Social media, political donations and incumbency advantage in america
Maria Petrova, Ananya Sen, Pinar Yildirim
New communication technologies change just how people become informed and stay connected, and will also change voter behaviour. This column runs on the dataset covering 1,814 candidates for the united states Senate with Twitter accounts to analyse how utilizing a new social media technology can overcome the barriers of communicating with voters. Candidates receive more campaign donations once they join Twitter, but adopting the technology appears to help only new, inexperienced politicians. This shows that new technologies can ease entry to politics for new candidates and promote political competition.
The benefit incumbents experience in electoral races in america is well-documented. Politicians re-running for a seat achieve re-election rates of around 90%. Incumbents are included in the media more regularly and receive more endorsements in comparison to their opponents, and take advantage of the sources of the offices they held (Levitt and Wolfram 1997). Incumbency advantage creates a tall barrier for political newcomers and challengers, and pre-empts competition (Ansolabehere and Snyder 2000, Prat 2002, Prior 2006).
As well, researcher have discovered that that new communication technologies can transform voter behaviour (Falck et al. 2014, Campante et al. 2016), policy outcomes (Gavazza et al. 2015), and protest participation (Enikolopov et al. 2016, Qin et al. 2017), in fact it is possible these new technologies may also help politicians to see their constituencies also to raise money for political campaigns.
In a recently available study, we analyse one channel by which newcomers in politics can overcome the barriers of communicating with voters: using online networks (Petrova et al. 2016). Specifically, we study the impact of politicians’ adoption of a fresh communication technology, namely Twitter, on the campaign contributions they receive. Theoretically, social media give a relatively cheap technology for communicating with voters weighed against traditional media, particularly for politicians who think it is costly to attain out to voters to see them about their candidacy and policies. How politicians’ adoption of online networks influences electoral process used, however, remains largely unknown.
The study runs on the dataset connecting the Twitter accounts of just one 1,814 politicians who ran for the united states Congress between 20 to the political donations they received over this era, which are reported by the Federal Election Commission. Our analysis is targeted on the non-public Twitter accounts of politicians instead of accounts focused on a political campaign, and we only consider smaller amounts in political donations ($1,000 and under). This enables us to declare that the donations will result from ordinary citizens instead of lobbyists.
We compare the change in the donations received by politicians running for Congress before and after joining Twitter, exploiting the variation in Twitter usage across geographical locations in accordance with other websites. We make reference to this variation as ‘Twitter penetration’. Twitter penetration impacts the reach of a politician’s message, and therefore the profit she can receive upon adopting Twitter as a political communication channel.
The results claim that using Twitter helps politicians to improve more contributions, and a lot more in the areas with an increased Twitter penetration. Importantly, the upsurge in support is observed for politicians who’ve never been elected to the Congress before, however, not for the experienced candidates who’ve better opportunities to see and fundraise. We find that within per month of joining Twitter, the aggregate political contributions to the average new politician increases by 1-3% of most donations under $1,000 received throughout a campaign. Figure 1 demonstrates the primary finding of our study. You will find a discontinuous upsurge in donations after a politician joins Twitter and the shift is higher in areas where Twitter penetration or usage is higher in accordance with all sites.
Figure 1 . Aggregate standardised political donations received in low and high Twitter penetration areas (before and after five weeks of joining Twitter)
Our empirical strategy exploits the complete timing of opening a Twitter account through the use of variation in donations before and after joining Twitter and across regions of different Twitter penetration. We likewise incorporate politician-month fixed effects to regulate for politician-specific unobserved time-varying factors, such as for example being more progressive-minded, more tech-savvy, or coming to a different stage of campaigning.
To credibly identify the causal impact of Twitter adoption on campaign donations, several competing hypotheses should be eliminated. We specifically worry about the unobservable factors that could influence both a politician’s decision to become listed on Twitter and the donations she raises, including other campaign activities or exogenous events which coincide with enough time of opening a merchant account and vary systematically between low and high Twitter adoption regions. We test for a number of alternative explanations to find that there surely is no discontinuous upsurge in general campaign expenditure, advertising on TV, or media and blog coverage of the politicians around enough time of their Twitter entry, thus making certain the email address details are not driven by the partnership between campaign contributions and these variables.
How persuasive is communicating with voters over Twitter when compared to other channels of communication? We compare Twitter to other political communication and persuasion channels including direct mailing and TV advertising using the persuasion rate formula defined by DellaVigna and Gentzkow (2010). The measure summarises the percentage of receivers of a note, who change their action due to being persuaded by the message. We find that Twitter’s persuasion rate is just about 1%, which is to the persuasion rates of 1% for direct mailing (Gerber and Green 2000) and 0.1%-1% for political advertising (Spenkuch and Toniatti 2016). While Twitter’s effectiveness comes even close to that of direct mailing and short political ads on TV, remember that the price of advertising and direct mailing could be significantly higher in comparison to that of operating an individual social media account.
How come adoption of Twitter only raise the donations received by inexperienced politicians, rather than those received by the experienced ones? Communication via Twitter can – similar to advertising – both inform and persuade individuals (Nelson 1974). Specifically, information on Twitter could increase knowing of the candidates for potential donors who have no idea new candidates or their policy positions and will bring new donors to the candidates. Alternatively, microblogging through Twitter may persuade those that already are alert to the candidate and will raise the size of donations.
Our findings claim that it is much more likely for Twitter communication to improve awareness of less popular politicians among the voter base. Several findings from the analysis support this argument. First, we find that the upsurge in donations mainly originates from donors who’ve never donated to the candidate before rather than from those that did. Second, we show that increases in size are higher for candidates who are relatively less popular by comparing the change in donations for the politicians running for the home of Representatives and for the Senate. Since every state has only two senators, plus they are appointed for six years instead of the two-year terms of the representatives for the home, voters will tend to be more acquainted with the candidates for the Senate. Backing this hypothesis, we find that politicians running for the home experience an increased shift in support than those running for the Senate.
Another little bit of evidence originates from a comparison of the result of Twitter adoption in states with substitute political communication channels. Specifically, politicians running in states with lower newspaper circulation where communication through other political channels could be more limited visit a greater upsurge in political donations in comparison to politicians from states where newspaper circulation numbers are high.
Since tweets only include 140 characters, a meaningful analysis of their content could be difficult. Nevertheless, through textual and sentiment analysis, you’ll be able to report several observations. Parsing the textual information in tweets, we find that politicians who are sending more informative messages (by including a hyperlink/URL to more information) see higher gains from opening a merchant account. Moreover, use of a far more inclusive language (e.g. words such as for example “we” rather than “I”) correlates with higher gains. We also apply the linguistic inquiry and word count method (Pennebaker et al., 2015) to analyse the sentiment in tweets on emotional, social, and thinking styles 1 and show a politician with a ‘plugged in’ social style is much more likely to visit a greater upsurge in donations.
The election outcomes in america are thought to depend on three Ms: money, machine, and media. With political campaigns becoming more and more more expensive, an all natural concern is whether challengers have sufficient opportunities to talk to the electorate also to fundraise.
The broad implication of our study is that adoption and usage of social media 2 offers a comparatively cost-effective alternate technology to talk to the electorate, and reduces the gap in fund-raising opportunities between new and experienced politicians, which, subsequently, reduces barriers to entry to nation-wide politics and increases political competition.
Ansolabehere, S and J M Snyder (2000), “Old voters, new voters, and the non-public vote: Using redistricting to gauge the incumbency advantage”, American Journal of Political Science 44 (1), 17–34.
Campante, F, R Durante and F Sobbrio (2016), “Politics 2.0.: The Multifaceted Aftereffect of Broabdband Internet on Political Participation”, working paper.
DellaVigna, S and M Gentzkow (2010), “Persuasion: Empirical evidence”, Annual Overview of Economics 2, 643-69.
Falck, O, R Gold, and S Heblich (2014), "E-Lections: Voting Behavior and the Internet", American Economic Review 104 (7): 2238-2265
Enikolopov, R, A Makarin and M Petrova (2016) “Social Media and Protest Participation: Evidence from Russia”, CEPR Discussion Paper 11254
Gerber, A S and D P Green (2000), “The consequences of canvassing, calls, and direct mail on voter turnout: A field experiment”, American Political Science Review 94 (03), 653-663.
Halberstam, Y and B G Knight (2016), “Homophily, group size, and the diffusion of political information in internet sites: Evidence from Twitter”, Journal of Public Economics 143, 73-88.
Levitt, S D and C D Wolfram (1997), “Decomposing the resources of incumbency advantage in america House”, Legislative Studies Quarterly 22 (1), 45-60.
Nelson, P (1974), “Advertising as information”, Journal of Political Economy 82(4), 729-754.
Pennebaker, J W, R L Boyd, K Jordan and K Blackburn (2015), “The development and psychometric properties of LIWC2015”, UT Faculty/Researcher Works.
Petrova, M, A Sen, and P Yildirim (2016), “Social Media and Political Donations: New Technology and Incumbency Advantage in america”, University of Pennsylvania working paper.
Prat, A (2002), "Campaign advertising and voter welfare", The Overview of Economic Studies 69 (4), 999-1017.
Prior, M (2006), “The incumbent in the living room: The rise of television and the incumbency advantage in US House elections”, Journal of Politics 68 (3), 657-673.
Spenkuch, J L and D Toniatti (2016), “Political advertising and election outcomes”, Offered by SSRN 2613987.
 As the emotional categories relate with one’s positive to negative emotions, social style indicates a amount of social openness and engagement, and thinking style indicates the usage of logic or senses in expression of opinions.
 Importantly, our findings aren’t specific to Twitter, we find similar but slightly weaker results analysing the adoption of Facebook by politicians.