Do Political Protests Matter? Evidence from the Tea Party Movement*

TL;DR: This article studied the effect of the Tea Party movement in the United States, which rose to prominence through a series of rallies across the country on April 15, Tax Day, 2009 and found that large rallies cause an increase in turnout in favor of the Republicans in the 2010 Congressional elections, and increase the likelihood that incumbent Democratic representatives retire.

Abstract: Can protests cause political change, or are they merely symptoms of underlying shifts in policy preferences? This paper studies the effect of the Tea Party movement in the United States, which rose to prominence through a series of rallies across the country on April 15, Tax Day, 2009. To identify the causal effect of protests, we use an instrumental variables approach that exploits variation in rainfall on the day of the coordinated rallies. Weather on Tax Day robustly predicts rally attendance and the subsequent local strength of the movement as measured by donations, media coverage, social networking activity, and later events. We show that larger rallies cause an increase in turnout in favor of the Republicans in the 2010 Congressional elections, and increase the likelihood that incumbent Democratic representatives retire. Incumbent policymaking is affected as well: representatives respond to large protests in their district by voting more conservatively in Congress. Finally, the estimates imply significant multiplier effects: for every protester, Republican votes increase by seven to fourteen votes. Together our results show that protests can build political movements that ultimately affect policy, and they suggest that it is unlikely that these effects arise solely through the standard channel of private-information revelation.

Summary

II.A. Tea Party Goals and Organization

• The 1773 Boston Tea Party has been a potent symbol for American anti-tax activists over the past few decades, and its iconic value has regularly been exploited for protests and fund-raisers (e.g. Holmes 1991, Levenson 2007).
• Though the movement is unified by opposition to the Democrat-dominated federal government and mostly supports Republican candidates for office, it is not explicitly partisan.
• The outbreak of the 2008 financial crisis triggered a substantial policy response from both the outgoing Bush administration and the incoming Obama administration.
• As a broader protest movement started to take shape in the form of online and real-life “Tea Party” groups, plans for larger coordinated protests culminated in the first large national showing of activism on April 15, 2009 (Tax Day), when the groups held a large number of rallies across the United States.
• In the remainder of their paper, the authors study the significance of the 2009 Tax Day rallies to the effectiveness of these efforts in the year and a half that followed, up to the 2010 midterm elections.

III Data and Summary Statistics

• To construct their dataset the authors extract information from a number of sources in order to collect data on rainfall, Tax Day rally attendance, Tea Party activism, media coverage, political beliefs, voting outcomes, and policymaking.
• The following subsections present these sources and how they are matched.

III.A. Rainfall Data

• Information on precipitation comes from the National Oceanic and Atmospheric Administration and contains data from approximately 12,000 weather stations over the period 1980-2010.
• The authors construct their rainfall measure by aggregating the weather station data to the appropriate geographic level (county or congressional district) and then extract the mean daily rainfall (in inches).
• In their baseline measure, days with rainfall below 0.10 inches count as non-rainy; higher precipitation levels are defined as rainy.
• Based on historical weather observations for the period 1980-2008, the authors also create a measure of the probability that a county or congressional district experiences at least 0.10 inches of rain on a given day in April (see Section IV.A.).

III.B. Rally Attendance, Movement, and PAC Contributions Data

• If less-attended rallies occur in counties with small populations, and these events are more likely to be neglected, the measurement error would be correlated with population size.
• These groups maintain their own social networking sites, with minimal privacy protections, allowing the IREHR to collect data on a daily basis since 2010.
• Included are typically the leadership of local chapters.
• In addition to the membership measures the authors also gauge local Tea Party activism by including attendance data for 2010 Tax Day rallies from EconomyPolitics (2010).

III.C. Media Coverage

• To measure local media coverage of the protests the authors use news articles from the NewsLibrary database matched to Audit Bureau of Circulations county-level circulation data.
• Newslibrary.com archives over 4,000 titles, but not those of large national newspapers such as the Wall Street Journal or the New York Times.
• The authors collect information on all articles from newspapers with circulation over 15,000 containing the phrase “Tea Party” from January 1, 2009 through June 20, 2010 and merge these data to geographic regions using the county-level circulation information, ending up with 255 publications across 46 states.
• Over the full time period these publications contain some 40,000 articles including the term “Tea Party.”.

III.D. Political Beliefs

• To study whether the rallies affected public support for the Tea Party movement as well as political beliefs more broadly, the authors use the Evaluations of Government and Society Study from the American National Election Studies (ANES).
• Interviews were conducted in October 2010, weeks before the midterm election, and include a set of questions about the Tea Party and political opinions associated with the movement’s leaders.
• The data also contain socio-demographic variables, voting behavior in the 2008 election, and each respondent’s reported likelihood of voting in the 2010 midterm election.
• The authors have information for a total of 42 states at the congressional-district level that they match with rainfall, census, and survey data.
• Finally, sociodemographic county and district level data (income, population, race, immigrants, and unemployment) come from the 2000 and 2010 Census and the 2009 American Community Survey.

IV Empirical Framework

• The main challenge in measuring the effectiveness of political protests is that unobserved political beliefs or a culture of activism are likely to be correlated with both the number of protesters and other political behavior such as voting.
• This exclusion restriction seems plausible, though a valid concern is that bad weather may also make a rally less pleasant for actual attendees, energizing attendees and the consequent movement less.
• The authors also present reduced-form effects of protest day rainfall for all outcomes, where the exclusion restriction is not a necessary identifying assumption 7Table A.2 reports the county and district voting outcomes in levels.
• Again, there is no significant difference across rainy and non-rainy counties or districts.

IV.A. Specifications and Hypotheses

• To estimate the effect of the protests, the authors first investigate whether rainfall decreases attendance by regressing the number of protesters in a county on a dummy variable that indicates whether there was significant rain in the county.
• In order to exploit weather variation across counties with similar baseline likelihoods of rainfall on the protest day, the authors control for the rain probability flexibly.
• 10Dickens (1990) demonstrates that population-weighting in geographically grouped data is only desirable when group sizes are small enough that the variance of the average of individuals is larger than the variance of the group component.
• By estimating week-by-week effects using an equation analogous to equation (2), the authors can test whether rainfall affects media coverage immediately after the rallies, as well as whether there is an effect on later events that were important to the Tea Party movement.
• Pre-determined election covariates, from 2008, are dummies indicating whether the respondent voted for the Republican Party in the election for the House of Representatives.

IV.B. Exogeneity Check

• A key identifying assumption is that rainfall in equation (1) is uncorrelated with other determinants of political outcomes.
• As shown in Tables I, A.1, and A.2 rainy and non-rainy counties and districts are quite similar on average.
• Table II shows that the rainfall dummy in their specification is not significantly correlated with any of the pre-rally political outcomes.
• Table A.5 presents district level estimates for the regression equivalent of equation (5) for the ACU’s roll-call scores in years before the rally, where covariates analogous to those in equation (5) are included for roll-call scores and election outcomes in preceding years.
• Together, Tables II, A.4, and A.5 lend credibility to their identification strategy.

V Results

• V.A. The Effect of Rainfall on Rally Attendance Table III presents estimations of equation (1).
• To address the possibility of measurement error and under-reporting (see section III.B.), column (2) estimates the effect using the highest reported attendance across the three sources.
• Column (3) shows a significant relationship when the authors instead use the precipitation amount (hundreds of inches), and column (4) shows that rainfall decreases attendance when the equation is estimated on the sample of 542 counties for which there was a reported rally.
• The authors do so by dividing counties into categories reflecting whether they are Republican leaning, Democratic leaning, or swing counties in the 2010 midterm election, based on the predicted Republican vote share from previous elections and socio-demographics.
• 16Since the natural logarithm is undefined at zero, this regression is estimated on the sample of reported rallies.

V.B. Movement Outcomes

• One of the primary mechanisms through which protests are thought to influence policy is by strengthening associated political movements.
• Historically, it has been difficult to obtain data on this type of activity, but the Tea Party’s on-line-era birth allows us to measure local activism along some dimensions.
• When making the additional assumption that the effects are driven purely by larger rally attendance, the estimates of equation (3) imply that a one percentage point increase in the number of protesters causes a 0.093 percentage point increase in the share of the local population joining the Tea Party movement as organizers (column (2)).
• By excluding observations where there were zero votes for a party in the preceding elections (i.e. where races were uncontested).

V.C. Monetary Contributions

• Table IV, columns (5) through (8) above, presents the effect of rain on the day of the rally on contributions to The authors Country Deserves Better PAC.
• The estimate on the rain dummy in column (5) is for 2009 post-rally contributions, and imply that good weather caused approximately a $0.00032 per-capita increase (significant at the 5% level).
• 18This effect is quite large, but partly comes from the fact that attendance in 2010 was lower across the board than in 2009.
• Together with the effects presented in columns (1) through (4), their results show that political rallies can trigger both growth of and support for a movement as individuals volunteer as organizers and contribute monetarily, and that such effects can last for extended periods of time.

V.D. Media Coverage

• An additional mechanism through which protests can create support for a movement and further its policy agenda is media coverage.
• That is, if mass media report on the protests and the policies promoted by the protesters, the movement may itself be energized, or get the attention of the general population and, ultimately, policy-makers, who are known to respond strongly to news coverage (Eisensee and Strömberg, 2007).
• As expected, rain on April 15, 2009 has no significant effect on the level of media coverage prior to the Tax Day rallies, which are marked in red.
• For most of the sample, the measured effect is slightly negative (though close to zero) and statistically insignificant.
• Interestingly, all four statistically significant dates correspond to important events for the Tea Party movement.

V.E. Political Beliefs

• As mentioned in Section II, Tea Party protesters commonly display discontent with the state of affairs in the country, and the movement, broadly speaking promotes a conservative-libertarian political agenda.
• Table V presents survey evidence, from about two weeks before Election Day 2010, showing that the protests increase popular support for the movement, and that the local population in areas with large rallies adopts political opinions typically expressed by the protesters and the Tea Party’s leaders.
• 20 Respondents in non-rainy districts are approximately 6 percentage points more likely to express strong support for the Tea Party movement (column (1)).
• The results are robust to the exclusion of the controls, with similar point estimates, and significance at least at the same levels (results not shown for brevity).
• 21For example, she participated as a speaker in the rallies organized throughout the country as a part of the Tea Party Express Bus Tours of 2009 and 2010.

V.F. Election Outcomes

• Columns (3) and (4) show that there is little evidence that the protests impact votes for the Democratic Party, which suggests that the Tea Party protests, together with the consequent increase in media coverage and 22See Section VI for an extensive discussion of how the authors construct the average-effect measure for their study.
• The effects are non-trivial, implying that lack of rain increases the Republican vote share by 1.55 percentage points (significant at the 5% level), and that a 0.1 percentage point increase in the share of the population protesting leads to a 1.9 percentage-point increase in the Republican vote share.
• Finally, if the functional form is misspecified or the exclusion restriction of equation (2) is violated, the estimates would misrepresent the true average causal effect.
• According to their data across all 542 reported rallies, there were an approximate 440,000-810,000 individuals protesting nation-wide on Tax Day 2009.

V.G. Policy Outcomes

• Ultimately people care about political rallies and movements because they have the potential to 24The calculations are based on multiplying the total number of protesters with the per-protester scaled estimates .
• Columns (1) through (4) indicate that rain on the date of the rally has significant effects on voting records in 2009 and 2010, in spite of the fact that Representatives from rainy and non-rainy rally districts had similar voting records through 2008.
• As the ACU score is based on 24 roll call votes on which the ACU has an explicit position, with one vote for the conservative position giving a score of 100/24, the effect of non-rainy rallies corresponds to approximately 1 additional conservative vote in 2010.
• 25In column (1) the authors estimate the effect on the full sample of all congressmen.
• This suggests that the Tea Party protests were effective in shifting the electorate towards more conservative policies (as shown in Tables VI and VII), forcing some Democratic incumbents into retirement.

VI.A. Specification

• First, in Tables A.9a and A.9b the authors show that the main results are robust to how the covariates are specified.
• First, the authors calculate standard errors that account for spatial dependence parametrically, following the procedure developed by Conley (1999).
• Only 0.4% of the placebo estimates of the effect of rainfall on Tea Party protesters in 2009 are more negative than the actual estimate, and 1.0% of the estimates are larger in absolute magnitude.
• Finally, their findings do not rest on any individual result alone, but on the fact that so many different 34By requiring that there is at least some non-trivial amount of variation in rainfall across counties, the authors help avoid that each placebo estimate is driven by a few outlier counties.
• The standardized effect across all outcomes shows that the true average effect has a larger negative value than any given placebo draw, with only 2.9 percent of the placebos being larger in absolute magnitude.

VII.A. A Simple Information Revelation Model

• The authors can think of gc,t as corresponding to the left-right political spectrum on the real line, where a higher gc,t corresponds to more conservative roll-call voting.
• Suppose that it is always optimal for the incumbent to set policy gc,t equal to the median voter’s preferred policy.
• The authors assume that people protest sincerely, because they like to express their political preferences, and that the payoff from protesting, h(gi,c), is strictly increasing in the benefit of the proposed policy, h′ > 0.36.
• This difference will reflect the difference between incumbents’ expectations of the median voter’s bliss policy in the two types of districts: βt = E[gc,t(rain)− gc,t(sun)]= E[gc|rain]− E[gc|sun] (9) The authors question is what this framework predicts for the reduced-form effect of weather on policy, βt.
• The results in Table VIII show that the effects in 2010 are, if anything, larger than the effects in 2009.

VII.B. Alternative Mechanisms

• If learning does not fully explain their results, a natural question is what does.
• This would explain why a shift occurred in the voting population towards the conservative position, and why that shift went beyond those voters initially involved in the Tax Day rallies.
• This mechanism goes a long way in explaining their findings regarding incumbent behavior.
• Helpful if further research pinpointed the precise mechanisms through which protests affect voting behavior and policymaking, and under which conditions.

VII.C. Conclusion

• The authors show that larger political protests can both strengthen the movement they support, and help advance the political and policy agenda of the movement.
• In addition, the authors provide evidence that these effects were driven by a persistent increase in the movement’s strength.
• Protests led to more grassroots organizing, to larger subsequent protests and monetary contributions, and to stronger conservative beliefs, as documented qualitatively by Skocpol and Williamson (2011).
• The authors results suggest that political activism does not derive its usefulness solely from the provision of information or its consumption value, but that the interactions produced at rallies and protests can affect citizens’ social contexts in ways such that a movement for political change persists autonomously.
• Stockholm University Harvard Kennedy School American Enterprise Institute Harvard Kennedy School.

Figures

Table II. Exogeneity Check at the County Level

Table I presents summary statistics for our county level pre-rally variables.6 It shows that the counties that were plagued by rain are fairly similar in terms of past voting behavior, past donations to Our Country Deserves Better PAC, population, racial composition of the population, and unemployment compared to the rainless counties. The district level analogue, Table A.1, displays a

Table I presents summary statistics for our county level pre-rally variables.6 It shows that the counties that were plagued by rain are fairly similar in terms of past voting behavior, past donations to Our Country Deserves Better PAC, population, racial composition of the population, and unemployment compared to the rainless counties. The district level analogue, Table A.1, displays a

Full text

ZurichOpenRepositoryand
Archive
UniversityofZurich
UniversityLibrary
Strickhofstrasse39
CH-8057Zurich
www.zora.uzh.ch
Year:2013
Dopoliticalprotestsmatter?EvidencefromtheTeaPartymovement
Madestam,Andreas;Shoag,Daniel;Veuger,Stan;Yanagizawa-Drott,David
Abstract:Canprotestscausepoliticalchange,oraretheymerelysymptomsofunderlyingshiftsinpolicy
preferences?WeaddressthisquestionbystudyingtheTeaPartymovementintheUnitedStates,which
rosetoprominencethroughcoordinatedralliesacrossthecountryonTaxDay,April15,2009.Weexploit
variationinrainfallonthedayoftheseralliesasanexogenoussourceofvariationinattendance.Weshow
thatgoodweatheratthisinitial,coordinatingeventhadsignicantconsequencesforthesubsequentlocal
strengthofthemovement,increasedpublicsupportforTeaPartypositions,andledtomoreRepublican
votesinthe2010midtermelections. Policymakingwasalsoaected,asincumbentsrespondedtolarge
protestsintheirdistrictbyvotingmoreconservativelyinCongress.Ourestimatessuggestsignicant
multipliereects:anadditionalprotesterincreasedthenumberofRepublicanvotesbyafactorwellabove
1.Togetherourresultsshowthatprotestscanbuildpoliticalmovementsthatultimatelyaectpolicy
makingandthattheydosobyinuencingpoliticalviewsratherthansolelythroughtherevelationof
existingpoliticalpreferences.
DOI:https://doi.org/10.1093/qje/qjt021
PostedattheZurichOpenRepositoryandArchive,UniversityofZurich
ZORAURL:https://doi.org/10.5167/uzh-137648
JournalArticle
AcceptedVersion
Originallypublishedat:
Madestam,Andreas;Shoag,Daniel;Veuger,Stan;Yanagizawa-Drott,David(2013).Dopoliticalprotests
matter?EvidencefromtheTeaPartymovement.QuarterlyJournalofEconomics,128(4):1633-1685.
DOI:https://doi.org/10.1093/qje/qjt021

DO POLITICAL PROTESTS MATTER?
EVIDENCE FROM THE TEA PARTY MOVEMENT
∗
Andreas Madestam
Daniel Shoag
Stan Veuger
†
David Yanagizawa-Drott
July 2, 2013
Abstract
Can protests cause political change, or are they merely symptoms of underlying shifts in policy
preferences? We address this question by studying the Tea Party movement in the United
States, which rose to prominence through coordinated rallies across the country on Tax Day,
April 15, 2009. We exploit variation in rainfall on the day of these rallies as an exogenous
source of variation in attendance. We show that good weather at this initial, coordinating event
had significant consequences for the subsequent local strength of the movement, increased
public support for Tea Party positions, and led to more Republican votes in the 2010 midterm
elections. Policymaking was also affected, as incumbents responded to large protests in their
district by voting more conservatively in Congress. Our estimates suggest significant multiplier
effects: an additional protester increased the number of Republican votes by a factor well above
one. Together our results show that protests can build political movements that ultimately affect
policymaking, and that they do so by influencing political views rather than solely through the
revelation of existing political preferences.
∗
We are grateful to Edward Glaeser, Andrew Hall, Lawrence Katz, Nathan Nunn, Rohini Pande, Jeremy Stein,
Justin Wolfers, and three anonymous referees, as well as to seminar participants at the AEA/ASSA Annual Meeting,
the American Enterprise Institute, Americans for Tax Reform, BI Norwegian Business School, Bocconi University, the
Brookings Institution, the Federal Reserve Board of Governors, the Graduate Institute Geneva, the Harvard Kennedy
School, the Harvard Departments of Economics and Government, the Heritage Foundation, MIT, and the Stockholm
Institute of Transition Economics for valuable comments. Emma Bennett, Annalise Blum, and Itai Nixon provided
excellent research assistance. We also thank Devin Burghart for providing us with Tea Party membership and rally
attendance data.
†
Corresponding author: American Enterprise Institute, 1150 Seventeenth Street NW, Washington, DC 20036, stan.
veuger@aei.org.

I Introduction
How does political change come about? While freedom of speech and assembly are central pil-
lars of democracy, recognized as intrinsically valuable, it is unclear how effective the exercise of
these freedoms is in bringing about change. Though there are numerous historical episodes where
political change has been associated with political demonstrations, such as the French Revolution,
the Civil Rights movement, and the recent Arab Spring, it is unclear to what extent these protests
caused change. Protests are likely to occur alongside other changes in society, and it is difficult
to disentangle whether they cause political change or simply reflect unobservable changes in pref-
erences and beliefs. Empirical evidence of the causal effects of protests is scarce. In fact, to our
knowledge, there is almost no empirical work quantifying the causal effects of protests on subse-
quent political outcomes. It is an open question to what extent political protests can cause political
change, and this paper sheds light on these issues.
More specifically, we investigate the impact of the Tea Party movement protests in the United
States on policymaking and citizen political behavior. The Tea Party movement is a conservative-
libertarian political movement in the United States that has organized protests and supported can-
didates for elected office since 2009. This setting is a well-suited testing ground for hypotheses
regarding the effectiveness of political protests and one of the few such settings for which exten-
sive data are available. The movement propagates an agenda that is systematically to the right
of the status quo, which makes the measurement of policy changes in the direction desired by the
movement straightforward. In addition, the largest protests in the early stage of the movement were
the nation-wide 2009 Tax Day Rallies. As this date was preset, it allows us to test whether the size
of local protests on Tax Day affected subsequent local political outcomes.
The main empirical challenge in estimating the impact of protests is that unobservable political
preferences are likely to determine both the number of protesters and policy outcomes. A naive
2

regression of policy on protest size is therefore unlikely to reflect a causal effect. We address this
problem by exploiting variation in rainfall during the day of the protest. The idea is simple: people
are more prone to participate in protests if it does not rain. Conditional on the likelihood of rain,
rainfall is a random event, arguably uncorrelated with other factors that affect political outcomes.
Under the assumption that absence of rainfall affects policy and voting behavior only through the
number of protesters, this allows us to estimate the impact of protest size using an instrumental
variables approach. Even when relaxing this assumption, our estimates demonstrate the overall
importance of these initial events to the movement’s success.
We use data from a large number of sources to measure the influence of the Tax Day protests on the
Tea Party. The importance of the initial protests to local movement strength is evident in outcomes
as diverse as participation in Tea Party online social networks, political action committee contri-
butions, the number of protesters at subsequent protests, and survey measures of local political
beliefs. We show that these political protests and the movements they build affect policymaking
and voting behavior as well. Incumbent representatives vote more conservatively following large
protests in their district, and a rain-free rally in a district increases the likelihood that a Democratic
incumbent retires. Larger protests increase turnout in the 2010 elections, primarily favoring Re-
publican candidates. In particular, our baseline estimate shows that a 0.1 percentage-point increase
in the share of the population protesting corresponds to a 1.9 percentage-point increase in the share
of Republican votes. The Tea Party protests thus seem to have caused a shift to the right in terms
of policymaking, both directly and through the selection of politicians in elections.
In addition to providing exogenous variation in rally outcomes, variation in rainfall can be used to
assess the statistical significance of these results. We compare the effect of rainfall on the true date
of the rally to the distribution of placebo estimates of rainfall on other days. We find that none of the
placebo dates in 1980-2008 produce a cumulative effect as large as the effect estimated for Tax Day
2009. This finding, when combined with numerous additional robustness checks, demonstrates the
3

reliability and significance of the results.
Our results relate to the large body of empirical and theoretical work that has attempted to explain
which factors drive political participation. Most empirical work on why people vote has identified
simple correlations between political activism and citizen characteristics (see e.g. Blaise 2000 for a
review). Papers that inform us about the the determinants of protest participation include Cicchetti
et al. (1971), Finkel and Opp (1991), and Finkel and Muller (1998), but there is little research
on the causal impact of political rallies. An exception is Collins and Margo’s (2004, 2007) work
on the effects of the riots following the assassination of Martin Luther King Jr. on income, labor,
and housing market outcomes for African-Americans. Similar in spirit to this paper, they exploit
rain during the month of April 1968 as an instrument for riot severity. Madestam and Yanagizawa-
Drott’s (2011) use of daily rainfall to generate variation in outdoor participation on the Fourth of
July to study the impact of celebrating Independence Day is another example of such an approach.
Theoretical work has generally suggested that a sense of civic duty or consumption value drives po-
litical involvement (Downs 1957; Riker and Ordeshook 1968; Coate and Conlin 2004; Feddersen
and Sandroni 2006). Political theorists rationalizing why people protest offer explanations based
on the importance of peer pressure within smaller political groups (McCarthy and Zald 1977; Uh-
laner 1989; Oberschall 1994), on people’s (unrealistic) perception that that they can be politically
influential (Opp 1989), and on bandwagon effects (Kuran 1989). However, these results leave the
question of why protests would matter as instruments for political change unanswered.
One attempt to answer this question focuses on social dynamics within groups and networks of cit-
izens, and their influence on individuals’ desire to attain certain political goals (Zuckerman 2005).
Another influential strand of papers, written by Lohmann (1993, 1994a, 1994b), emphasizes the
role of information.
1
Lohmann (1993, 1994a) models the role of visible political activism in reveal-
1
See also Bueno de Mesquita (2010) for an information model where a revolutionary vanguard engages in public
violence to mobilize protesters.
4