for group 1 is denoted tr
and is plotted in Figure 10.3. Here Ur(pz) is referred to as groups 1's
response function" because it tells group 1 what level of pressure is bes:
terms of u,elfare) in response to group Z's level of pressure. Note that di.r:
increasing in p2. The more pressure that group 2 exerts, the lower r,
influence of group 1. To ol{set this reduced influence, group l finds it t'F-
to invest more resources in greater pressure.
A political equilibrium is defined as a pair of pressure levels sucl,
neither group has an incentive to change their decision. ln other worJ,
pair (pi, Pj) ir a pohtical equilibrium if, given that group 2 applies pre,
l)i, th.r-, pi is the pressure rhat maximizes group 1's welfare cind, giver
group 1 applies pressure pi, then pi is the pressure that maximizes grc"':
welfare. A polirical equilibrium is then defined by the intersection of th .

#lu

-
 The Theor-v of Regulation 321

:isponse functions [Link],(p:) and t,(p,), as at that intersection both interest

lue* -\{odel :. are simultaneously optimizing. r2
. he political equilibrium has both interest gr(lLrps investing in pressure so
' Lnfluence the politicai process. The optimal pressure for each group rs
nuch dependent on the level of pressure exerted by the other group,
..e what determines regulatory policy is relatiue int'luence. As a result, the
-.pz) ::Jing problem inherent in ail groups is not as important as had been
usly thought. Since all groups are subject to free riding, what is important
. -ebtiue severity of free riding. When the free riding problem is less severe
-'ti@i
-- Lrp 1 than in group 2 (perhaps because group t has fewer members),
1 will have a relative advantage over group 2. This is regardless of
-:.er group t has a severe free riding problem in some absolute sense.
,.---Yz@t)
-.nother important property to note about the equilibrium is that it is not
' optimal. Both groups could invest fewer resources and achieve the same
of relative influence. Since relative influence is all that matters, the
I outcome would be the same but at a lower cost for both groups. As
: ,,mpl€, consider the case of competition among cable operators for the
,

: :elevision franchise in the New York City boroughs of Brooklyn, Queens,

Island, and the Bronx:

[franchise] applicants haue hired influential lawlers and public'relations

t, a roster of whom reads like a Who's Who rtf former citl and state
. . [A utce president for one of the applicants] contends that,.these friends
1.;;1LL (who tlpical\ command fees of about $5,000 per month) hst'e tended
wlcel one another out.23

. for influence in the political process uses up

:r,Ltion among groups
.urc resources to obtain the wealth transfer, resulting in a Pareto-
.:eater pfesSUfe -': - - -nt outcome. The logic behlnd this result is exactly the same as the
lply too much -.hind the Pareto inefficiency of the Cournot outcome in the oligopoly
dre greater the inI
ve influence of gn
. lsee Chapter 5).
..'en the establishment of a political equilibrium, we want to use it tt,
.r. Taking into ,., .
.:e testable hypotheses concerning properties of regulation. One important
:rrmal value oi: - jerived is that if the marginal deadweight loss from regulation increases,
::-rup 1 is denote: -
:le amount of regulatory activity decreases (measured by the amounr oi
to as grouPs ^ ,
transfer T). An increase in the marginal .leadu'eight lc,ss resulr. tn
. .l .,f pressure i' --
-:.-UIe . Nore thar - applying more pressure for any level of pressure for group 1 because
-.rst forego more wealth for any amount of transfer T due to the greater
- .'rerts, the lou -: . :.sht loss associated. Hence, group 2's best response function shifts from
ce, group I finds it o
:.,{r!(pl) (see Figure 10.1). With a higher marginal deadweight loss,
.--- . u'ill get a smaller wealth transfer for any given tax of group 2. Since
;ressure levels
&cision. In other wonh - has less of an incentive to invest resources to increase regulatory
':group2applies . rt will apply less pressure. This is represented by its best response
e shifting from rfr(P,) to U?(P:) in response to an increase in the marginal
- s welfare and, "-. '-'
rhat maximizes .: -::ht loss frorn regulation. As a result, the new political equilibrium is
- u'l-rich entails more pressure by group 2 and less pressure by gror-rp 1.
. -'
:l-re intersection
322, Cfurprer 10 Introcluction to Economic Regrrlation

Since T : Il(pr,p2) and it is increasing in p, and decreasir-rg in p], t

musr be true that ii(p!, p!) < I'(pi, pi). Hence, as meastrrecl bv thc a
of wealth transfer, rcgulatory activity is reducecl becau'.e of an increasc ::
marginal deadrveight ltlss associ:rted rvith it'
An irnportanr impiicatiitn of this resr-rit is that legLllat(lry ptllicies th-,-
welfare-improving are more likelv ttt be in'rplen'rented th:rn ones that arr
Suppose rhar industry A is a natural monopoly and indtrstr)' B is compe:
Tl-re dead*,eighr welfarc loss frttm regulatir-rg inclustry B is greater than
industry A, ceteris pirribus, because inclustry B is already irchieving a
oprimum *,hil. ir-rdr.o,ry A is r-urt. The irnplic:rtion of the abovc analysis r,
the greater n'rarginirl deadu,eight loss associ'.lted with regulation of rnclll''
*.ui. rhat more pressure will be apphed for regulation in industry A
industry B. The Becker model suggests that industries plagued by
failures (so that the marginal deadweight loss from regulation is relati'e '
or even negarive) are more likely to be regulated. The beneficiary group.
:

rvill apply more pressure' (

grenter potential fbr gains, so th:rt they
f,o.*..I by regultrtion will nor be l-rarmed as much because of ti-rc

deadweiglrt 1oir, ,n that they will apply less pressure against regulation
Ir., ."[Link].u.t ro the Stigler/Peltzrnan model of regulation, the Becker
provides a theoretical foundation for NPT. lyhere there are market fa
ih.." potential welfare gains from regulation. Some interest groups
"."
to gain iot f.o- regulation, while other groups stand to lose.a. little
(r
"
,u [Link]...r, groups in industries not subject to market failure) because
"relaiively
absence of large deadweight welfare losses. As a result,
relatively great pressure for regulation o{ industries subject to market
Ho*",r.r, ih" g..k"r model, in contrast to NPT, does not state that re
occurs or-rly *h.r-, there is a market failure. What determines regulatory
is the relative influence of interest groups, and this influence is det
not only by the welfare effects of regulation but also by the relative
of interest groups in applying pressure to legislators and regulators'

Taxationby Regulntian One of the perplexing aspects ofeconomic r

is the commo1 ,.rr" of cross-subsidization; that is, where the revenue
sale of, say, good 1 subsidizes the sale of another good, say, gtlod I
specifically, the price of good I exceecls its average cost u'hile,the pricc
i foll, b"lu* its marginai cosr. Such pricing bchar,'ior is perplexing, I
can be incctnsistent with troth profit mtrximizaticlrt and welfalre maxi
An explanation for cross-subsiclization was provided by Richar.l
an imporrant article in 1971.r+ He put forth thc tl-resis tl-rat or-re of tl-re
.rf regtrlation is to assist the got,ernment in its role of redistributing
In tlris iight, [Link].n is ilterpretecl as a me:lns for redi
wealth from One group of consllmers ttt a secrtnd group of cclnsu
example, pricc regtrlation entails charging a [Link] price for prcNi.l:'-
[Link]."^[Link]. Th.,s a conslrmer tvho lives in:r city, u'here thc r:.
.ort nf hooking him up to the system is very lor'v, pays the samc -
consumcr who lives in a rural area, where tlie marginal cost of 1-r'