Book Review-US voting

Blind Retrospection Electoral Responses to Drought, Flu, and Shark Attacks

Christopher H. Achen

Deparlment of Politics, Princeton University ;-' :;t i;''';

,r ''' 1 ';' -'

: '

Larry M. Bartels

Department of Politics and Woodrow Wilson School of Public and International Affairs.

Princeton University r ;iti'ii;i' i ti l: J;t' ji,ir ",)t{i Jr' i i.,itl

Revised: 27 January 2004

Abstract

Students of democratic politics have long believed that voters punish incumbents for hard times.

Governments bear the responsibility for the economy in the modern era, so that replacing incompetent

managers with capable alternatives appears to be a well-informed, rational act. However, this vision of

a sophisticated retrospective electorate does not bear close examination. We llnd that voters regularly

punish governments for acts of God. including droughts, floods, and shark attacks. As long as

responsibility for the event itself (or more commonly, for its amelioration) can somehow be attributed

to the government in a story persuasive within the folk culture, the electorate will take out its

frustrations on the incumbents and vote for out-parties. Thus, voters in pain are not necessarily

irrational. but they are ignorant about both science and politics, and that makes them gullible when

ambitious demagogues seek to profit from their misery. Neither conventional understandings of

democratic responsiveness nor rational choice interpretations of retrospective voting survive under this

interpretation of voting behavior.

Blind Retrospection

Electoral Responses to Drought, Flu, and Shark Attacksr

And Moses stretched forth his rod over the land of Egypt, and the Lord brought an east

wind upon the land all that day. and all that night; and when it was morning. the east wind

brought the locusts. And the locusts went up over all the land of Egypt, and rested in all the

coasts of Egypt: very grievous were they; befbre them there were no such locusts as they,

neither after thern shall be such. For they covered the face of the whole earth, so that the land

was darkened;and they did eat every herb of the land, and allthe fruit of the trees which the

hail had left: and there remained not any green thing irT the trees. or in the herbs of the field,

thror-rgh all the land of Egypt.

Then Pharaoh called for Moses and Aaron in haste; and he said, I have sinned against the

Lord l orrr God. and against you.

-Exodus 10: 13- l6 (King James version)

When collective misfoftune strikes a society, somebody has to take the blame. For ancient Israel,

disasters were God's punishment lbr sin-perhaps the ruler's sin, perhaps Israel's. Theology did not

single out the guilty pafiy, but it structured the search and set limits on what counted as a credible

explanation.

1 Earlier versions of this repoft were presented at the 2002 Annual Meeting of the Arnerican Political Science

Association and in seminars atthe Juan March Institute, the University of Michigan's Centerfbr Political

Studies, and Princeton University's Center for the Study of Democratic Politics. We are grateful to various

colleagues and friends for helpfLrl discussions and comments; to Jonathan Ladd for organizing data; to Helene

Wood fbr graphical assistance;to Sasha Achen, Bryan Jones, John Londregan. Arlene Saxonhouse. W. phillips

Shively, John Wilkerson, and Natasha Zharinova for pointing us to examples and references; and to John

Blydenburgh and David Mayhew fbr help with New Jersey political history. Achen also expresses his thalks to

the Center fbr the Study of Democratic Politics. Princeton University, and to the Department of Political

- Science, University of Michigan, for their financial sr-rpport of a fellowship year.

In the theology of classical Egypt, pharaohs were divine beings responsible for making the Nile

flood annually. When it failed to do so, as happened repeatedly in the famines and political disorder of

the First Intermediate Period (ca. 2200 BCE), some scholars believe that the pharaoh's reign was

shortened, and perhaps his life as well (Bell 1971; Hassan 1994).

Through the centuries, competitors of the ruler have been well aware that disaster presents them

with an opporlunity. When disasters take on catastrophic dimensions. not just the ruler but the entire

regime can come under suspicion. Writing of the Black Death in the 14th century, which may have

killed a third or more of the European population, Herlihy (1997.64) remarks:

The plague also discredited the leaders of society, its governors, priests, and intellectuals, and the

laws and theories supporled by them. These elites were obviously failing in their prime social

function, the def-ense of the common welfare. in the name of which they enjoyed their privileges.

During the plague years, spontaneous religious and political movements arose to threaten church

and government. A few of these bands agitated against the groaning inequities of medieval society.

Most were less attractive. Some, fired by conspiracy theories, targeted disliked minorities such as

beggars and Jews. Across Europe. thousands of Jews were murdered befbre the traditional spiritual

and secular authorities managed to halt the fury of the mobs (Herlihy 1997,66).

Contemporary democratic rulers have little aura of divinity about them, nor have they faced

Biblical famines or medieval plagues. Nonetheless, the citizenry continues to hold them responsible

for routine hardships and misfbrtune when election time comes. In this paper we examine electoral

responses to natural disasters that are clearly beyond the control of incumbent politicians. We find that

voters regularly punish incumbent governments for such events. as long as they can find some

psychologically appealing connection-whether plausible or not-between the disaster and the

government.

fsection on "pocketbook voting!!-v61g1s punishing the incumbents for hard economic times-

omitted herel

Shark Attacks in New Jersey, 1916: The voters Bite Back

On the four-day Fourth of July weekend in 1916, the beaches of New Jersey were packed with

crowds happy to escape the summer heat of nearby cities.2 On Saturday, July l, a young Ivy League

graduate from Philadelphia, Charles Vansant, was swimming jLrst beyond the breakers in four feet of

water at Beach Haven. He was attacked by a shark. Skillful lif'eguards managed to get him to shore,

but he died soon after from blood loss.

Five days later, a young Swiss bellhop named Charles Bruder, a strong swimmer like Vansant,

also ventured out past the lifelines at Spring Lake beach, some forty five miles nor-th of Beach Haven.

He, too, was attacked by a shark. Though rescued by lifeguards in a small boat, he died of his wounds

befbre reaching shore.

Nearly all of the diminished numbers of Jersey Shore swimmers stuck close to shore in the days

after the two deaths. However, no one worried about boys swimming in a creek on July 12 in the town

of Matawan, about two miles fiom open water. One was attacked and killed by a shark, as was a

young man from the town who dove in to recover the boy's body. Downstream, another group of boys

were swimming at the same time in ignorance of the attacks. Within half an hour, one of them had his

leg mauled by a passing shark. However" he was quickly pulled liom the water. reached the local

hospital, and survived.

By this time, the mounting panic reached a crescendo. Even the distant San Francisco Chronicle

had a front-page headline on.Iuly 14: EAST COAST BEGINS wAR oN RAVtrNOUS MAN-

' Unles, otherwise noted, the historical renditiorr follows Fernicola (2001). the most cornplete account. See also Capuzzo (2001).

EATERS (Fernicola 2001, 87). Steel mesh was being installed at beaches. Bounties were offered. and

sharks were killed in sizable numbers along the shore. Finally, one great white shark was hauled in

near Matawan Creek with what appeared to be human bones in its stomach. Perhaps for that reason,

the attacks stopped. ending the most serious string of shark-related fatalities in American history.

Before the attacks, no arm of government had patrolled for sharks or set up barriers against them

in New Jersey, since there had never been a recorded shark attack in the history of the state. Indeed"

prominent American scientists doubted that unprovoked shark attacks on human beings ever occurred,

certainly not as t-ar north as New Jersey.3 (Fernicola 2001, 22). The general climate of skepticism lecl

the l/ev' York Times to place its article about the flrst attack only on page 18, headlined "Dies After

Attack by Fish"-no doubt a consolation to the New.lersey resort owners, who were anxious to avoid

publicity.4

In the aftermath of the attacks. governments, particularly the federal government, were called on

fbr help. The resorts were losing money rapidly, with a quarter million dollars in reservations

cancelled within a week. Some resorts had 75 percent vacancy rates in the midst of their high season

(Captzzo 2001. 274). Losses may have amounted to perhaps as much as $1 million for the season

altogether, a sizable sum in 1918 (Fernicola 2001 .174). Letters poured into Cor-rgressional offlces

from the affected counties, demanding federal action, though there was little any government agency

'' Indeed. two scientists who were later called in to investigate the attacks. Dr. Jolrn T. Nicols. an ichthyologist and director of the Fishes Wing of the American Museum of Natural History. and Dr. Frederick Lucas, director of the museum. had recently coauthored with athird scientist an article arguingthat unprovoked sharks never attack human beings.

a Parallels to the film "Jaws" arrd its sequels are no accident. Peter Benchley. the author of the book on rvhicl.r

the film was based, is a New Jersey resident, and the film version. thor-rgh set on Long Island, New york.

inclr-rdes a reference to the l9l6 New Jersev attacks.

could do. Fernicola (2001, 70) describes the atmosphere, as the shark attacks entered popular imagery

and became a metaphor for other political crises as well:

Newspaper cartoons now poftrayed Wilson's chances fbr reelection in November, using the shark

fin as the symbol for his potential loss. The black fin labeled "def-eat" was shown slicing through

shark-inf-ested northeast regions. Other political cartoons of the day showed lawyers, represented

by sharks heading toward a beleaguered sailboat, embossed with "Union Bank." At the stern of

the bank boat. a chewed and legless victim dangled over the gunnel depicting "deposits."

As it happened, the Secretary of the Treasury, William McAdoo, had a summer home in Spring

Lake and was in residence at the time of the second attack. Joseph Tumulty, Wilson's powerful aide

fbr political affairs. had a summer home in Asbury Park, about five miles norlh of Spring Lake.

President Wilson himself, a former president of Princeton University and former governor of New

Jersey, had been looking for a summer White House in New Jersey as well, and chose a hotel in

Asbury Park, moving there shortly after the attacks ended. Thus the attacks received immediate

federal attention.

Wilson held a Cabinet meeting to discuss the attacks (Fernicola 2001, 70), but the Bureau of

Fisheries could suggest nothing beyond killing sharks at random and warning bathers. "No cefiainly

efTective preventive measure could be recommended," they said (Capuzzo 2001, 277). The president

could only direct the Coast Guard to inspect the beaches and patrol the water. However, the problem

disappeared and autumn arrived before much could be done. By election time in November, Wilson

ll'as back at his Asbury Park headquarters, but other election issues, notably potential U.S. entry into

World War I. took over the headlines (Link \954.247 -251). In the end, Wilson lost nearly all the

northeastern and Great Lakes states, including New Jersey, but managed to squeak out his re-election

by adding most of the Great Plains. Mountain States and West to the Democrats' customary Solid

South.

Did the shark attacks influence the presidential election in the affected areas of New Jersey?

Hitherlo, sharks have not been suspects in any electoral analysis. Nonetheless, if our argument is

correct. they should have reduced Wilson's vote. First. the attacks were a natural disaster causing

several deaths plus considerable emotional and financial distress to entire communities. Second, the

government was thought to be responsible for dealing with the crisis, and high federal officials were

present at the scene from the beginning. Third, the election lbllowed the crisis quickly enough that the

summer's events would have been fresh in the minds of the voters. The fact that no government has

any influence over sharks should have been irrelevant.

The evidence for a shark effect turns out to be rather strong. We now-turn to the first piece of

that evidence, using election returns at the New Jersey county level. The Wilson vote in 1916 is the

variable to be explained. Our key independent variable is "beach county," defined as Monmouth,

Ocean, Atlantic. and Cape May counties. These were, and are, the classic "Jersey Shore" counties

listed in the guidebooks, whose beach areas are heavily dependent upon summer tourism. They are the

places in which the shark attacks would have had the most pronounced economic effects. The attacks

themselves took place in Monmouth (three deaths) and Ocean (one).

[Aseveral-pagediscussionofvariousregressionanalysesisomittedhere@

irg The main finding is:]

The estimated negative effect on Wilson's vote in the beach counties is a little more than 3

percentage points" with a 95% confidence interval confined between 1.2 and 5.2. The shark attacks

indeed seem to have had an impact-about one-fourth the eftbct that the Great Depression had on

Herberl Hoover's vote in New Jersey 16 years later.5

' Hoover's vote share in New Jersey fell from 59.8%in l92B to 41.6o/o ir"r 1932.

6

We underlook two additional investigations with different samples. First, we examined the vote

in the first two shore townships where the attacks took place.6 Both Beach Haven and Spring Lake

were small, stable communities. making comparison sensible.T Figure 2 shows the vote change fbr

Wilson between l9l2 and 1916 in these two communities, and compales it with the change in their

respective counties and in New.lersey as a whole. Both townships show remarkable drops in Wilson's

support. 11 points in Beach Haven and 9 in Spring Lake, far more than the negligible changes in the

Wilson vote in their counties and in the state. These are vote losses equal to those Herbert Hoover

suffbred statewide in New Jersey in 1932 at the height of the Great Depression. It is apparent that

something drastically reduced enthusiasm for Woodrow Wilson in these two townships.

*** Figure 2 ***

We also investigated whether Beach Haven and Spring Lake were typical of beach areas. To

answer this question, we examined the townships in Ocean County near the water. Ocean was chosen

because it has many beach communities, nearly all on a bank of land clearly separated from the

mainland. Thus there is no difficulty in separating those seven communities right on the beach from

the twelve near the beach but not on it.8 The western border of the near-beach area was set to the

.r "{_Matarvalr Township and Matarvan Borough, where the final two shark deatlrs occurred, were excluded from

this analysis since they are not beach resort communities and thr-rs suffered no widespread economic loss from

their shark attacks or anyone elt"'t] [n any case, the rapid growth in the number of voters in both places between 1912 and 1916 makes comparison impossible;morethan aquarterof the l9l6 voters in Matawan township had not been there in 1912.

7 Beach Haven cast I I 2 votes fbr president in 1 912 and 1 1 9 in 1 9 1 6. The corresponding numbers for Spring

Lake are 271 and265.

t One beach township, Sea Side Park. apparently split into two between I912 and 1916 and jointly nearly

doubled in size; we dropped it from the analysis.

current New Jersey turnpike, which runs within a few miles of the shore in Ocean County. These two

areas had nearly identical Democratic percentages for Wilsoninlgl2 (36.3% at the beach and34.loh

in the near-beach), and thus are comparable.

In each area, we compared Wilson's vote percentages in 1912 and 1916. If the argument of this

paper is correct, the beach voters should show the largest drop in support for Wilson, while the near-

beach citizenry should be largely unaffected. The actual vote change turns out to be a drop of 8.2

percentage points in the beach area. compared to a tiny 0.2 percentage point gain in the near beach, an

easily statistically significant difl'erence.e Again, we flnd that disaft-ection lbr Wilson was widespread

in the beach areas whose livelihood was most directly affected by the shark attacks. far different from

the otherwise comparable areas next door. where Wilson's vote was nearly constant.l0

In summary, then, every indication in the New Jersey vote returns fbr 1912 and 191 6 is that the

horrifying shark attacks during the summer of 1916 reduced Wilson's vote in the fall. Retrospection

here was surely blind. If bathers insist on swimming in the ocean, governments then and now can do

nothing about shark attacks, as the subsequent attacks in New Jersey in 1960 and the regular

encottnters in Florida, California, South Africa. and Australia demonstrate (Fernicola 2001, ch. 5).

" Therervere3llbeach and2645 non-beachvoters in1912.and349 beachand2B5gnon-beachvotersin 1916. This comparison includes Point Pleasant Beach Boror,rgh as pafi of the near-beach. In spite of its name, the

overwhelming bLrlk of its population lived in Point Pleasant, which is not on the beach. However^ this

borough's30percentincreaseinthevotefrom 19l2to lgl6isthelargestofanybeachornear-beach

community. rnaking its two presidential years less comparable and suggesting that it shor,rld be excluded fiom

the analysis. If Point Pleasant Beach Borough is excluded, the near-beach vote change alters fiom +0.2

percentage points to -0.6, still very different from the -8.2 effect at the beach. Sirnilarly, if all beach and non-

beachcommunitieswithmorethan20%o increaseinthevotefrom l912to lgl6areexcluded,thenon-beach

vote change becomes -0.5, while the beach change is -l l.B. In shoft, these alternate versions of the sample lead to precisely the same substantive conclusion.

"' Tlie sarr-re finding fiom the Ocean County township sample holds when rnedians are used irr place of means, and when (weighted or unweighted) regressions are run with the townships as units of observation.

Shark attacks are natural disasters in the purest sense of the term. and they have no governmental

solution. Yet the voters punished any*ay. "

Of course, it is possible that the voters did not blame the government for the attacks themselves.

but did blame it for not helping them with their economic distress. In that case, retrospection might not

be blind. No doubt voters told themselves something like that at the time. Yet in the case of the

sharks, it is not clear what the government could have done to help the local economy. The truth could

not be covered up. The vacationers could not be compelled to come to the beach, nor could the sharks

be forced to stay away. Of course, from the perspective of a century later, it is obvious that extending

welfare benefits and unemployment compensation would have helped. But these social programs did

not exist at the time, they could not have been put in place quickly, and no one would expect them to

be enacted in response to a single local disaster in any case. In sum, for the case of the New Jersey

shark attacks, "failed disaster assistance" seems a w-eak hypothesis driven by insufficient historical

perspective.

[section on droughts and 1918 flu epidemic omitted here]

Conclusion

Our account of democratic politics strikes directly at key assumptions in two different

contemporary schools of thought. Perhaps most obviously, it questions the ability of ordinary citizens

to assess their public life critically, listen to the proposals for change coming from contenders for

public office" and then choose between the candidates in accordance with their own values. Like most

survey researchers who have talked extensively to real voters, we believe that few such citizens exist.

" On l7 December 1967 Australian prime minister Harold Holt disappeared while swimming in shark-infested waters at Cheviot Beach near Poftsea, Victoria. His body was never found. Being devotees of democracy,

however, we disapprove of this apparent attempt by the sharks to cut outthe middleman.

The present paper is one more item of evidence. The central fact about democracies is that the voters

understand little beyond their own and their community's pain and pleasure, and they think about

causes and effects as the popular culture advises them to think. The romantic vision of thoughtful

democratic participation in the common life is largely mythical. Democracy must be defended some

other way, if it is to be defended at all.

Our work also strikes a blow at the customary tallback position for contemporary defenders of

democracy, namely the view that the voters may know very little, but they can recognize good and bad

government performances when they see them. Hence they can choose retrospectively in a defensible

way. In most recent scholarly accounts, retrospection is a natural and rational feature of democratic

politics. In our view it is natural but not so obviously rational. Voters operating on the basis of a

valid, detailed understanding of cause and ellbct in the realm of public policy could reward good

performance while ridding themselves of leaders who are malevolent or incompetent. But real voters

oflen have only a vague, more or less primitive understanding of the connections (if any) between

incumbent politicians' actions and their own pain or pleasure. As a result, rational retrospective voting

is harder than it seems, and blind retrospection sometimes produces consistently misguided patterns of

electoral rewards and punishments.

What we have not oft-ered here is any systematic account of the circumstances under which

citizens will find-and accept-a cultural understanding that holds public officials responsible for

changes in the public's welfare. We know that the framing of news by the mass media may increase or

decrease the likelihood that citizens will attribute responsibility for social problems to the government

(Iyengar 1991). We know that politicians themselves may be more or less successful in "managing

blame," exploiting competing explanations to exonerate themselves (McGraw 1991). l2 These

rr Ideological commitments may play a significant independent role in elite constrr"rctions olexplanations fbr natural or social disasters, as with the Federalists' and Republicans'competing explanations of the yello1v Fever

epidemic of 1793 (Pernick l9l2). Physicians "divided bitterly overthe cause of the epidemic," with

10

alternate explanations are always present: some medieval towns blamed the plague on prostitutes,

beggars. or foreign agents (Herlihy 1997 , 65-67); some New Jersey residents in 1916 thought that

German U-boats might have induced the sharks to attack (Fernicola 2001, 166-170'); some Americans

in the grip of the Spanish Influenza pandemic two years later feared that "plague germs were inserled

into aspirin made by the German drug company Bayer" (Kolata 1999, 3).

When is one explanation accepted rather than another? Much seems to depend on plausibility

within the folk culture. Unfortunately, a general theory of political accountability explaining when and

why specific attributions or evasions of responsibility actually work is nowhere in sight. The

development of such a theory strikes us as a very high priority for students of democratic politics.

We end, then, on a discouraging note. For those who take the evidence about voter capacities

seriously, neither Rousseau nor Downs will save us. Democracies take their electoral direction liom

human beings with fewer capacities for self-government than either writer imagined. Under sufficient

pressure, those voters may lash out blindly. Such events are not bizarre historical footnotes rendered

irrelevant by modern education and hygiene. They are inevitable consequences of human cognitive

limitations-limitations which democratic government has not altered. Thus, as Sophocles taught and

as the destruction of the Weimar Republic reminds us, when the inevitable hard times appear, tragedy

may ensue.

Republicans generally attributing it to poor sanitation, climatic conditions, and the unhealthy location of Philadelphia. while Federalists blamed disembarking refugees from Haiti (Pernick 1912.562-563). In fact,

Pernick notes, "both sides were right."

11

References

..- Abney,F.Glenn,andLarryB.Hill. l966. "Natural DisastersasaPolitical Variable: TheEffectofaHurricane

on an Urban Election." American Political Science Reyiew 60:974-981.

Achen, Christopher H., and Larry M. Bartels. 2002. "lgnorance and Bliss in Democratic Politics: Party

Competition with Uninformed Voters." Prepared lbr presentation at the Annual Meeting of the Midwest

Political Science Association. Chicago.

ArceneaLtx, Kevin, and Robert M. Stein. No date. "Who is Held Responsible When Things Go Wrong? Policy

Responsibility in Urban Elections." Unpublished paper, Rice University.

Barnhart,JohnD. 1925. "Rainfall andthePopulistPartyinNebraska." AmericanPoliticalscienceReview l9:

527-540.

Bell,Barbara. l9Tl. "TheDarkAgesinAncier-rtHistoryl: TheFirstDarkAgeinEgypt." Americun,Iournol of Archaeology 7 5: l -26.

Berelson, Bernard R.. Paul F. Lazarsfeld, and William N. McPhee. 1954. Voting; A Study of Opinion

Frtrtnation in a Presidential C'crmpazgzz. Chicago: University of Chicago Press.

Blum, John M. 1951. .Ioe Tumulty and the Wilson Era. Boston: Houghton Mif}'lin. Campbell, Angus, Philip E. Converse, Warren E. Miller. and Donald E. Stokes. 1960. The American Voter.

New York: John Wiley & Sons.

Capuzzo, Michael. 2001. Close to Shore. New York: Broadway Books.

Connors, Richard J. 1971. A Cycle of Power; The Career of'Jerselt City Mayor Frank Hague. Metuchen, NJ: Scarecrow Press.

Crosby, Alfred W. 1989. America'.s Forgotten Panclemic; The Influenza o/'t9tB. Cambridge: Carnbridge University Press.

Cuddy, Edward. 1969. "lrish-Americans and the 1916 Election: An Episode in Immigrant Ad.justment."

American Quarterly 2l: 228-243. Fiorina. Morris P. 1981. Retrospective Voting in American National Eleclion,s. New Haven: Yale Universitv

Press.

Hassan,Fekri A. 1994. "NileFloodsandPolitical DisorderinEarlyEgypt." InH.NuzhetDalfesetal.,eds. Third Millenniwn BC Climute C'hange and Old World Cctllapse. Berlin: Spring.

Herlihy, David. 1991 . The Black Death and the Transfbrmation of the West. Cambridge. MA: Harvard University Press.

Irving,.lohn A. 1959. The Social Credit Moveruent in Alberla. Toronto: University of Toronto Press. Iyengar, Shanto. 1991. 1s Any-one Respon.sible'? How Television Frarnes Politicql Issues. Chicago: University

of Chicago Press.

- Kershaw, Ian. 1983. Popular Opinion and Political Dissent in the Thirct Reich. Oxford: Oxfbrd University Press.

12

Key. V. O.. Jr. 1966. The Respon.sible Eleclorttte; RotionaliQ in Presidential Voting 1936-1960. Cambridge, MA: Harvard University Press.

Kolata, Gina. 1999. Flu. The Story of the Great Influenza Pandemic o/'1918 ond the Search.for the Virtts That Ccrused 1r. New York: Farrar. Straus and Giroux.

Kramer, Gerald H. 1971. "Short-term Fluctuations in U.S. Voting Behavior: l896-1964." American Political Science Reviey., 65: 131 - I 43.

Leary.WilliarnM..Jr. 1967. "WoodrowWilson, IrishAmericans,andtheElectionoflgl6." Jottrnalof

American History 51 57-72. Lewis-Beck, Michael S. 1988. Econontics and Electictns; The Major Western Democracie,r. Ann Arbor, MI:

University of Michigan Press.

Link, Arlhur S. 1947. Wilson; The Road tct the White Hou,se. Princeton, NJ: Princeton University Press. Link. Arthr-rr S. 1954. Woodrov'Wilson and the Progres.sive Era 1910-1917. New York: Harper and Brothers. Lipset. Seymour Martin. 1959. Agrorian Socielism. Berkeley, CA: University of California Press. Markus, Gregory. 1988. "The Irnpact of Personal and National E,conomic Conditions on the Presidential Vote:

A Pooled Cross-Sectional Analysis." American Journal o/'Potitical Science 32: l3l-154.

McGraw, Kathleen M. I991. "Managitrg Blame: An Experimental Testof the Effects of Political Accounts." Anterican Political Science Reviev'85: I 133-1151.

McKean. Dayton David. 1940. The Boss. Boston: Houghton Mifflin.

Miller, Raymorrd Curlis. 1925. "The Background of Populisrn in Kansas." Mississippi Valley Hi,ytorical Reyiew l1:469-489.

Monypenny, Williarn F. and George E. Buckle. 1929. The Life of Benlcrmin Di.sraeii, vol. II, rev. ed. London:

John Murray.

Neustadt, Richard E., and Harvey V. Fineberg. 1983. The Epidemic Thcrt Never Was; Policymaking ancl the

Swine Flu Scare. New York: Vintage Books.

Pernick, Marlin S. 1912. "Politics. Parties, and Pestilence: Epidemic Yellow Fever in Philadelphia and the Rise of the First Party System." williaru and Mary Quarterly. Third Series,29:559-586.

Phillips, Kevin. 1969. The Ernerging Republican Majority-. New Rochelle. NY: Arlington House. Pyle, Gerald F. 1986. The Di.ffusion of Influenza; Potterns and Paradigm.s. Lanham, MD: Rowman &

Littlefield.

Reynolds, .lohn F. 1988. Testing Dentocracy; Electorol Behavior and Progressive Re.form in lVew, ,Iersey, 1880-1920. Chapel Hill: University of North Carolina press.

Rusk. Jerrold. 2001. A Statistical Histor.y of the Aruerican Electorate. Washington. DC: CQ press.

Sackett, william E. 1914. Motlern Battles of rrenton,vol. II. New york: Neale. Sophocles. 1991. Sophocles I,2nd ed. Translated and with an introduction by David Greene. Chicago:

University of Chicago Press.

l3

Stone. Deborah A. 1989. "Causal Stories and the Formation of Policy Agendas." Politicctl Science QtLurterll, 101:281 -300.

Tufte, Edward R. 1978. Political Control o/-lhe Econorruy. Princeton, N.l: Princeton University Press.

Wolfers. Justin. 2002. "Are Voters Rational? Evidence from Gubernatorial E,lections." Research Paper No.

1730. Craduate School of Business. Stanford Universitv.

14

Figure 2. Change in Woodrow Wilson's Vote in New Jersey,1912-1916, in Two Beach Resort Counties with Fatal Shark Attacks during July 1916

2

New Jersey

o I

2

Ocean County

Monmouth County o '6a o oo G

oo

Spring Lake township (site of Monmouth Co, shark attack)

10

12

Beach Haven township (site of Ocean Co. shark attack)

l5

Figure 3.

1: Near-beach townships

o

- l

-2

'5

6 ^,

o-

,8

E.9-

Change in Woodrow Wilson's Vote in New Jersey, 1912-'1916, in Ocean County Townships

Beach townships

16