DOCUMENT RESUME

ED 203 335 CS 206 365

AUTHOR Wirth, Michael 0.: Vollert, James A.,. TITLE The Effects of market Structure on Television News

Pricing.PUB DATE Aug 81NOTE 18p.: Paper presented At the Annual Me of the

Association for. Education in Journalism h, EastLansing, MI, August B-11, 1981).

EDRS PRICE MF01/PC01 Plus Postage.DESCRIPTORS *Advertising: Broadcast Industry: Cable Television:

-)

*Commercial Television: *Competition: FederalRegulation:. Multiple Regression Analysis; News Media:*Television Commercials: *Television Research: *UnitCosts

IDENTIFIERS *Market Value

ABSTRACTMultiple regression techniques were used to examine

the business side of local television news operations for November1978..Research questions examined the effect of several variables onlocal television news prices (advertising rates), including type ofownership, network affiliation/signal type, market size, cablenetwork penetration, market concentration', and market prosperity. It.was expected that higher levels of market concentration (making themarket less competitive) should lisAd to. higher news prices. Data forthe study came from 297 commercial.' television stations in 105'television markets. The.results conformed to theoreticalexpectations. Specifically, ownership type (group or cross media) had'little effect on television news advertising prices: cable networkpenetration had a significant negative effect on such prices: andmarket concentration-had a significant positive Impact on price. Theprimary policy recommendations suggested by these results are thatthe Federal Communications Commission (FCC) should continue to pursueprocompetitiie policies that are most likely toincrease the numbetofitelevision channels. from which consumers can choose: and that !"`regulation of the commercial television industry with respect to,theFCC's scheme of taxation by regulation (public interest program/requirements) must be discontinued as competition is increased.:-(Author/RL)

********************************pg*******************Sig****************Reproduttions supplied bp,; ORS are the best that 'can be made ,*

from thei;Mt*ginal document. *

e r

flo

44.a. urommaar OF EDUCATIONNATIONAL INSTITUTE OF EDUCATION

EDUCATIONAL RESOURCES INFORMATIONIftCENTER IERICI

his document has been reproduced asreceived from the person or organizationoriginating it.

I I-Minor changes have been made to improvereproduction quality.

Points of view or opinions stated In this docu.ment do not necessarily represent official NMposition or policy.

a

I

THEORY AND METHODOLOGY

THE EFFECTS OF MARKET

STRUCTURE ON TELEVISION

NEWS PRICING

By-

Michael 0. Wirth and James A. WollertUniversity of Denver and Memphis State University

Denver, Colorado

"PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY

Mirh.9131 0. Wirth

Jame-A. Wollcrt

TO THE EDUCATIONAL RESOURCESINFORMATION CENTER (ERIC)."

41:

(

Presented' to the Theory and Methodology Division,Association for Education in"Journalisrn, annual convention, Michigan StateUniversity, East Lansing,1,11,,August 1931

INTRODUCTION

The mafket.for local teleleision.news provides the main source of

competition-among localtelevision stations. Thousands'of d011ars are

spent.annually by local stations to get, the advice of news consultants

regarding who to put on the.air, what kinds of stories to cover, and how

to provide broadcast coverage of those stories selected. Likewise, many

stations have recently made the large capital investments needed to convert

lrom film news equi6ent to electronic news gathering system's (ENG). From

an audience perspective, local news cdnAtitioteS the most important social

1 benefit which.has been derived from the Federal CoMmunications Commission's

(FCC) local system of.broadcasting. "However; the market forjoal -

television news has rece,iveit.vety little'attgptIon when it comes to economicy.

,

research.

Paqt economic studdes,.which haVe focused'on local television news,.

1-were conducted by Litman -and Prisuta

2. Both of these studies centered on

the development of a market abate instability'nstability model to explain the observed. A.

variation in local televisi n news market shares.. Both agree that the local

_ ..

television news market constitutes an oligopoly (there ate few enough sellers,,

such that they recognize their mutual dnterdependence. Litman concludes tha

market share instability' s strongly related to,the,numbercif market competit rs1

and that programs preceding the news create a strong audience - fl9w effect3

.

1.-

-. .c, 'A./

The present, study is the fist to examine the 'business. sides L/tif local, .

television news. Since ielevisiOn audiences are not able to form individual

contracts with stations affering local newscasts; the relevant price to be ,

examined is the price charged advertisers for purchasing access to television

'-

new ienCes4.

: Past studies of television station pricing have been,' ,

.','

conducted by Owens, Lago6

, Peterman1,- and Mirth and Allen

8. 'All of

.

thete.. .

.

studies examined station=level pricing-during prime time,andfor from .9 a.m. to

rs

Midnight. The FCC's Crossmedia Ownership Itulemaking9provided the primarx

focal point around whicireach of these studies centered. The major source

a 'dontroversy stemmed frOm whether viewing audiefnce is or is not an exogenouslY

10determined initependent variable . Owen argued that audience is an endogenous:

11.variable and that it should not he included in a reduced form price equation .

Lego12

and Peterman13

.disagreed. Wirth and Allen argued that when local pro-

grqmming is being examined, audiehce.should be treated as an endogenous

.variable; when national progra4ming is the main focus, however; audience is

mord properly viewed as an exogenous variable14

.

.

In the present ca.e, local

news audience should be viewed as an endogenous variable and shoulde excluded

from all reduced form estimates. However, since its exclusion (inclusion) is

gontroversial,,the regression's provided.herein alternately include and exclUde

news audience asa Predictor variable.

This.article examines the following research questions concerning local

televisiOn'news pricing:.k(1).Does the type of ownership which characterizes a

particular station have an effect on. local television mews prices? (2) Does

network affiliation and/or, .Signal type affeCt news prices? (3) What kind of't

impact does market'size have on 1°41 news prices? (4) Does cable penetration

have gn-effect on local news'prices?'.(5) Does market concentration have an-

,impact on. local news prices? And (6) What kind of impact does market prosper6

have on, lochl television news prices?

The basiC theoretical expectitiOils are simple. It islOxpected that

4 station of market level characteristics which are-associated with higher levels

of market concentratiout(in other words the market is less competitive) will

lead to higher-news' prices-all other things equal.

The next section'describes the methodological approach taken, specifies

the model used, defines the variables included and provid's re4 levant :hypotheses.:

This is followed bya resultspection. The parer c.anclndes with a discussion.

of the study's policy implications: 4 .1

',METHOD

a

This,cross-sectional analysis focuseS on 1978 local television news list

prices. Data were obtained from a variety of industry sources. Ordinary least

squares (OLS) multiple regression techniques were used to analyze the news

pricing behavior of local stations. A random start, systematic samplingI

4

technique with a skip interval of two was utilized to provide a sample of 105

TV markets; data were included from 297 commercial TV. stations.

Two functional forms were used to analyze news pricing at noon, in the

.early evening and in the late evening: linear and ddublelog. The doublelog

functional form is the preferred form for

-foremost, utilization of the doublelog form minimizds heteroscedasticity

a variety of reasons. First and

r problems. Since the data being analyzed represent both large and small

1.

.

televisiOn stations, a linear analysis cannot be expected to have a homosce--L

, .

I'dastic error term15

. In the prdsent case, a doublelog functional fOrm is also

preferred because it aFeounts for more of the variance of the dependent

variables. Since there is no a priori problem with utilizing a doublelog,

form, only our doublelog results are repoTt.ed het-et

The general format of the model can be expressed:

(1) Yi = f(Y.1' 3(k,

) + ui

where Y. are the news price variables, the XI( are exogenous, and the ui are1

random error terms. Some of theXk are continuos; others are.bina y variables.

,

The form used to estimate the relatiOnship betweerilocal television news prices .

1

. 1._

and a number of different station and market level characteristics was:-...,

(2)1n(Ygi )=ai +811n(CAM.Eix100)+b21n(HERF.x100) + b3ln(CSIHH

i/100)

1

+ b41n(AUDHH./11000) 4- b TVRAD

1+ b

6 iGROUP + b NP. +A XMED1A

1 5 . 7i8 i

+ b9 iABC -+ b

10CBS

i+ b

11NBC

i+ b

12VHF

i+ b

13R10

i+ b

14R25.

+ bi5R501: b16R1,00i + bi7R15.0i +

3 - 4

Where Yi rerresents

NOON NEWS PRJCE

4:ARLiiVENINGNEWS RICE

..

LATE EVENINGNEWS PRICE

18and where:

CABLE

HERF

CSIHH

AUDHH

TVRAI)

GROUP

4

the following dependent vartables: 17

The pricestation's

The pricestation's

The price'station'S

of purchasing A 30 second spot on the ithmidday news.report.

.

v.

of purchasing a 30seco.nd spot on .the. ithearlyevening.news.

of purchasing a 30 second spot on the ithlate evening news:-

. t

Csible peiietr tion in the ith, station!s.Area of Dominant ,

.

). .

.

Influence ( DI) of operation. An'increase in a market'scable. penetration should lead to an increase in thecompetition faced by local stations. Consequently, an.inverse relationship. should exist between cable penetrationand prik.19,

Each market's Herfindahl,Index. -Since,larger Herfinda1ll-Index values are associated with morehighlY concentratedmarkets, a positive relationship should exist betweedAhemarket'Herfindahl Index and the ith-stationiaprices.

--..-

fotal consumer spendable income in the ith ADI per ADIAusehold. Income per household should, be -

related to price, since advertiSers'shouid value richeraudiences more than poorer ones.

zl _

Average quarter hour hou sehold audience for the ith",station's Monday through Friday local news broadcasts(noon, early evening, and Gate evening). This variableis alternately included and Occluded from the'analyses.When it is included, a positive relationship 'should °OcistbecauSe advertisers can be expecd to pay a higher.totalprice to reach larger audiences.,

A dummy variable with a value of 1 if a television stationowns a'radio station whose city of license is in the /msADI, 0 otherwise. Televisioastation ownership of a radiostation in its .market of operation shouldincrease the'firm's market Rawer and J. therefore expected to lead tohigher prices.

A dummy variable ,with a value of 1 if TV station,is,owned by an entity Tiich owns at least one other TVstaliOn, 0 otherwise. The effect of grew ownership cannotbe predicted B. priori.2

4

rit

4,

NP A dummy variable with a value of 1 lf a TV stationis owned by a daily newsp'aper, 0 otherwise, Theeffect of newspaper.wner5hip cannot be predicteda priori,25

XMEDIA - -A ,dummy-variable with a ligue of 1 i.Pa TV stationis Owned 13S, a daily*newspaper which is publiShedin-ti same Area'of Dohinant Influence (ADI),.0 otherwise.In.theory, televition,stations which are owned by '

daily.newspaperl in the same market should possess alargerdegree.of market power than do. the stations withwhich they compete. 'his should lead to higher prices:26

ABC - A dummy variable with A value of 1 if a TV station is,)

affiliated with the ABC Televis.Sibn Network, 0 otherwise.4r,

) , -

CBS .17., A dummy vari* le with avalue of 1-'if a'TV-Statipn is

I, ,Ifiliated w th the CBS Televisi Newtork, 0 otherwise

..), ..

NBC -.f..-li dummy vafialYle with a value, o, 1 if,a TV station is'affiliated with ,the NBC Television Netwoik, 0 otherw se

, 'Ea01.of these .0ariables should have a positive impac onlocal.news p /ice since they are beinglcompared to th

ce priceicharged by independent stations.':?f .

.,

.VHF A dummy variable with a value of 1 a'TV'st dim is a ,

VHF sta ion, O'otherwise, VHF stations shoul ,have "

signifiC ntly'higher prices than their UHF 'cdunterparts'since VII V-Iold a si nificot technical advantage (*ettimos.28

R10

R25

.4 41 /4,

- A dummy.variablewith.a value of 1 if a tV station isin the top 10 telev sjien markets,. 0 otherwise.

---".-4.-___)-

A dummy variable with a.1 value Of 1 if a TV( station islocated injparkets 11 to 25, 0 otherwise..

R50 ledummy4ariable with a value of 1 if a TV statibn islocated in,markets 26 to 50, 0 otherwise.

.,

R100 Adummy variable with a value of 1 if a TV station is-, located-in markets 51 to 100, 0 otherwise.

A dummy,variable WItha vane of 1 if a TV station islocated in markets 1,01 to 150, 0 otherwise. It isexpected that an increase-it market rank (towards the

.

top 10) will lead to an increase in price.29

u is a random erior"term, assumed to be distributed normally with a mean of.zeto.

R150

40'4-

0.

This analysis. is primarily concerned with the price effects: (1). of cable,. . r

television penetration, (2) of increased Maqe concentration, and (3) of

Various types of station ownershii. The results from testing thesestimating

equation and the preceding hypotheses are.detailed in the next section.

4

RESULTS.0"

Variable means and standard deviations are provided for a number of

important local news related variables (see Table 1). Both the average total

4list price and the overage cost per thousand householdOincreases as the

) broadcipt day progresses from noon news to late evening news. The average list

price costiper chousand'households found in Table 1 would seem to be on t he

high side:, This probably results for two 1$easons. First, the prices found

in Standard Rate and.Data Service are list prices which could be expected to

exceed actual local news transaction prices.' Second,.adVertiSer demand for_

access to local news audiences can be expected to exceed the numbeJ of access

opportunities available. This should result Itihig4r costs per thousand and

higherlorices:. Further inspe tion of Table 1 suggests that there is a great

deal.ofiivar bility present i each of the news specific varfiables provided.

The. results from using

are found in Table 2. Si a doublelog funetidpal form was utilized,

the nonstandardized regression coefficients provided are directly interpretable

data to estimate six news Price equatio

1

as elasticitie,30 . The average quarter hour .114ehold'audience which4.

corresponded tp a.given news program was alternately 1nel/tided and 'bxcluded.e"p

One hundred and thity staeions were includedin the noon new&,pride analysis;

239 in the early'even ng analysis: and 229 in the late evening analsis.

1

111

TABLE 1

Local Televion News List Price Variable Means and Standard Deviations

Variable Mean'

Noon News Price .-, $126

Noon News Audience (HHs) 32,300

CPMHH for Noon News $5.63

Early Evening News Price $268

StandardDeviation

$146 134

32,400 134

$6.52 134 '

, $400 248

Early Evening News Audience(HHs) 53,40U 59,000 \. 248

-..

- CP,M gor Early Evening News .$6.97 $8.00 248

Lae Evening News Price $325 1 $595

A

Late Evening News Audippce(HMO. 48,500 r 60,8

,

6MHH for Late Evening News $6.99

Wp

rP 9

$3.93

4

ID

236

236

236\

/4

Lnsprction of the fesults provided in Table 2 suggests that including

news audience .as a predletorNariable results in major conflicts with respect

to a number of hypothesized.relationships. Most notably, network affiliates,

VHF stations, and ratio s operatini 'outside of the Top 50 television markets

all appear to charge lower prices who the endogenous audience variable is

included as a predictor variable. Likewise, market concentration, as measured

by the Herfindahl Index, is much less of a factor, and actually has wosigni-,..e

ficant.negative effect on the price charged at noon. The differences found may

in fact be true one since the primary reason for expecting positive relation-,

4:, ships between price, network affiliat7iou, VHF signal type and market size

stem fro an expectation that audience size will be larger ceteris paribus.

However, if one agrees with the theory that local news auVence is an endogenous

variable with respect to local new price, the coefficients provided in the

Table 2 regre4sions which include audience as a predictor varlable.are biased

and inconsistent,31

. ,the rest of the discussion focuses on the

correctly specified redUced form estimates (those without audience).-!

Overall, the lion-audience news price models found in Table 2 yield good

estimates of individual station prices. This contention iS'rsupported by the

high R2's associated with each equation estimated (.68, .66, and ..73) and the

low.coefficients of variation 40.1 percent, 9.8 percet, and 9.7 percent).

Additionally, thp non - audience equation estimates provided are generallyJeci,

consistent with our hypotheses. '

Most notably, a one percent increase in t9,1)1 cable penetration caused

'1978 local news prices to fall by .21 to .27 peftrent. This result suggests

that cable television-systems are having and will -have significant negative

effects on broadcastreveaues (and ultimately profits).

. TABLE 2

Regressions on Television News Advertising Prices:Doublelog Fonn*(t-statistics In Parentheses).

flip<.1 for a 1 talltest; bp<.05 for a I

--Deundent Variables Ad Price Ad PriceNoon News Noon News

Pred ic tor VariaViZn--_

. I

-.03 -.05(0..42) 0.52)

.09 .15n

(1.00) (1.49)

-.09 -.11(1.02) (1.08)

Same-Mkt NeWspaper-TV dombo -.02 -.07

(0.19)

1

(0.46)

ABC Affiliate .02 .34a/ (0.10) (1.44)

.46b;CBS Affil ate .05

(0.245 (2.10)

TV-Radio Combo

Croup-Owned

Newspaper-Owned

NBC Affiliate

VHF

t TOP 10

11-25

26-50

51-100

101-150

Cable Penetration

Herfindahl Index

.41g.41-.'02

(0.08) (1.86)

-.19 .14

(1.25) (0.89)

§3b

1.53c

(2.17) (5.25)

.15 .94c

(0.61) (3.91)

. 06 .73c

(0.30) (3.40)

-.31b

.19(1.74) L (1.00)

-.34b

A -.07(2.06) (0.36)

-.21c F .25c

(3.00) (3.05)

-.20c. -.05(2.50) (0.60)

Consumer SpendableIncome/HH -.02

(0.06)

No. of HHs Viewing' .38c

(6.31)

Constant Term, 5.00

R2

.761.

Standard Error of Est. .394'

Coeffic' n 'Variation .087

130

.24

(0.55)

3.251

.686

.457

.101

130 239

tail test: cp<.01 for a 1

Ad Price Ad PriceEarly Even- Early Even-lug_ News irlf News

.03 .08

(4.55) (1.08) '

tall test)

,Ad PriceLate Even-in_g,_ News

--.08"

(1.47)

Ad PriceLille Even-

_ntligLiews

'-.05 -

(0.71)

.04 .09 -.02 -.01(0.53) (4.11) (0.41) ' (0.12)

.02 .05 -.09 .04.

(0.31) (0.54) (0.14) (.42)

ik'..-

-.002 -.09 -.01 -.178

(0.0Q) (0.71) (0.06) (0.58)

-.33 -.24 -.24 .3,0

(0.73) 7 (0.46) (0.86) (0.78)

-.20 . -.003 -.14 .42(0.44) (0.00) (0.50) (1.09)

-.24 -.03 -.26 43(0.52) 0.06) (0.92) (1.13)

-;05 .22c -.11a

.31c

(0.54) 2.37) (1.42) (3.20)

.98c

2.33c

.59c

2.30c(3.75) (9.35) (2.70) (9.23)

.48c

1.53c .27a

1.47c(2.32) (7.81) (1.58) (7.42)

.20 1.11c -9 .99c

(1.14) (6.62) (0.62) (5.76)

-.05 .69c -.20b1

.53c

(0.33) (4.88) (1.72) (3.65) t

a-.19 .

c32 -.31c .10

(1.49) (2.47) ,p2) (0.72)

-.14c -.21c -.14c -.27C/(2;53) (3.1Q) - (2.95) (4.(91

.08 .36c -.03 .22c(1.10) (4.97) (0.52) (2.98).

--.04 .02 .39a .83c

(0.14) (.05) (1.57). (2.46)

.45c .67c

(8.92) (13.90)

4.13 3.55 2.00 -.10

.753 .664 .860 .732

.437 .508 .368 .508

.084 .098 .070 .097

239 229 229

IThe market concentration results (lirfindal Index) found in Table 2

fulfill our expectation that nn lrcas, in market concentration mit! lead

to an increase in price.. Market concentration had no effect on noon now

prices. However,' a one percent -Increase in market concentration (te

Herfindahl Index) resulted in .36 pefcent higher early evening new prices;

and .22 percent higher late evening news prices. Likewise;, higher Income

audiences appear to be worth more to late evening-news advertisers. A one

percent increase in a market's consumer spendable income per household led

to akincrewie,in late news prices of .83'percent.

C)

As for the wnership.type variables examined (TV radio combinations,

group-owned statibns,newspaper-owned stations, Ind same-market newspaper-TV

.

combinations), there is little to report. Television stations which own radio

stations in their market of operation charged somewhat lower (but not signi-

ficantly s "priCes,at noon and in the late evening and somewhat higher prices

for the earl ,evening news. Group -gowned stations charged 15 percent higher

prices at noon than did all other stations.and nine percent more for early

4'evening news spots. However, their late night prices were no different from

all other stations, ceteris varills. 'Newspaper-owned stations had somewhat

lower prices at no no re difference existed for early evening and.late

evening newsca s. Finally television stations which are owned by daily

newspapers i the same,ADI charged lower prices (but not significantly so),

at noon, in. the earl evening and for the late evening news.

Network affili ion had the expected effect on news prices'except'in the

early eIrni w ere no difference was found between the price charged by

atfiliate and those charged IRY independents. the major caveat which exists

a

with spect to our affiliation results is that only a small number of inde-t

pen nt stations were included in the finaj estimate of each equation (six in

the no analysis, one in the early evening analysis and two in the late

evening ana lysis). Cohsequntly, the most important affiliation results are found

12

by comparilg ABC affiliates toCBS affiliates to NBC affiliates. The/

results coAtained in Table 2 suggest, that news advertising prices wee

highAse do CBS and NBC stations in 1978.. _

a.--,

Ni_Finally, the signal type and market rank coefficients confl-ito our

, xpectationst VHF stations charged significantly higherfticeS 41,Aeanly4 6

1

.0 . s. .

evening and late evening news advertisers than did UHF'stAions (22 and 31.. ,

49

p rcene-more respeltivelx). Likewise; stations operating in larger markets)

. consistently charged advertigers higher prices for access to their audiences.1,T.

. ..)

., . , vFor examplestations,operating in the Top 10 television 4arke4s chArged

. ,

.1.

advertisers 230 percent higher rates in 1978 force late evening news spot .

.0.

, ) . il

than did stations operating in markets 151+ In markets 11 to 25, the, 1

'differential was 147 percent. Thel;est of the.

coefficients are interpreted

in a similar manner:

1DISCUSSION

A number of policy quegtions were posed at the beginning of thig article.

These questions can now be answered in a definitive manner. "Four types of

.

ownership were examined. Only in the case of group-owned stations was a

significant relationship discovered: that group-Owned stations charged

<:10

*significantly higher noon news prices in 1978.

On the other hand, significant relationships were found between the

remaining policy variables and the price charged news advertisers. Specifi-

cally, cable penetration was found to be inversely related to news prices and

market concentration was found to have a'pasitive effect on price.

The former results suggest one of two things. Either same-market cross-

media combinations do not possess more market power than all other stations,

ceteris paribus, or they were not exercising,,the extra market power which they

Al possessed with respect to the prices charged local news advertiserk in 1978.32'

13 (

The analysis presented herein cannot determine which of these represents the

true market condition. However, these results clearly suggest that the4

prices charged by the ownership types included in this study were consistent

with the prices being charged by all other stations. In other words, no overt

exercise of power pver price was found with respect to the ownership types studied.

The cable penetration results are consistent with the audience diNkersion

studies submitted ag-part of the FCC's Cable Television Economic Inquiry.33

.

There is no question but that incr ses in cable penetration, all-other things,

equal, result'in reduced broadcast audiences. The results of this,study.

,provide an estimate of the. relationship between cable penetration and the

priCes charged to gain aotess to, the audiences produced by each station'simost

important type of local program. Clearly, the negative effect'cable television

has on broaddast audiences results in lower advertising prices (and ultimately

loWer revenues).34

No broadcaster has yet been forced off the air because of

cable. television. However, as cable penetration increases, broadcasters can be

expected to achieve levels of profitability which are more consistent with normalits

economic returns. As cable competition reduces the excess profits earned by

tfc

broadcdtters to normal levels, the FCC's taxation by regulation scheme will have

35to be modified.

c.

The market. concentration results indicate that stations which operate in

markets with less direct television competition and/or with more uneven competition

charge higher news prices,tall other things equal. Telecommunication policy

makers haye been struggling with this problem for many years. Since only a

limit number df over-the-air commercial television channels have proved viable1.

given the FCC's policies and table of allocations, the primary method forY

reducing t'L undesirable effects of television market concentration, inwasing

'0.th number of competitors, seemed out of reach. Yecentlyi2 however, the FCC has

taken a number of steps'which should increase competition in television markets

12

in the long run. Examples of such pro-competitive policy include: a

laissez faire approach to cable television, allowing for the licensing of

over-the-air pay televisio4 n st/tions, allowing for the development of thousands

of low power televm on stat&s, and considering the use of direct broadcast

satellites for program deli ery to the home. Such policieg should redUce the

market powir possessed b any given television firm. Sowever, as indicated

in tffe discussion of the cable penetration results, the FCC'should not be

allowed to indefinitely maintain its scheme of taxation by regulation as it

increases the compe4ition television broadcasters must face.

7 /

c".

a

FOOTtibiES

1Barry Litmn, "Market Share-Instability in LoCal Television News,"

Journal of Broadcasting 24 (Fall .1980), pp. 499-514.

2Robert Prisuta, "Local Television News as an Oligopolistic Industry:Pilot Study," Journal of Broadcasting 23 (Winter 1979), pp. 61-68.

.3-Litman, op.. cit.., p. 512.

44See generally, BruQe-014en, 'Jack Mebe and Willard Manning, ' TelevisionEconomics (Lexington, Mass..:. D. Heath 'and Company, 1974, p. 10.

5Bruce Owen, "Newspaper and Television Station Joint Ownershipi" AntitrustIF Bullptin 18 (1973), pp. 787-807.

6A.M. Lago, "The Price Effects of Joint Mass Communication-Ownership,Antitrust Bulletin 16 (1971), pp. 789-813.

7J. L. Peterman, "Concentration of Control and the Priceof Television Time,"

American Economic Review 61 (1971, Papers and. Proceedings), pp. 74-80.

8Micilael 0. Wirth and Bruce T. Allen, "qossmedia OwnershipYRegulatory

Struiny, and Pricing Behavior," Joulrnal of Economics and Business 33(Fall 1980), pp. 28-42 r

9Federal Communications Com ion, Notice of Proposed Rulemaking 45 Fed. Reg.

5963 (1970).

10See Michael Wirth and Bruce Allen, "Another Look at Crossmedia Ownership,"Antitrust Bulletin 24-(1979), pp. 88-90 for a brief discussion of thespecific issues involved.

11of. cit., p. 7,91.

12tago,El. cit., pp. 790-792.

13Peterman, Ea. cit., pp. 77-78.

14Wirth and Allen, 1979), Ea. cit., p.09.

15R. S. Pindyck and D. L. Rubinfeld, Econometric Models and Economic Fore sts,

(New York: McGraw-Hill, 1976), pp. 17-18.

16.The AUDHH variable is presented as part of the overall news price model even

though its values are alternately included and excluded in the analysis.

1 T rNa.

List prices for each_skf .the depedent variables were obtained from StandardRate and Data Service: Spot Television Rates and Data. When rate cardswith frequency discounts were encountered the one time only spot rate wasutilized.

4

1

14

'6

18.'The 1974 Broadcast Yearbook was utilized as the main data source for thepredictor variables included in the news price models. The only variableswhich did"not .cdme from thii source were: each mAket's Consumer spendableincome per household, which came from a 1979 edition of SRDS: Spot

Television Rates and Data; each market's cable penetration, which came fromthe_1979 Television Factbook; and the news audience variables and thHerfindalh Index,wtich used dap obtained from individual market sufeeps byARBITRON TV (November'1978).

19This hypothesis finds support in the audience diversion studidone by

Gdorge.R. Schink and Sheela Thanawala, The Impact's of Cable TV on LocalStation Audience National Association of Broadcasters Final Report(March l9787; Also, see Comments of the National Cable TelevisionAssociation (March 15, 1978), FCC Docket No; 21284; -For more recent eNtidenceof the effect of cable television on those markets hardest hit by cable'.competition, see Paul" Bortz and Michael Winth,iet. al., The Economics ofTelevision Station Operation in 100-Plus Markets Nltional Association of

-'Broadcasters Final Report (February 1981).

2dEach market's Herfindahl Index was calculated by using the following formula:

H = E S.2

where Siwas the audience (HH) share of the ith station from

i=1

9 a.m.. to Midnight in November 1978. When an ADI was occupied,by'emly one -'

TV station, the index attained its maximum value of vi.O. The value declineswith increases in tie number of stations and increases with rising. inequalityamong any given nurnBer of stations. See F. M. Scherer, Inchistrilal Market

Structurie and Economic Performance (Chicago: Rand McNally, 1970), pp. 51-52for a complete explanation 'of how .the index works and-for support, of the

stated hypothesis.

21 Support for this hypothesis is found ii Owen, cm. cit., p. 806; Lago, cla. cit.,p. 802; Peterman, cla. cit., pp. 77-79: and Wirth and Wen, (1980), Ea. cit.,

pp. 36-37.

22This hypothesis is consistent with the results reported by Lago, cm. cit.,

A

p. 802; and Wirth and Allen, (1980), .11a. titI'

p. 38:/

23See Wirth and Allen, (1980), lap. cit., p. 41.

ier24See Barry Litman, "Public Interest Programming and the Carroll Doctrine:A Re-Examination," Journal of Broadcasting 23 (Winter 1979), pp. 56-57; Also

see Michael Wirth and James-Wollerts ,"Public Interest)Programming: Taxation

by Regulation," Journal of Broadcasting 23 (Summer 1979Wp..323.

25Ibid.

J,

26All of the'previously mentioned eeonpmetric'stud4s concerned with this issueadopted this hypothesiS. Only_in the case of Owen, cm. cit., p'. 806 wasthis hypothepis found to be empiriCally valft. Wirth and Allen, (1980),

op. cit., found a negative re]ationship. After testing a number/ of

different theories; they reached the conclusion that regulatery scrutiny

had caused crossmedia owned stations to hold the line on prices ininflationary times and that this had resulted in lower crossmedia prices.

1

1715

0 4.

4 '1 ..

27This hypothesis is supported by Owen, RE.

.'

"cit., p' 806:.Lngo, ox.. cit.,

F. 802; anck by Wirth and Allen, (100) , ,s12! cir; ,p.

4.

28' -

All of the previously mentioned studies which focused on*

televisionpricing in the context of fhe erospmedia question support this hypothesisL. ,

. ,

r 4 \,

29See Wirth and Allen, (1,980), cap, dit., p. 36 for Support o this

.fl

, .

30Regressioncoefficients whic are obtained from dolublelog functional forms,

,allow us to directly sw r the question, what is the percentage change,In ,a dependent variable given the percentage change ima given predictor'variable, all other things held constant.

.31See Pindyck and Rubinfeld,,RE. cit., 132-137. An additional course ofaction will be pursue.in.future re eSrcil to.eliminate the problems associa:'ted with using audience As a predictor variable. To accomplish this, arecursive' equation system Will be developed. -. ,/

° ..

,

32See Wirth, and Allen, (1080) op. eit' pp. 40-0 for disdussion. o f these'

3Fed)61 Communieati stommissioil,' I wiry Into the Economic ReqationshipBetween Televisi Bibadtasting and .able ''Television (1970, Docket No.. 21284:,

34See Schink and Thanatralgc.22.'cit.; Commints of -the National Cable Televisio9Association, Op, cit.; and BortzLanclih, Ea. cit.

35See Carroll Broadcasting Companyv. Federartorknmnientions Commission 258 F.2d

440 (1980);'llichard.Posner',4"Taxaeion by Rdgmlation," Bell Journal of,Economicsand.Management Science 2,(1971), p. 22 provides a,completediscussion of taxation by 'regulation in general; and Wirth and Wollert:, per. cll.

issues.

16

1s