Transcript slides

What influences advertising price in television
channels?: An empirical analysis on the Spanish
market.
Germà Bel & Laia Domènech
(Universitat de Barcelona & ppre-IREA)
December 2008
2008
Introduction
•
Motivation: Changes in market conditions: new operators, financial deficit
incurred by public operators, development of new tecnologies and new
regulatory framework.
This situacion has increased pressures in advertising market, the main source
of revenue for public and private broadcasters.
•
Aim of the paper: We empirically analyze the factors explaining ads price in
free-to-air TV market.
•
Contribution:
1.
We consider how ad prices are affected by factors such as:

Ownership

Contracting out
2.
Provide new evidences about the relationship between ad price, ad time and
share.
3.
We study the Spanish case.
1
Plan of the presentation
1.
Motivation
2.
Aim of the paper
3.
Relationship to the literature
4.
Data
5.
Methodology
6.
Estimation and results
7.
Concluding remarks
2
1. The free- to-air television market in Spain
The free-to-air market in Spain is undergoing quick and intense change.
The main characteristics of the free-to-air television market in Spain are:
a) There are public and private operators in this markets. Television channels
can be regional or national.
•
National TV channels compete with the regional TV channels for advertisers.
•
Regional channels have an important weight in the market.
•
Some of they use in their broadcasting other co-official languages different from
Spanish.
These network enjoy a certain degree of monopoly over their
viewers (there are not relevant competing channels in the same language)
b) Many Television channels (public and private, regional and national)
contract out the management of their sales of ad time.
How do the channels sell their ad time?
»
Channel sells ad time to advertisers.
»
Channel contracts out to private firm the ad sale management.
3
1. The free- to-air television market in Spain
c)
Significant financial deficit incurred by Spanish public broadcasters, national as well
regional.
For this reason, several
reorganization plans.
governments
have
designed
and
implemented
The financial stress is due both to the inefficiency of the existing financing system
and to a loss of audience.
Funding of public service broadcasting (SPAIN): Advertising and budgetary
founds. (No fee)
Consequently, public broadcasters compete with private television channels for TV
advertising.
Crowding out effect.
d) The number of private channels in Spain has increased in 2006.
New private national channels: Cuatro and La Sexta
Conclusion: This, together with the financial stress placed of public operators, has
increased pressuresnin the advertising market.
4
1. The free- to-air television market in Spain
This mixed oligopoly type of market has some peculiars characteristics:
There are public and private operators
There are regional and national operators, and
Small percentage variations in the share level translate into large effects on revenues
from advertising.
5
Source: Comisión del mercado de las telecomunicaciones (2007)
2. Aim of the paper
What influences TV advertising price?:
The effect of television channel ownership on advertising price.
The existence of externalities.
Consumers don’t want to consume advertising. (Negative
externality)
Advertisers are willing to pay more for high level of audience.
The relationship between price and management of their sales of
advertising time.
The relationship between price and specific characteristics of
programming and viewers.
6
3. Relationship to the literature
• Recent literature has focused on the Two-Sided Markets concept. (Armstrong 2006)
(Rochet y Tirole 2003).
• This concept has been used to understand the television market.
• Anderson & Coate (2005), Reisinger (2004), Bel et al. (2007) and King et al. (2008)
have offered recent studies focusing on two sided markets in television.
• We understand Two-Sided Market as:
Advertising (-)
Viewers
Programs
Attention (+)
Advertisers
Pay for advertising
7
3. Relationship to the literature
Variables
Share
Hypothesis
Positive relationship
advertising price
Minutes
Negative relationship with
share and price.
Contracting
out
Positive
price
Theoretical basis
Share generates a positive externality to
advertisers. The ad time generates a negative
externality to share. Cumminghan & Alexander
(2004), Anderson & Coate (2005), Reisinger
(2004).
Empirical evidences
Positive relationship with price, and negative with ad
time. Elaticities between 0.83 and 1.4. Goettler
(1999), Wilbur (2008), Brown y Cavazos (2005),
Brown y Alexander (2004) y Kasuga y Shishikura
(2006).
Negative relationship with share and price. Wilbur
(2008).
with
Generates a negative externality to share.
Negative relationship with ad price. Anderson &
Coate (2005), Reisinger (2004) and Bel, Calzada
& Insa (2007)
Not consider in the literature
ownership
Differences
between
commercial
and
public
channel Important variable
on price
Differences between commercial and public
channels.
Hansen & Keiding (2006) and Kind, Nilssen &
SØgard (2007)
Demographic
variables
Demographic
affect to ad price
variables
No specific hypothesis in the literature
Economic
variables
High per capita income
generates a positive effect
on ad price.
No specific hypothesis in the literature
Differences between commercial and public channels
Alcock & Docwra (2006), Delaney & O’Toole (2006),
Wurff & Cuilemberg (2001) & Withers (2001).
No differences between commercial and public TVs.
Tsourvakas (2004).
The literature consider demographic composition
Goettler (1999), Wilbur (2007), Hammervold &
Solberg (2006), Withers (2001) yBrown & Cavazos
(2005), Brown & Alexander (2004) and Kasuga &
Shishikura (2006).
Economic variables used in the empirical literature.
Withers (2001).
Brown & Cavazos (2005), Brown & Alexander (2004),
Kasuga & Shishikura (2006) show a positive
correlation between ad price and income.
relationship
with
Not considered in the literature
8
3. Relationship to the literature
According to the literature, we consider how ad price are affected :
Advertising minutes.
Audience share.
Viewers characteristics
In addition, we examine two effects :
The effect on prices the existence of a public operator.
The effect on prices that one channel contracting out the
management of their ad time sales.
9
4. Data
We want to explain TV advertising price. To conduct our estimation we use panel
data, with semester data, for 2002 thought 2006, and for 13 broadcasters (public and
private).
Variable
Source
Advertising revenues
Infoadex
National share
TNS-SOFRES
Advertising minutes
TNS-SOFRES
Percentage of population 14-29 yeras old
INE
Percentage of population 30-44 years old
INE
Percentatge of women
INE
Regional groos domestic product per
capita
INE
Football
Channels
Percentage of entertainment programs
TNS-SOFRES
Percentage of cultural programs
TNS-SOFRES
10
5. Methodology
The methodology used in the paper:
• Parametric analysis:
We estimate the price equation by means of two-stage least square estimator.
The model is estimated in two stages to avoid endogeneity problems caused by advertising
price, share and advertising time.
• Non- parametric analysis: median spline
Median spline between share and ad time
10000
0
0
5000
10
Median spline
20
30
15000
Median spline between share and advertising price
0
10000
20000
pmin
30000
0
10000
20000
Mpubli
30000
40000
11
6. Estimation and results
We estimate the following function:
log( price )  e   1 log(min utesit )   2 log( share it )   3 contractingout it 
  4 public it   5 per14  29jt   6 per 30  44 jt   7 women jt  8 log( regional _ gdp jt )  
Dependent variable:
– Pminit : Ad price per minute for broadcaster i at moment t
Variables de control:
– minutes it: minutes of advertising in network i at moment t.
– share it: the percentage of viewers who are watching TV in network i at moment t .
– contracting out : management of network i sales of advertising time .
– public : ownership of network i
– per14-29 : percentage of population 14 -29 years old in territory j at moment t
– per30-44 : percentage of population 30 -44 years old in territory j at moment t
– women : percentage of women in territory j at moment t
– regional_gdp: regional gross domestic product per capita in territory j at moment t
12
6. Estimation and results
Price equation (dependent variable: lpmin)
Lminutes
Lshare
Contracting out
Ownership
lPer14-29
lPer30-44
Women
lregional_gdp
Tendency
Intercept
N
R2
Test F (Joint significance)
Test Sargan (Over identification test of all
instruments)
Test Anderson canonical correlations
(under identification and weak identification)
Lshare
Shea’s partial R2 (excluded instruments)
Test
F
(Significance
of
excluded
instruments)
Lmin
Shea’s partial R2 (excluded instruments)
Test
F
(Significance
of
excluded
instruments)
OLS
-0.264
(0.272)
0.848
(0.069)***
0.171
(0.017)**
-0.478
(0.154)***
5.814
(0.914)***
-9.282
(1.263)***
2.856
(1.999)
2.903
(0.299)***
0.071
(0.171)***
-19.173
(9.024)**
122
0.90
355.43***
--
(IV/2SLS)
-0.444
(0.231)**
0.947
(0.096)***
0.085
(0.156)
-0.380
(0.165)**
4.515
(2.181)**
-7.316
(2.643)***
3.030
(1.936)
2.555
(0.644)***
0.060
(0.032)*
-17.034
(9.864)*
108
0.89
86.64***
5.586
--
32.852***
0.547
(16.29)***
--
0.311
(6.76)***
13
Note 1: Robust standard errors in parenthesis.
Note 2: *** significant at 1% level,; ** significant at 5% level; * significant at 10% level.
Note 3: The instruments for minutes and share are log (entertainment), football, language, log(minutesT-1) and log(cultural).
7. Concluding remarks
Our results show:
– The share is a key variable in advertising price. (High elasticity, 0.947, of ad
price with respect to share).
– Negative relationship between ad time and ad price. Advertisers pay more
for ads on networks that have less total ad time.
– We obtain a negative and significant relationship between public ownership
and ad price.
– The variables contracting out and women are not significant relationship.
– We find a positive relationship between price and the percentage of
population 14 and 29 years old and the strength of GDP in the region .
14
Thank you
[email protected]
15