It’s the economy, stupid! David Gross, Holo Zheng, IMBA, Bar-­‐Ilan University, 2014-­‐2015

!
Introduc4on
As much as naConal security is considered to be the first duty of any government, naConal elecCons are generally decided on the issue of economic competence. This was most clearly seen in the UK General ElecCon of 1945, where aNer prosecuCng the Second World War successfully, it was widely expected that Winston Churchill was would be re-­‐elected with a handsome margin. Not only did the Labour party win, but it did so in a landslide, indicaCng that the public wanted economic issues to be placed at the top of the naConal agenda. In 1979, a similar phenomenon occurred, where aNer years of insCtuConal obstrucCon, strikes and poliCcal chicanery by the trade unions, Margaret Thatcher was returned as Prime Minister with a specific mandate to reform the economic landscape in the UK. The conduct of the Labour Government between 1997-­‐2001 in fiscal and economic terms went a long way in burnishing its poliCcal credibility when the BriCsh public were asked as to whether the government was behaving competently on economic issues, even when it when it was much more unpopular overall. The General ElecCon in Britain this year will see a stark choice between a ConservaCve Party that has brought the deficit down by a third, in the main by cuXng spending, with a Labour Party that will seek to raise taxes on the wealthier segments of society and increase public spending. The laYer are already facing an uphill baYle in terms of repairing the damage done to their economic credibility aNer losing the last elecCon. The primary importance of economic competence has also taken hold in Israel over the past five years. In the summer of 2011, there were unprecedented civil protests as to the cost of living in the 1

country. It was the primary subject of naConal elecCons in 2013, where the surprise of the elecCon was the nominally centrist party, Yesh ACd, gain nineteen seats as the second largest party in the Knesset. Their elecCon campaign was premised on fiscal discipline, reducing the deficit and doing away with budgetary imbalances that had accrued because of certain sectors of Israeli society to the detriment of the middle class. The upcoming naConal elecCons look to also be centred on economic issues, where the rise of yet another centrist party, Koolanu, led by a former CommunicaCons Minister, Moshe Kachlon, who previously succeeded in introducing compeCCon into the mobile phone market and reducing costs by 90%. Therefore, it would be sensible to suggest that having a good reputaCon for economic competence would be one of if not the foremost factor for conCnued electoral success.

!
Review
of Literature It has become a staple of poliCcal life to follow the trends of polls. Throughput the world polling firms, market research companies and similar organisaCons have made it their business to track public opinion on mulCple levels and planes. This data has become increasingly valuable for predicCng all manner of poliCcal happenings, which in turn has a massive impact on economic stability and the ability to rely on that stability in the long term. AggregaCng this data has been especially helpful to certain companies in correctly predicCng the outcomes of the 2008 and 2012 US PresidenCal ElecCons and the 2014 midterm elecCons. The best exponent of this is the founder of the FiveThirtyEight.com website, Nathan Silver.1 In terms of designing the method of data collecCon, one is slightly spoilt for choice with respect with the possible ways that we could construct a mechanism to collect sufficient data. Having said that

1 Retrieved from hYp://fivethirtyeight.com/interacCves/pollster-­‐raCngs/ 9/1/2015

2

any answer we have to the hypothesis will be as a result of construcCng a correlaConal study as opposed to an experimental one. YouGov is probably the largest internet market research company in the United Kingdom. In a recent poll it took between December 17th and 18th 2014 as to voCng intenCons, it took a sample of 1981 adults in Britain.2 Clearly, taking a sample is far more realisCc proposiCon than polling an enCre populaCon of over sixty million people. In addiCon, the sample was three Cmes larger than the minimum size of a sample (600) for it to be considered significant. The main findings can be seen below:

!
Westminster VI 2010 Vote Gender Age

Total

Con

Lab

Lib Dem

UKIP

Con

Lab

Lib Dem

Male

Female

18-24

25-3 9

40-59

60+

1981

482

566

99

259

591

465

407

961

1020

236

501

678

567

1981

449

590

94

264

559

519

410

998

983

155

385

847

594

! ! !
One
of the common of features of all polls that are taken across all firms is that the upper row takes into account the strength of intenCon to vote in the next elecCon, meaning that the raw values (in the second row) can be set against each other.

2 Retrieved from hYp://cdn.yougov.com/cumulus_uploads/document/6s63ripmxv/YG-­‐Archive-­‐Pol-­‐Sun-­‐

results-­‐181214.pdf 10/1/2015

3

Another leading market research company that maintains similar tracking polls is ICM, which usually publishes in the Guardian newspaper. Their methodology is broadly similar, though their sample is considerably smaller than YouGov’s.3*

! !
Total Con Lab Westminster VI 2010 Vote Gender Age Lib Dem UKIP Othe r Con Lab Lib Dem Male Female 18-24 25-34 35-64 65+

1001

195

222

53

93

70

262

229

102

490

511

122

164

509

206

1001

158

202

61

91

70

214

179

131

471

530

65

111

532

293

! ! !
What
is clear from both of the above polls is that there are a considerable amount of variables simply in the collecCon of voCng intenCons. These polls however, are not sufficient to work out what the underlying reasons to those intenCons are. VoCng intenCons are generally derived what the primary issues on the naConal agenda at the moment and which party is perceived to be best placed or equipped to meet the challenges. The graphic below shows how events moved electoral paYerns in the year following the Yom Kippur War and the subsequent arms embargo:4

3 Retrieved from hYp://www.icmunlimited.com/data/media/pdf/2014_dec_guardian_poll.pdf 10/1/2015

*The following table was edited with geographical informaCon not displayed.
4
Retrieved from hYps://www.ipsos-­‐mori.com/researchpublicaCons/researcharchive/3455/EconomistIpsos-­‐

MORI-­‐September-­‐2014-­‐Issues-­‐Index.aspx#gallery[m]/1/ 11/1/2015

4

!

!

!
The
subsequent graphic seemingly shows that the economy dropped to second place in September 2014 in the runup to the UK General ElecCons:5

!

5 Ibid.

5

While this would seem to indicate that the economy was not highest in the public’s consciousness, the reality is that much of today’s scepCcism is of immigraCon comes as a result of a saturated employment market, strain on public services and the inevitable higher taxes that would follow that would be set to alleviate the strain, which comes back to economic credibility of the government of the day.

!

6

! !
Design
In simple terms, the goal of this study is to see if one independent variable i.e. economic competence could be idenCfied to predict or even dictate the dependent variable – victory in a naConal elecCon. In theory, this should be able to be seen by a simple T-­‐test: By seeing which poliCcal party had more credibility on the economy, we would then see whether that would be mirrored into an elecCon victory. The nature of naConal polls mean that it would be a two group independent T-­‐test, as respondents express a preference for a party. On a conceptual level, “economic competence” is a broad and amorphous topic. In addiCon, it is only a small proporCon of the public are really in a posiCon to speak with authority on economic maYers and the relaCve competences of the various parCes, when in fact there is a parCcular sub-­‐category within the gamut of economic state of the country in quesCon, such as unemployment, career prospects and overextended government borrowing. For example, an Ipsos-­‐Mori poll in the United Kingdom takes a poll every month as to voters opinions regarding the importance of various issues. In May 2010, the month in which the General ElecCon was held, 71% of respondents held that the economy was of the most naConal importance.6 On the other hand, when quesConed as to whether the cost of living, food prices or inflaCon were of primary importance, the number of respondents who agreed only numbered around 10%. Yet this is perhaps too simplisCc a measure to really be able to say with any certainty whether this was the only reason or even a primary reason. It would be easy to claim that incompetence of one party would result in a de facto support for the other.

6 Retrieved from hYps://www.ipsos-­‐mori.com/researchpublicaCons/researcharchive/2905/Issues-­‐Index-­‐2012-­‐

onwards.aspx?view=wide#2010 11/1/2015

7

This requires us to suggest a more wide ranging method of saCsfying our hypothesis. The next method that comes to mind is the Analysis of Variance or ANOVA method. The ANOVA method seeks to where there are three opCons or condiCons of the independent variable. In this case it would be a parCcular poliCcal factor, idenCfied by the public to be the most important issue of the day, which in turn, would dictate the outcome of the elecCon, the dependent variable. The poll below was compiled in response to the following quesCon: “With the Economy facing the challenges it does in the months ahead, who do you trust more to steward the economy at the next General ElecCon, David Cameron and George Osborne – or Ed Miliband and Ed Balls?”7 The breakdown between all affiliated voters showed that all voters not affiliated with the Labour Party and therefore “Ed Miliband and Ed Balls” all perceive the current Prime Minister and Chancellor of the Exchequer to be more credible on the economy. While the outcome of the elecCon has yet to be decided, the fact that the Labour Party is not seen as credible on the economy means that they are seriously disadvantaged when it comes to implemenCng an elecCon campaign that will engage with this issue.

7 Retrieved from hYp://lordashcroNpolls.com/wp-­‐content/uploads/2014/10/ANP-­‐summary-­‐140627.pdf

18/1/2015

8

As the table below shows, the first response to the poll is as follows

A factorial ANOVA study would also be of parCcular use to poliCcal parCes when deciding the direcCon, tone and content of their campaign. Focussing on a parCcular issue which the public are 9

not either interested in or aware of to the detriment of other more central issues has proven criCcal in the outcome of an elecCon victory. A prime example of this would be to see the outcome of the UK General ElecCon in the 2001. The ConservaCve Party, having lost the previous elecCon resoundingly, based their enCre campaign on “Saving the Pound”.8 The response by the BriCsh public was unequivocal9:

!
Seats
Labour Conservative Liberal Democrats 412 166 52

Votes
10,724,895 8,357,622 4,812,833

% Share
40.7 31.7 18.3

! ! !
What
was even more impressive about this victory was that the Labour Party succeeded in increasing an already formidable majority because they campaigned on increased funding for public services while not overtly raising taxes, and more specifically not raising income tax, clearly, a much more posiCve, aYracCve proposiCon. In the upcoming UK General ElecCon, Ipsos Mori, who have been tracking the relaCve importance of issues at that moment, found that the state of the economy was not the most important issue of the day, but rather immigraCon.10 For the purposes of our discussion, what was even more interesCng was that the poll saw that the state of the NaConal Health Service on par as the state of the economy. This has seen the Labour Party, who suffer from

8 Retrieved From hYp://news.bbc.co.uk/1/hi/uk_poliCcs/1163591.stm 18/1/2015 9 Retrieved from hYp://www.ukpoliCcal.info/2001.htm 18/1/2015 10 Retrieved from hYps://www.ipsos-­‐mori.com/researchpublicaCons/researcharchive/2905/Issues-­‐Index-­‐2012-­‐

onwards.aspx?view=wide 18/1/2015

10

low credibility regarding the state of the economy, focus on healthcare as a means of gaining a poliCcal advantage. The most important issue as seen in that same poll was immigraCon.11 One important point when discussing the role of immigraCon in the context of an elecCon campaign is that the arguments for and against are almost always founded on economic arguments, whether as a posiCve in that many immigrants in the European Union context would do various menial jobs that BriCsh people would turn their noses up at. AlternaCvely, others would argue that too much immigraCon puts a strain on public services, as well as prevenCng naCve ciCzens from entering the job market, even at the most basic level. In either instance, it might be argued that all of these concerns really speak to several facets regarding overall economic prosperity, whether it be job security, unemployment or concern about the strain of public services. InacCon regarding immigraCon over the past twenty years has seen the rise of smaller parCes that have sprung up in reacCon, and both the ConservaCve and Labour parCes, despite it being the most important issue, have done their utmost not to engage the issue. We see, therefore, that the issues of the day have enormous influence to the content of the poliCcal agenda of poliCcal parCes and interact in mulC-­‐layered and someCmes counter-­‐intuiCve ways.

!
In
conclusion, having treated various opCons, our hypothesis should be phrased as follows: In order to correctly determine electoral victory, one has to idenCfy the most important electoral issue of the moment which will have most bearing on the public’s choice. Necessarily, the need to weigh up various opCons means that we are best served by insCtuCng a Factorial ANOVA test.

! !
11
Ibid.

11

! !
Method
We are using two tests in this design, independent T-­‐test and ANOVA. As for independent T-­‐test, we want to see if one independent variable i.e. economic competence could be idenCfied to predict the dependent variable – victory in a naConal elecCon. To operaConalize it, we firstly choose two parCes with different economic competence. Secondly, we use the group/party’s ranking in pools as an ordinal measurement of its economic competence, rather than its real power, which is hardly measurable, over specific economic issues. Thirdly, we measure elecCon victory by how many votes each party has. Last but least, we gather data from a certain number of elecCons during one specific Cme duraCon, and set the criteria. If the significance level is lower than 5%, then we call it significant, or in another word, the economic competence can be idenCfied to predict the victory in a naConal elecCon. In the process P value and t raCo will be calculated. To simulate a more realisCc scenario, we also use two-­‐way ANOVA, which enables us to find out if elecCon result is responsive to interacCon between two independent variables e.g. security concerns and economic competence. By so doing we divide the total variability among values into four components. Prism tabulates the percentage of the variability due to interacCon between the row and column factor, the percentage due to the row factor, and the percentage due to the column factor. The remainder of the variaCon is among replicates (also called residual variaCon). To operaConalize it, we firstly choose 3 parCes of different levels of economic competence ranging from very low to middle level to very high. In addiCon, we label their security concern into different level, also according to the pool, and gather the elecCon results during a certain period of Cme. By so doing we come up with 6 groups of numbers, which shows the interacCon between economic concerns, e.g. party with highest economic competence and the lowest security concern is more

12

successful in elecCon than party with slightly lower economic competence and a very high security concern. In the process an ANOVA table will be formed and P value will be calculated. Implica4ons The finding we have from the tests will help us understand if the economic competence of a party has a decisive influence on the elecCon result and whether the interacCon between economic competence and the security concern of a party can influence the elecCon result. This will not just help the party make beYer strategic decision on what their image should be built on, but also making pracCcal decisions how to split a limited budget between different issues. Moreover, knowing the interacCon between two factors can raise the cost efficiency of a coherent campaign, by using one factor fostering the other or at least stop it from being an obstacle for the other to thrive.

13