SlideShare a Scribd company logo
1 of 12
Turn left or right:
Are political preferences in the Netherlands changed
               by media developments?

      ARIMA models and impact assessment analysis




                           Assignment 2




            Mark Boukes (markboukes@Hotmail.com)
                             5616298




                      1st semester 2010/2011
                      Dynamic Data Analysis
                    Lecturer: Dr. R. Vliegenthart
                         November 23, 2010




                                 Communication Science (Research MSc)
                               Faculty of Social and Behavioural Sciences
                                                 University of Amsterdam
Table of contents


INTRODUCTION.............................................................................................................................................1
METHOD........................................................................................................................................................2
RESULTS........................................................................................................................................................3
    ARIMA MODEL.......................................................................................................................................................3
    IMPACT ASSESSMENT ANALYSIS........................................................................................................................................5
CONCLUSION.................................................................................................................................................6
REFERENCES..................................................................................................................................................7
APPENDIX 1: DO FILE......................................................................................................................................I
Introduction
The public broadcasting system in the Netherlands is part of a recent political discussion:
should the state invest the same amount of money in it or can the budget for Nederland 1, 2
and 3 be reduced to limit the budget deficit of the state? Advocates of public broadcasting
point to the (democratic) benefits of this system: it provides citizens with news from different
stands, involves people with politics and culture, and gives a voice to minority groups in
society without having much commercial pressure. According to former chairman of the
BBC, Gavyn Davies (1999), should public broadcasters “… inform, educate and entertain in a
way which the private sector, left unregulated, would not do.” (p. 10) However, opponents of
public broadcasting think that financing television channels by the state is false competition to
players on the market that gather revenues by advertising only. Those opponents claim also
that public broadcasting is a waste of tax money, because they believe that commercial parties
can fulfil the tasks of public broadcasting as long as there is no false competition. General
manager of RTL 4 (commercial broadcaster in the Netherlands) Bert Habets thinks that a first
step in decreasing the budget of public broadcasting is cooperation with commercial channels
on areas as sports and entertainment.1 Shortly, the position of the Dutch public broadcasting
system is under pressure and therefore I investigated its impact.
        The purpose of this study is to research how the introduction of commercial television
in the Netherlands affected political attitudes of citizens. Many studies before investigated the
effects of media use on political attitudes, like political interest, political knowledge, vote
intention, government approval etc. Curran, Iyengar, Brink Lund and Salovaara-Moring
(2009) found that, compared to people in countries with a market model, people had more
knowledge about both international and domestic news topics like the Kyoto Protocol and
news about the French president in countries with a public broadcasting system. Aarts and
Semetko (2003) found a similar effect in the Netherlands. Watching more public broadcasting
instead of commercial broadcasters had positive effects on voter turnout, but also on political
cognition and efficacy. DellaVigna and Kaplan (2007) also found evidence for the importance
of introducing new channels to citizens. They showed that voters in towns in the United
States, in which Fox News was introduced in 2000, were more likely to vote Republican in
the first coming election. I aimed with my study to see if the introduction of commercial
channels to the Dutch range of television channels had a similar effect. The research question
was therefore: Did the introduction of commercial broadcasting to Dutch television affect the
average political preference on a left-right dimension?
1
 http://www.rtl.nl/%28/actueel/rtlboulevard/nieuws/articleview/%29/components/actueel/rtlboulevard/2010/11_
november/entertainment/rtl_baas_handreiking_po.xml

                                                                                                          1
To answer this question, I used data from 1977 until 2000, which were analyzed with a
ARIMA model and an impact assessment analysis. As no one conducted such a research before,
I hope to find new information that gives more insight in the importance of public
broadcasting in the Netherlands and enriches the debate about the existence and financing of
public broadcasting.

Method
To investigate which factors have an effect on the left-right preferences of Dutch citizens, I used
a dataset that contains information about this for a long time period. The NIPO Weeksurveys
1962-2000 was found on https://easy.dans.knaw.nl/dms and contained for the period
1977-2000, 1.086.336 individual answers for the following question, Here you see seven boxes
between the words left and right. Could you indicate on this scale how left, right or in between
your political opinion lies? The observations were transformed in such a way that the mean
answer for every week was reported, because the answers were reported individually and
aggregate level data is needed to answer the research question,. This resulted in 1226 weekly
items containing the value for the average left-right preference of Dutch citizens.
       I included also data for various happenings that I have in mind to effect political
preferences of voters, to investigate whether they indeed had an effect. Various dummy
variables were created, to study the impact of the introduction of commercial television on the
left-right preferences of citizens. One for the introduction of RTL4 to the Dutch television
viewers on October 2 1989 (dummy variables were 0 before this date and 1 after this date), and
the other for the moment (starting in 1992) it was allowed for commercial television channels to
broadcast in the Netherlands. Also dummy variables were created for the introduction of the
other Dutch commercial channels that attract(ed) many viewers: Veronica, RTL5 and SBS6. All
those effects were expected to have a permanent effect, which came about rather abrupt. In
addition, a variable was created adding the values of the entrance of the different channels
together; for instance, after RTL4 was introduced this variable was one, after Veronica was
introduced this value was two, etc. The effect of this variable can thus be said to be gradual and
permanent. The following events were also taken into account by making dummy variables for
the periods that represent them, to control for other developments in Dutch and the global
society. Moments that were taken into account are the hijack of a train by Moluccans from May
23 1977 untill June 11 1977; the second oil crisis in 1979/1980, the weeks before, during and
after European elections (in 1979, 1984, 1989, 1994, 1999); the fall of the Iron Curtain starting
in April 1989 in Poland and lasting until the end of that year when also the Berlin Wall fell;
Dutch politician Bolkestein who held a speech about the integration of ethnic minorities in
Dutch society on September 6 1991 and finally the happenings in Srebrenica in July 1995. To

                                                                                                    2
analyse the effects of those events, first an adequate ARIMA model is developed, which is
followed by an impact assessment analysis of the different events.

Results
I specify in this results section how the ARIMA model was created and thereafter what the
results were of the impact assessment analysis to include events in the analysis.

ARIMA model
The timeseries dataset is first being analysed according to the ARIMA-framework described by
Vliegenthart (n.d.). The first step was to establish whether the data was stationary. Therefore, I
made a graph was made (see Figure 1) to inspect the data in a convenient manner and three
augmented Dickey Fuller (ADF) tests were used to check whether the series was a random walk
without drift, a random walk with drift or a random walk with drift and trend (see Table 1). One
of the results of this ADF indicates that the null hypotheses of non-stationarity should not be
rejected and therefore it was necessary to integrate the data and check again for stationarity.
This time only significant results were found; the integrated series was stationary (see Figure 2).




Figure 1. The mean score on a political left-right dimension over time.

Table 1. The various augmented Dickey-Fuller tests.
                                                  Augmented Dickey-Fuller test
Random walk without drift                         -0.374*
Random walk with drift                            -12.779
Random walk with drift and trend                  -21.805

After integrating
Random walk without drift                         -57.741
Random walk with drift                            -57.718
Random walk with drift and trend                  -57.696
Note. All tests without a * indicate the presence of an unit root.




                                                                                                     3
Figure 2. The differenced value for the mean score on a political left-right dimension over time.

The next step was predicting the data as good as possible by accounting for its past, either with
autoregressive (AR) terms, moving average (MA) terms or both. This was done by inspecting
the autocorrelation (ACF) and partial autocorrelation functions (PACF) (see the graphs for both in
Figure 2). Those functions suggest a moving average at lag 1; there was one peak in the ACF
and a decaying order in the PACF. Therefore, a ARIMA(0,1,1) model was used which indeed
has a significant effect for this moving average term. This model was tested for autocorrelation
and heteroscedasticity with the Ljung–Box Q test statistic for both the residuals as well as the
squared residuals. The (significant) results for both tests indicate that there was
autocorrelation present and thus not only white noise. Therefore the model was extended with
a moving average term at lag 2 for tho reasons. First, because the ACF and PACF showed a
peak at the second lag, however, the graphs did not give a clear indication for adding an
autoregressive (AR) or moving average (MA) term. Second, adding a MA term increased the
fit more than adding a AR term according to the values of the Akaike Information Criterion
(AIC) and Schwarz Bayesian Criterion (BIC), therefore I choose to use a resulting ARIMA
(0,1,2) model.
        The values of this resulting model can be found in Table 2. The results of the Ljung–
Box Q test statistic for this model was insignificant for the residuals, therefore we can assume
the residuals to be white noise. However significant results were found for the squared
residuals with the Ljung–Box Q test statistic, which indicates the presence of
heteroscedasticity. However for this moment no attention is paid to this, but hopefully it will be
solved with ARCH and GARCH models later in this Dynamic Data Analysis course.




                                                                                                    4
Figure 3. ACF and PCF for the differenced mean score on a political left-right dimension.



Table 2. ARIMA model for mean score on a political left-right dimension over time.
                                         ARIMA (0,1,2)
Constant                                 -.000 (.000)
MA(1)                                    -.767 (.026)*
MA(2)                                    -.091 ( .026)*

Ljung-Box Q(20) residuals                20.61
Ljung-Box Q(20) residuals²               103.34*

AIC                                      -3583.93
BIC                                      -3563.49
Note. Unstandardized coefficients. Standard errors in parentheses; * p<.001

Impact assessment analysis
I created various dummy variables were created to test if different events had an effect on the
mean score for the political preference on a left-right dimension over time. The duration of
those events could be temporary or permanent and the onset could be gradual or abrupt
(McCleary & Hay, 1980). How the duration and onset were implemented was based on
simple theory, common sense. The various events were incorporated in the ARIMA-model
that was established before. Because the timeseries used as a dependent variable was
differenced, also differenced values of the independent (event) variables were used in the
analysis.
        Three models were the most interesting to study the impact of television developments
on political preferences; the one that makes clear what happened after RTL4’s introduction,
the one for the moment commercial broadcasters had permission to broadcast in the
Netherlands, and the one that can be used for the gradual introduction of more commercial


                                                                                               5
broadcasters in the Netherlands. However, as Table 3 shows, none of those developments had
an effect on the mean score for the political left-right dimension.

Table 3. Influence on media developments on the mean score on a political left-right dimension over time.
                                                                                  Permission     All TVchannels
                                            ARIMA (0,1,2) RTL4-model
                                                                                     model          intro-model
Constant                                    -.000 (.000)         -.000 (.000)    -.000 (.000)    -.000 (.000)
MA(1)                                       -.767 (.026)*        -.767 (.026)*   -.766 (.026)*   -.767 (.026)*
MA(2)                                       -.091 (.026)*        -.089 (.026)*   -.092 (.026)*   -.090 (.026)*
Introduction RTL4                                                .015 (.048)
Permission for commercial broadcasting                                           .020 (.026)
Introduction different TV channels                                                               0.007 (0.017)

Ljung-Box Q(20) residuals                   20.61                20.59           20.98           20.55
Ljung-Box Q(20) residuals²                  103.34*              103.54*         103.71*         103.79*

AIC                                         -3583.93             -3582.18        -3582.36        -3582.22
BIC                                         -3563.49             -3556.64        -3556.81        -3556.68
Note. Unstandardized coefficients. Standard errors in parentheses; * p<.001
To check whether this dependent timeseries variable is not influenced by any factor, the
analysis were repeated for various historical events that were specified in the methods section
that might have affected the political preferences of people. Also for those events the
differenced series were used in the analysis. The results showed that only the second oil crisis
in 1979/1980 had an effect on political preferences. As this effect was positive, it can be said
that this crisis caused an increase in the preference for right-wing politics. Perhaps more
remarkable is that all other variables had no effect on the average political preference in the
Netherlands, both when tested alone in the ARIMA(0,1,2)-model or when tested together as
shown in Table 4.

Conclusion
My main aim in this paper was to study the impact of the introduction of commercial
broadcasting to Dutch viewers. It seems that this had no effect on the political preferences of
Dutch people; the mean score on a left-right scale was not affected by the various events. To
do this analysis first a proper ARIMA-model was build by integrating it one time and taking
the first and the second lag into account with moving average terms. The resulting model did
reflect white noise in the residuals, however the squared residuals did still correlate, which
indicates heteroscedasticity. Thereafter, events like the introduction of RTL4, the moment of
permission for commercial broadcasting and the introduction of different commercial
channels were included in the model. All three however seemed to have no effect. Therefore,
other (historical) events were also taken into account to see if this dataseries was not affected


                                                                                                       6
at all by events. Only the second oil crisis seemed to affect peoples’ political preferences.
Most events however are unlikely to influence the average political preference on a left-right
dimension; the media developments did not achieve it and also most historical events did not
have an effect.

Table 4. Influence on historical events on the mean score on a political left-right dimension over time.
                                                    Historical events model
Constant                                            -.000 (.000)
MA(1)                                               -.767 (.026)**
MA(2)                                               -.091 (.026)**
Introduction different TV channels                  .009 (.016)
Train hijack                                        -.025 (.054)
Second oil crisis                                   .036 (0.015)*
European elections                                  .014 (.017)
Fall of Iron Curtain                                .022 (.027)
Bolkestein’s speech                                 .020 (.298)
Srebrenica                                          .013 (.036)

Ljung-Box Q(20) residuals                           21.43
Ljung-Box Q(20) residuals²                          101.81**

AIC                                                 -3575.97
BIC                                                 -3519.77
Note. Unstandardized coefficients. Standard errors in parentheses; *p<0.05, ** p<.001


References
Aarts, K., & Semetko, H. A. (2003). The divided electorate: Media use and political
        involvement. The Journal of Politics, 65(3), 759-784.

Curran, J., Iyengar, S., Brink Lund, A., & Salovaara-Moring, I. (2009). Media system, public
        knowledge and democracy. European Journal of Communication, 24(1), 5-26.

Davies, G. (1999). The Future Funding of the BBC. Report of the Independent Review Panel
        to Department of Culture, Media and Sport. Retrieved on November 23, 2010, from
        http://news.bbc.co.uk/hi/english/static/bbc_funding_review/reviewco.pdf

DellaVigna, S. & Kaplan, E. (2007). The Fox News effect: Media bias and voting. Quarterly
       Journal of Economics, 122(3), 1187-1234.

McCleary, R., & Hay, R. (1980). Applied Time Series Analysis for the Social Sciences.
       London: Sage.

Vliegenthart, R. (n.d.). Moving up. Applying aggregate level time series analysis in
       communication science. Unpublished manuscript.

                                                                                                      7
Appendix 1: Do file

* declare data to be time series
replace nr = nr + 766 + 52 + 52
tsset nr, weekly

* making differences
gen dleftright=leftright - l.leftright

* graphic
twoway (tsline leftright, lcolor(black))


* 3 tests for stationarity, 1)without constant=random walk without drift,
2)normal ADF: random walk with drift, 3) ADF trend: random walk with drift
and trend
dfuller leftright, noconstant
dfuller leftright
dfuller leftright, trend

twoway (tsline d.leftright, lcolor(black))

dfuller d.leftright, noconstant
dfuller d.leftright
dfuller d.leftright, trend

* arima model
ac leftright
pac leftright

ac d.leftright
pac d.leftright

arima d.leftright, ma(1,2)
estat ic
predict r, res
wntestq r, lags(20)
gen r_s=r*r
wntestq r_s, lags(20)
ac r
pac r
ac r_s
pac r_s
drop r r_s

* create intervention for RTL. eerste uitzending was op 2 oktober 1989 =
week 40
gen RTL=0
replace RTL = 1 if nr>1520

*Veronica
gen veronica=0
replace veronica = 1 if nr>1825

*rtl5 2 oktober 1993
gen rtl5=0
replace rtl5 = 1 if nr>1725

*sbs6 28 augustus 1995
gen sbs6=0

                                                                             i
replace sbs6 = 1 if nr>1823

*Media development
gen media = RTL+ veronica+ rtl5+ sbs6

*in 1992 werden binnenlandse commerciële omroepen toegestaan
gen comm=0
replace comm = 1 if nr>1635

*EU elections 1979,1984, 1989,1994,1999
gen euelect=0
replace euelect = 1 if nr>=991 & nr<=993
replace euelect = 1 if nr>=1245 & nr<=1247
replace euelect = 1 if nr>=1502 & nr<=1504
replace euelect = 1 if nr>=1759 & nr<=1761
replace euelect = 1 if nr>=2016 & nr<=2018

*Fall of iron curtain starting in Poland, then berlin wall, untill and of
1989, minus christmas weeks
gen IronCurtain=0
replace IronCurtain = 1 if nr>=1493 & nr<=1529

*Bolkestein 6 september 1991
gen bolkestein=0
replace bolkestein= 1 if nr==1618

*Molukse acties: Treinkaping bij De Punt 23 mei (21)-11 juni 1977(23)
gen moluk=0
replace moluk= 1 if nr>=904 & nr<=906

*2e oliecrisis: March 26 1979 (13): OPEC makes full 14.5 percent price
increase for 1979 effective on April 1
*December 1980 (48): Collapse of OPEC's pricing structure. Saudis use $32
per barrel marker, others use $36 per barrel benchmark.
gen oil=0
replace oil= 1 if nr>=983 & nr<=1067

*Srebrenica 11 juli 1995 (28)
gen srebrenica=0
replace srebrenica= 1 if nr>=1818 & nr<=1828

arima d.leftright d.srebrenica, ma(1,2)
estat ic
predict r, res
wntestq r, lags(20)
gen r_s = r*r
wntestq r_s, lags(20)
drop r r_s

arima d.leftright d.media d.euelect d.IronCurtain d.bolkestein d.moluk
d.oil d.srebrenica, ma(1,2)
estat ic
predict r, res
wntestq r, lags(20)
gen r_s = r*r
wntestq r_s, lags(20)
drop r r_s




                                                                            ii

More Related Content

Similar to Turn left or right: Are political preferences in the Netherlands changed by media developments? - ARIMA models and impact assessment analysis

Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
IbrarHussain105
 
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docxAnn Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
durantheseldine
 
Political (In)Stability and Public Policy Transplantation: a Macedonian Case
Political (In)Stability and Public Policy Transplantation: a Macedonian CasePolitical (In)Stability and Public Policy Transplantation: a Macedonian Case
Political (In)Stability and Public Policy Transplantation: a Macedonian Case
jpsjournal1
 
Demonstration Exercise 1109one ofE, ulicylone III.docx
Demonstration Exercise 1109one ofE, ulicylone III.docxDemonstration Exercise 1109one ofE, ulicylone III.docx
Demonstration Exercise 1109one ofE, ulicylone III.docx
theodorelove43763
 

Similar to Turn left or right: Are political preferences in the Netherlands changed by media developments? - ARIMA models and impact assessment analysis (20)

Turn left or right: How the economy affects political preferences and media c...
Turn left or right: How the economy affects political preferences and media c...Turn left or right: How the economy affects political preferences and media c...
Turn left or right: How the economy affects political preferences and media c...
 
Media, intention and final vote: A two-wave panel data study to the effects o...
Media, intention and final vote: A two-wave panel data study to the effects o...Media, intention and final vote: A two-wave panel data study to the effects o...
Media, intention and final vote: A two-wave panel data study to the effects o...
 
Who Sets the Agenda: Media or Parliament?: A panel data study to the agenda s...
Who Sets the Agenda: Media or Parliament?: A panel data study to the agenda s...Who Sets the Agenda: Media or Parliament?: A panel data study to the agenda s...
Who Sets the Agenda: Media or Parliament?: A panel data study to the agenda s...
 
Explaining political coverage in the Netherlands; Why Dutch voters are less c...
Explaining political coverage in the Netherlands; Why Dutch voters are less c...Explaining political coverage in the Netherlands; Why Dutch voters are less c...
Explaining political coverage in the Netherlands; Why Dutch voters are less c...
 
Curso lse
Curso lseCurso lse
Curso lse
 
Adrian Kay - The Dynamics of Public Policy
Adrian Kay - The Dynamics of Public PolicyAdrian Kay - The Dynamics of Public Policy
Adrian Kay - The Dynamics of Public Policy
 
Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
Intermedia Agenda Setting in the Social Media Age: How Traditional Players Do...
 
Sphere layout final-051415
Sphere layout final-051415Sphere layout final-051415
Sphere layout final-051415
 
9783631755938.pdf
9783631755938.pdf9783631755938.pdf
9783631755938.pdf
 
How the Serialization of News Affects People’s Attitudes Toward Politicians I...
How the Serialization of News Affects People’s Attitudes Toward Politicians I...How the Serialization of News Affects People’s Attitudes Toward Politicians I...
How the Serialization of News Affects People’s Attitudes Toward Politicians I...
 
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docxAnn Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
Ann Oper Res (2016) 2361–13DOI 10.1007s10479-015-1902-9.docx
 
Brand equity ebsco
Brand equity ebscoBrand equity ebsco
Brand equity ebsco
 
G0352037043
G0352037043G0352037043
G0352037043
 
Data4Impact Expert Workshop Report
Data4Impact Expert Workshop ReportData4Impact Expert Workshop Report
Data4Impact Expert Workshop Report
 
When East Meets West: Interpersonal Contact and the Demand for Democracy
When East Meets West: Interpersonal Contact and the Demand for DemocracyWhen East Meets West: Interpersonal Contact and the Demand for Democracy
When East Meets West: Interpersonal Contact and the Demand for Democracy
 
Book-public-policy-analysis.pdf
Book-public-policy-analysis.pdfBook-public-policy-analysis.pdf
Book-public-policy-analysis.pdf
 
OTT Services - Colour to the internet
OTT Services - Colour to the internetOTT Services - Colour to the internet
OTT Services - Colour to the internet
 
Political (In)Stability and Public Policy Transplantation: a Macedonian Case
Political (In)Stability and Public Policy Transplantation: a Macedonian CasePolitical (In)Stability and Public Policy Transplantation: a Macedonian Case
Political (In)Stability and Public Policy Transplantation: a Macedonian Case
 
Demonstration Exercise 1109one ofE, ulicylone III.docx
Demonstration Exercise 1109one ofE, ulicylone III.docxDemonstration Exercise 1109one ofE, ulicylone III.docx
Demonstration Exercise 1109one ofE, ulicylone III.docx
 
IRJET- Sentiment Analysis using Machine Learning
IRJET- Sentiment Analysis using Machine LearningIRJET- Sentiment Analysis using Machine Learning
IRJET- Sentiment Analysis using Machine Learning
 

More from Mark Boukes (University of Amsterdam)

More from Mark Boukes (University of Amsterdam) (20)

Etmaal 2014 presentation: on the effects of opinionated news
Etmaal 2014 presentation: on the effects of opinionated newsEtmaal 2014 presentation: on the effects of opinionated news
Etmaal 2014 presentation: on the effects of opinionated news
 
Appendix I - Transcripts of audio in the videos
Appendix I - Transcripts of audio in the videosAppendix I - Transcripts of audio in the videos
Appendix I - Transcripts of audio in the videos
 
Politicians in Talk Shows - Etmaal van de Communicatiewetenschap 2012
Politicians in Talk Shows - Etmaal van de Communicatiewetenschap 2012Politicians in Talk Shows - Etmaal van de Communicatiewetenschap 2012
Politicians in Talk Shows - Etmaal van de Communicatiewetenschap 2012
 
Cartoon controversy; why the Danish Mohammed cartoons could be published.
Cartoon controversy; why the Danish Mohammed cartoons could be published. Cartoon controversy; why the Danish Mohammed cartoons could be published.
Cartoon controversy; why the Danish Mohammed cartoons could be published.
 
Media attention in Belgium: How much influence do citizens and politicians ha...
Media attention in Belgium: How much influence do citizens and politicians ha...Media attention in Belgium: How much influence do citizens and politicians ha...
Media attention in Belgium: How much influence do citizens and politicians ha...
 
Sustainable Development in Popular Newspapers: How is coverage in De Telegraa...
Sustainable Development in Popular Newspapers: How is coverage in De Telegraa...Sustainable Development in Popular Newspapers: How is coverage in De Telegraa...
Sustainable Development in Popular Newspapers: How is coverage in De Telegraa...
 
Asymmetric media responses in the Dutch context: Does newspapers coverage res...
Asymmetric media responses in the Dutch context: Does newspapers coverage res...Asymmetric media responses in the Dutch context: Does newspapers coverage res...
Asymmetric media responses in the Dutch context: Does newspapers coverage res...
 
Sustainable development in three newspapers: How does coverage in a particula...
Sustainable development in three newspapers: How does coverage in a particula...Sustainable development in three newspapers: How does coverage in a particula...
Sustainable development in three newspapers: How does coverage in a particula...
 
Attention for de Publieke omroep in newspapers: public broadcasting in the n...
Attention for de Publieke omroep in newspapers:  public broadcasting in the n...Attention for de Publieke omroep in newspapers:  public broadcasting in the n...
Attention for de Publieke omroep in newspapers: public broadcasting in the n...
 
Public broadcasting; what should it add, what should be its role and what are...
Public broadcasting; what should it add, what should be its role and what are...Public broadcasting; what should it add, what should be its role and what are...
Public broadcasting; what should it add, what should be its role and what are...
 
Journalism and the media: the cartoon controversy: Why were they published?
Journalism and the media: the cartoon controversy: Why were they published?Journalism and the media: the cartoon controversy: Why were they published?
Journalism and the media: the cartoon controversy: Why were they published?
 
Climate change; explaining the differences in reporting
Climate change; explaining the differences in reportingClimate change; explaining the differences in reporting
Climate change; explaining the differences in reporting
 
Journalisten, politici en agendasetting
Journalisten, politici en agendasettingJournalisten, politici en agendasetting
Journalisten, politici en agendasetting
 
Reducing prejudice via mediated contact with immigrants - Proposal for an exp...
Reducing prejudice via mediated contact with immigrants - Proposal for an exp...Reducing prejudice via mediated contact with immigrants - Proposal for an exp...
Reducing prejudice via mediated contact with immigrants - Proposal for an exp...
 
Thematische congruentie en de invloed van advertenties
Thematische congruentie en de invloed van advertentiesThematische congruentie en de invloed van advertenties
Thematische congruentie en de invloed van advertenties
 
Structural equation modelling
Structural equation modellingStructural equation modelling
Structural equation modelling
 
Influence of media source on political interest;
Influence of media source on political interest;Influence of media source on political interest;
Influence of media source on political interest;
 
Interviewing ethnic minorities
Interviewing ethnic minoritiesInterviewing ethnic minorities
Interviewing ethnic minorities
 
Causality
CausalityCausality
Causality
 
Onbewust asocialer
Onbewust asocialerOnbewust asocialer
Onbewust asocialer
 

Recently uploaded

Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost LoverPowerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
PsychicRuben LoveSpells
 
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
Faga1939
 

Recently uploaded (20)

Transformative Leadership: N Chandrababu Naidu and TDP's Vision for Innovatio...
Transformative Leadership: N Chandrababu Naidu and TDP's Vision for Innovatio...Transformative Leadership: N Chandrababu Naidu and TDP's Vision for Innovatio...
Transformative Leadership: N Chandrababu Naidu and TDP's Vision for Innovatio...
 
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 46 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 46 (Gurgaon)Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 46 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 46 (Gurgaon)
 
declarationleaders_sd_re_greens_theleft_5.pdf
declarationleaders_sd_re_greens_theleft_5.pdfdeclarationleaders_sd_re_greens_theleft_5.pdf
declarationleaders_sd_re_greens_theleft_5.pdf
 
Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost LoverPowerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
Powerful Love Spells in Phoenix, AZ (310) 882-6330 Bring Back Lost Lover
 
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 47 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 47 (Gurgaon)Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 47 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 47 (Gurgaon)
 
Busty Desi⚡Call Girls in Vasundhara Ghaziabad >༒8448380779 Escort Service
Busty Desi⚡Call Girls in Vasundhara Ghaziabad >༒8448380779 Escort ServiceBusty Desi⚡Call Girls in Vasundhara Ghaziabad >༒8448380779 Escort Service
Busty Desi⚡Call Girls in Vasundhara Ghaziabad >༒8448380779 Escort Service
 
Group_5_US-China Trade War to understand the trade
Group_5_US-China Trade War to understand the tradeGroup_5_US-China Trade War to understand the trade
Group_5_US-China Trade War to understand the trade
 
Politician uddhav thackeray biography- Full Details
Politician uddhav thackeray biography- Full DetailsPolitician uddhav thackeray biography- Full Details
Politician uddhav thackeray biography- Full Details
 
China's soft power in 21st century .pptx
China's soft power in 21st century   .pptxChina's soft power in 21st century   .pptx
China's soft power in 21st century .pptx
 
05052024_First India Newspaper Jaipur.pdf
05052024_First India Newspaper Jaipur.pdf05052024_First India Newspaper Jaipur.pdf
05052024_First India Newspaper Jaipur.pdf
 
Embed-2 (1).pdfb[k[k[[k[kkkpkdpokkdpkopko
Embed-2 (1).pdfb[k[k[[k[kkkpkdpokkdpkopkoEmbed-2 (1).pdfb[k[k[[k[kkkpkdpokkdpkopko
Embed-2 (1).pdfb[k[k[[k[kkkpkdpokkdpkopko
 
04052024_First India Newspaper Jaipur.pdf
04052024_First India Newspaper Jaipur.pdf04052024_First India Newspaper Jaipur.pdf
04052024_First India Newspaper Jaipur.pdf
 
422524114-Patriarchy-Kamla-Bhasin gg.pdf
422524114-Patriarchy-Kamla-Bhasin gg.pdf422524114-Patriarchy-Kamla-Bhasin gg.pdf
422524114-Patriarchy-Kamla-Bhasin gg.pdf
 
KING VISHNU BHAGWANON KA BHAGWAN PARAMATMONKA PARATOMIC PARAMANU KASARVAMANVA...
KING VISHNU BHAGWANON KA BHAGWAN PARAMATMONKA PARATOMIC PARAMANU KASARVAMANVA...KING VISHNU BHAGWANON KA BHAGWAN PARAMATMONKA PARATOMIC PARAMANU KASARVAMANVA...
KING VISHNU BHAGWANON KA BHAGWAN PARAMATMONKA PARATOMIC PARAMANU KASARVAMANVA...
 
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
THE OBSTACLES THAT IMPEDE THE DEVELOPMENT OF BRAZIL IN THE CONTEMPORARY ERA A...
 
Enjoy Night ≽ 8448380779 ≼ Call Girls In Palam Vihar (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Palam Vihar (Gurgaon)Enjoy Night ≽ 8448380779 ≼ Call Girls In Palam Vihar (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Palam Vihar (Gurgaon)
 
Busty Desi⚡Call Girls in Sector 62 Noida Escorts >༒8448380779 Escort Service
Busty Desi⚡Call Girls in Sector 62 Noida Escorts >༒8448380779 Escort ServiceBusty Desi⚡Call Girls in Sector 62 Noida Escorts >༒8448380779 Escort Service
Busty Desi⚡Call Girls in Sector 62 Noida Escorts >༒8448380779 Escort Service
 
Embed-4.pdf lkdiinlajeklhndklheduhuekjdh
Embed-4.pdf lkdiinlajeklhndklheduhuekjdhEmbed-4.pdf lkdiinlajeklhndklheduhuekjdh
Embed-4.pdf lkdiinlajeklhndklheduhuekjdh
 
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 48 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 48 (Gurgaon)Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 48 (Gurgaon)
Enjoy Night ≽ 8448380779 ≼ Call Girls In Gurgaon Sector 48 (Gurgaon)
 
WhatsApp 📞 8448380779 ✅Call Girls In Chaura Sector 22 ( Noida)
WhatsApp 📞 8448380779 ✅Call Girls In Chaura Sector 22 ( Noida)WhatsApp 📞 8448380779 ✅Call Girls In Chaura Sector 22 ( Noida)
WhatsApp 📞 8448380779 ✅Call Girls In Chaura Sector 22 ( Noida)
 

Turn left or right: Are political preferences in the Netherlands changed by media developments? - ARIMA models and impact assessment analysis

  • 1. Turn left or right: Are political preferences in the Netherlands changed by media developments? ARIMA models and impact assessment analysis Assignment 2 Mark Boukes (markboukes@Hotmail.com) 5616298 1st semester 2010/2011 Dynamic Data Analysis Lecturer: Dr. R. Vliegenthart November 23, 2010 Communication Science (Research MSc) Faculty of Social and Behavioural Sciences University of Amsterdam
  • 2. Table of contents INTRODUCTION.............................................................................................................................................1 METHOD........................................................................................................................................................2 RESULTS........................................................................................................................................................3 ARIMA MODEL.......................................................................................................................................................3 IMPACT ASSESSMENT ANALYSIS........................................................................................................................................5 CONCLUSION.................................................................................................................................................6 REFERENCES..................................................................................................................................................7 APPENDIX 1: DO FILE......................................................................................................................................I
  • 3. Introduction The public broadcasting system in the Netherlands is part of a recent political discussion: should the state invest the same amount of money in it or can the budget for Nederland 1, 2 and 3 be reduced to limit the budget deficit of the state? Advocates of public broadcasting point to the (democratic) benefits of this system: it provides citizens with news from different stands, involves people with politics and culture, and gives a voice to minority groups in society without having much commercial pressure. According to former chairman of the BBC, Gavyn Davies (1999), should public broadcasters “… inform, educate and entertain in a way which the private sector, left unregulated, would not do.” (p. 10) However, opponents of public broadcasting think that financing television channels by the state is false competition to players on the market that gather revenues by advertising only. Those opponents claim also that public broadcasting is a waste of tax money, because they believe that commercial parties can fulfil the tasks of public broadcasting as long as there is no false competition. General manager of RTL 4 (commercial broadcaster in the Netherlands) Bert Habets thinks that a first step in decreasing the budget of public broadcasting is cooperation with commercial channels on areas as sports and entertainment.1 Shortly, the position of the Dutch public broadcasting system is under pressure and therefore I investigated its impact. The purpose of this study is to research how the introduction of commercial television in the Netherlands affected political attitudes of citizens. Many studies before investigated the effects of media use on political attitudes, like political interest, political knowledge, vote intention, government approval etc. Curran, Iyengar, Brink Lund and Salovaara-Moring (2009) found that, compared to people in countries with a market model, people had more knowledge about both international and domestic news topics like the Kyoto Protocol and news about the French president in countries with a public broadcasting system. Aarts and Semetko (2003) found a similar effect in the Netherlands. Watching more public broadcasting instead of commercial broadcasters had positive effects on voter turnout, but also on political cognition and efficacy. DellaVigna and Kaplan (2007) also found evidence for the importance of introducing new channels to citizens. They showed that voters in towns in the United States, in which Fox News was introduced in 2000, were more likely to vote Republican in the first coming election. I aimed with my study to see if the introduction of commercial channels to the Dutch range of television channels had a similar effect. The research question was therefore: Did the introduction of commercial broadcasting to Dutch television affect the average political preference on a left-right dimension? 1 http://www.rtl.nl/%28/actueel/rtlboulevard/nieuws/articleview/%29/components/actueel/rtlboulevard/2010/11_ november/entertainment/rtl_baas_handreiking_po.xml 1
  • 4. To answer this question, I used data from 1977 until 2000, which were analyzed with a ARIMA model and an impact assessment analysis. As no one conducted such a research before, I hope to find new information that gives more insight in the importance of public broadcasting in the Netherlands and enriches the debate about the existence and financing of public broadcasting. Method To investigate which factors have an effect on the left-right preferences of Dutch citizens, I used a dataset that contains information about this for a long time period. The NIPO Weeksurveys 1962-2000 was found on https://easy.dans.knaw.nl/dms and contained for the period 1977-2000, 1.086.336 individual answers for the following question, Here you see seven boxes between the words left and right. Could you indicate on this scale how left, right or in between your political opinion lies? The observations were transformed in such a way that the mean answer for every week was reported, because the answers were reported individually and aggregate level data is needed to answer the research question,. This resulted in 1226 weekly items containing the value for the average left-right preference of Dutch citizens. I included also data for various happenings that I have in mind to effect political preferences of voters, to investigate whether they indeed had an effect. Various dummy variables were created, to study the impact of the introduction of commercial television on the left-right preferences of citizens. One for the introduction of RTL4 to the Dutch television viewers on October 2 1989 (dummy variables were 0 before this date and 1 after this date), and the other for the moment (starting in 1992) it was allowed for commercial television channels to broadcast in the Netherlands. Also dummy variables were created for the introduction of the other Dutch commercial channels that attract(ed) many viewers: Veronica, RTL5 and SBS6. All those effects were expected to have a permanent effect, which came about rather abrupt. In addition, a variable was created adding the values of the entrance of the different channels together; for instance, after RTL4 was introduced this variable was one, after Veronica was introduced this value was two, etc. The effect of this variable can thus be said to be gradual and permanent. The following events were also taken into account by making dummy variables for the periods that represent them, to control for other developments in Dutch and the global society. Moments that were taken into account are the hijack of a train by Moluccans from May 23 1977 untill June 11 1977; the second oil crisis in 1979/1980, the weeks before, during and after European elections (in 1979, 1984, 1989, 1994, 1999); the fall of the Iron Curtain starting in April 1989 in Poland and lasting until the end of that year when also the Berlin Wall fell; Dutch politician Bolkestein who held a speech about the integration of ethnic minorities in Dutch society on September 6 1991 and finally the happenings in Srebrenica in July 1995. To 2
  • 5. analyse the effects of those events, first an adequate ARIMA model is developed, which is followed by an impact assessment analysis of the different events. Results I specify in this results section how the ARIMA model was created and thereafter what the results were of the impact assessment analysis to include events in the analysis. ARIMA model The timeseries dataset is first being analysed according to the ARIMA-framework described by Vliegenthart (n.d.). The first step was to establish whether the data was stationary. Therefore, I made a graph was made (see Figure 1) to inspect the data in a convenient manner and three augmented Dickey Fuller (ADF) tests were used to check whether the series was a random walk without drift, a random walk with drift or a random walk with drift and trend (see Table 1). One of the results of this ADF indicates that the null hypotheses of non-stationarity should not be rejected and therefore it was necessary to integrate the data and check again for stationarity. This time only significant results were found; the integrated series was stationary (see Figure 2). Figure 1. The mean score on a political left-right dimension over time. Table 1. The various augmented Dickey-Fuller tests. Augmented Dickey-Fuller test Random walk without drift -0.374* Random walk with drift -12.779 Random walk with drift and trend -21.805 After integrating Random walk without drift -57.741 Random walk with drift -57.718 Random walk with drift and trend -57.696 Note. All tests without a * indicate the presence of an unit root. 3
  • 6. Figure 2. The differenced value for the mean score on a political left-right dimension over time. The next step was predicting the data as good as possible by accounting for its past, either with autoregressive (AR) terms, moving average (MA) terms or both. This was done by inspecting the autocorrelation (ACF) and partial autocorrelation functions (PACF) (see the graphs for both in Figure 2). Those functions suggest a moving average at lag 1; there was one peak in the ACF and a decaying order in the PACF. Therefore, a ARIMA(0,1,1) model was used which indeed has a significant effect for this moving average term. This model was tested for autocorrelation and heteroscedasticity with the Ljung–Box Q test statistic for both the residuals as well as the squared residuals. The (significant) results for both tests indicate that there was autocorrelation present and thus not only white noise. Therefore the model was extended with a moving average term at lag 2 for tho reasons. First, because the ACF and PACF showed a peak at the second lag, however, the graphs did not give a clear indication for adding an autoregressive (AR) or moving average (MA) term. Second, adding a MA term increased the fit more than adding a AR term according to the values of the Akaike Information Criterion (AIC) and Schwarz Bayesian Criterion (BIC), therefore I choose to use a resulting ARIMA (0,1,2) model. The values of this resulting model can be found in Table 2. The results of the Ljung– Box Q test statistic for this model was insignificant for the residuals, therefore we can assume the residuals to be white noise. However significant results were found for the squared residuals with the Ljung–Box Q test statistic, which indicates the presence of heteroscedasticity. However for this moment no attention is paid to this, but hopefully it will be solved with ARCH and GARCH models later in this Dynamic Data Analysis course. 4
  • 7. Figure 3. ACF and PCF for the differenced mean score on a political left-right dimension. Table 2. ARIMA model for mean score on a political left-right dimension over time. ARIMA (0,1,2) Constant -.000 (.000) MA(1) -.767 (.026)* MA(2) -.091 ( .026)* Ljung-Box Q(20) residuals 20.61 Ljung-Box Q(20) residuals² 103.34* AIC -3583.93 BIC -3563.49 Note. Unstandardized coefficients. Standard errors in parentheses; * p<.001 Impact assessment analysis I created various dummy variables were created to test if different events had an effect on the mean score for the political preference on a left-right dimension over time. The duration of those events could be temporary or permanent and the onset could be gradual or abrupt (McCleary & Hay, 1980). How the duration and onset were implemented was based on simple theory, common sense. The various events were incorporated in the ARIMA-model that was established before. Because the timeseries used as a dependent variable was differenced, also differenced values of the independent (event) variables were used in the analysis. Three models were the most interesting to study the impact of television developments on political preferences; the one that makes clear what happened after RTL4’s introduction, the one for the moment commercial broadcasters had permission to broadcast in the Netherlands, and the one that can be used for the gradual introduction of more commercial 5
  • 8. broadcasters in the Netherlands. However, as Table 3 shows, none of those developments had an effect on the mean score for the political left-right dimension. Table 3. Influence on media developments on the mean score on a political left-right dimension over time. Permission All TVchannels ARIMA (0,1,2) RTL4-model model intro-model Constant -.000 (.000) -.000 (.000) -.000 (.000) -.000 (.000) MA(1) -.767 (.026)* -.767 (.026)* -.766 (.026)* -.767 (.026)* MA(2) -.091 (.026)* -.089 (.026)* -.092 (.026)* -.090 (.026)* Introduction RTL4 .015 (.048) Permission for commercial broadcasting .020 (.026) Introduction different TV channels 0.007 (0.017) Ljung-Box Q(20) residuals 20.61 20.59 20.98 20.55 Ljung-Box Q(20) residuals² 103.34* 103.54* 103.71* 103.79* AIC -3583.93 -3582.18 -3582.36 -3582.22 BIC -3563.49 -3556.64 -3556.81 -3556.68 Note. Unstandardized coefficients. Standard errors in parentheses; * p<.001 To check whether this dependent timeseries variable is not influenced by any factor, the analysis were repeated for various historical events that were specified in the methods section that might have affected the political preferences of people. Also for those events the differenced series were used in the analysis. The results showed that only the second oil crisis in 1979/1980 had an effect on political preferences. As this effect was positive, it can be said that this crisis caused an increase in the preference for right-wing politics. Perhaps more remarkable is that all other variables had no effect on the average political preference in the Netherlands, both when tested alone in the ARIMA(0,1,2)-model or when tested together as shown in Table 4. Conclusion My main aim in this paper was to study the impact of the introduction of commercial broadcasting to Dutch viewers. It seems that this had no effect on the political preferences of Dutch people; the mean score on a left-right scale was not affected by the various events. To do this analysis first a proper ARIMA-model was build by integrating it one time and taking the first and the second lag into account with moving average terms. The resulting model did reflect white noise in the residuals, however the squared residuals did still correlate, which indicates heteroscedasticity. Thereafter, events like the introduction of RTL4, the moment of permission for commercial broadcasting and the introduction of different commercial channels were included in the model. All three however seemed to have no effect. Therefore, other (historical) events were also taken into account to see if this dataseries was not affected 6
  • 9. at all by events. Only the second oil crisis seemed to affect peoples’ political preferences. Most events however are unlikely to influence the average political preference on a left-right dimension; the media developments did not achieve it and also most historical events did not have an effect. Table 4. Influence on historical events on the mean score on a political left-right dimension over time. Historical events model Constant -.000 (.000) MA(1) -.767 (.026)** MA(2) -.091 (.026)** Introduction different TV channels .009 (.016) Train hijack -.025 (.054) Second oil crisis .036 (0.015)* European elections .014 (.017) Fall of Iron Curtain .022 (.027) Bolkestein’s speech .020 (.298) Srebrenica .013 (.036) Ljung-Box Q(20) residuals 21.43 Ljung-Box Q(20) residuals² 101.81** AIC -3575.97 BIC -3519.77 Note. Unstandardized coefficients. Standard errors in parentheses; *p<0.05, ** p<.001 References Aarts, K., & Semetko, H. A. (2003). The divided electorate: Media use and political involvement. The Journal of Politics, 65(3), 759-784. Curran, J., Iyengar, S., Brink Lund, A., & Salovaara-Moring, I. (2009). Media system, public knowledge and democracy. European Journal of Communication, 24(1), 5-26. Davies, G. (1999). The Future Funding of the BBC. Report of the Independent Review Panel to Department of Culture, Media and Sport. Retrieved on November 23, 2010, from http://news.bbc.co.uk/hi/english/static/bbc_funding_review/reviewco.pdf DellaVigna, S. & Kaplan, E. (2007). The Fox News effect: Media bias and voting. Quarterly Journal of Economics, 122(3), 1187-1234. McCleary, R., & Hay, R. (1980). Applied Time Series Analysis for the Social Sciences. London: Sage. Vliegenthart, R. (n.d.). Moving up. Applying aggregate level time series analysis in communication science. Unpublished manuscript. 7
  • 10.
  • 11. Appendix 1: Do file * declare data to be time series replace nr = nr + 766 + 52 + 52 tsset nr, weekly * making differences gen dleftright=leftright - l.leftright * graphic twoway (tsline leftright, lcolor(black)) * 3 tests for stationarity, 1)without constant=random walk without drift, 2)normal ADF: random walk with drift, 3) ADF trend: random walk with drift and trend dfuller leftright, noconstant dfuller leftright dfuller leftright, trend twoway (tsline d.leftright, lcolor(black)) dfuller d.leftright, noconstant dfuller d.leftright dfuller d.leftright, trend * arima model ac leftright pac leftright ac d.leftright pac d.leftright arima d.leftright, ma(1,2) estat ic predict r, res wntestq r, lags(20) gen r_s=r*r wntestq r_s, lags(20) ac r pac r ac r_s pac r_s drop r r_s * create intervention for RTL. eerste uitzending was op 2 oktober 1989 = week 40 gen RTL=0 replace RTL = 1 if nr>1520 *Veronica gen veronica=0 replace veronica = 1 if nr>1825 *rtl5 2 oktober 1993 gen rtl5=0 replace rtl5 = 1 if nr>1725 *sbs6 28 augustus 1995 gen sbs6=0 i
  • 12. replace sbs6 = 1 if nr>1823 *Media development gen media = RTL+ veronica+ rtl5+ sbs6 *in 1992 werden binnenlandse commerciële omroepen toegestaan gen comm=0 replace comm = 1 if nr>1635 *EU elections 1979,1984, 1989,1994,1999 gen euelect=0 replace euelect = 1 if nr>=991 & nr<=993 replace euelect = 1 if nr>=1245 & nr<=1247 replace euelect = 1 if nr>=1502 & nr<=1504 replace euelect = 1 if nr>=1759 & nr<=1761 replace euelect = 1 if nr>=2016 & nr<=2018 *Fall of iron curtain starting in Poland, then berlin wall, untill and of 1989, minus christmas weeks gen IronCurtain=0 replace IronCurtain = 1 if nr>=1493 & nr<=1529 *Bolkestein 6 september 1991 gen bolkestein=0 replace bolkestein= 1 if nr==1618 *Molukse acties: Treinkaping bij De Punt 23 mei (21)-11 juni 1977(23) gen moluk=0 replace moluk= 1 if nr>=904 & nr<=906 *2e oliecrisis: March 26 1979 (13): OPEC makes full 14.5 percent price increase for 1979 effective on April 1 *December 1980 (48): Collapse of OPEC's pricing structure. Saudis use $32 per barrel marker, others use $36 per barrel benchmark. gen oil=0 replace oil= 1 if nr>=983 & nr<=1067 *Srebrenica 11 juli 1995 (28) gen srebrenica=0 replace srebrenica= 1 if nr>=1818 & nr<=1828 arima d.leftright d.srebrenica, ma(1,2) estat ic predict r, res wntestq r, lags(20) gen r_s = r*r wntestq r_s, lags(20) drop r r_s arima d.leftright d.media d.euelect d.IronCurtain d.bolkestein d.moluk d.oil d.srebrenica, ma(1,2) estat ic predict r, res wntestq r, lags(20) gen r_s = r*r wntestq r_s, lags(20) drop r r_s ii