Why don't we have third parties anymore? The history of electoral fusion

I was looking into the history of the end of electoral fusion in Kentucky, and it seems that the Ohio River Valley was particularly significant in the end of electoral fusion and, consequently, the decline of third parties in American politics. Here are selections from a couple of monographs; RTWT. The "Australian ballot" is the secret ballot, printed at public expense. Previously, voters would use ballots printed by individual political parties.

Duverger's Law, Fusion, and the Decline of American "Third" Parties, by Howard A. Scarrow (The Western Political Quarterly, Vol. 39, No. 4 (Dec., 1986), pp. 634-647):

During the latter half of the 19th century fusion candidacies were frequent at all levels of government. At the presidential level the first example is found in the election of 1856, when the American (Know-Nothing) party and the Whig party both backed the third party candidacy of Millard Fillmore. ...The last conspicuous example of a presidential fusion candidacy was that of William Jennings Bryan, who was the candidate of both the Democratic party and the Populist party in 1896.

...The introduction of the Australian ballot in the last decade of the century sounded the initial death knell for fusion candidates. The introduction of this ballot reform presented minor parties with the obstable of now having to gain access to the ballot by means of petition signatures, a burden not placed on parties which could gain automatic ballot access by virtue of having attracted a large number of votes in a previous election—that is, the two major parties. The decline of third parties in the United States has thus often been traced in part to this aspect of ballot reform. There was, however, another feature of the new ballot laws which has received insufficient attention... the new ballot laws presented a way for states to outlaw fusion candidacies. When the party-column type of ballot was authorized, the law could require that a candidate's name appear at only one place on the ballot. Where the office-block ballot format was introduced, the legislation could specify that a candidate have only one party label attached to his name on the ballot. By 1895 six states had enacted one of the other of these restrictions. By 1900 the list had grown to thirteen; by 1910 to twenty.

...From their very inception, anti-fusion laws were challenged in the courts by aggrieved candidates. Almost invariably, however, state courts refused to support the argument that the laws violated the respective state constitutions.

...Dissenting opinions, however, used equally righteous language to defend the fusion practice. Fusion candidacies, it was argued, were often the only way of defeating an entrenched and corrupt political machine. They also provided a healthy antidote to narrow partisanship; and they increased voter interest. In only one state, New York, did these latter opinions prevail. In 1911 that state's highest court struck down an anti-fusion statue which had been pushed through the legislature by Tammany forces stung by successful fusion campaigns in New York City.

It should be noted that electoral fusion was used in the 1933 election of Republican Fiorella LaGuardia and the eventual demise of Tammany Hall.

"A Place on the Ballot": Fusion Politics and Antifusion Laws, by Peter H. Arsinger (The American Historical Review, 85:2, April 1980, pp. 287-306):

...at least one little-known development in the electoral reform of the 1890s involved a conscious effort to shape the political arena by disrupting opposition parties, revising traditional campaign and voting practices, and ensuring Republican hegemony--all under the cover of procedural reform. This development was the adoption of the so-called antifusion laws, which also altered the political behavior characteristic of the Gilded Age, with varying effects on the role of third parties, modes of political participation, and the electoral process itself.

Fusion, or the electoral support of a single set of candidates by two or more parties, constituted a significant feature of late nineteenth-century politics, particularly in the Midwest and West, where full or partial fusion occurred in nearly every election. ...if fusion sometimes helped destroy individual third parties, it helped maintain a significant third party tradition by guaranteeing that dissenters' votes could be more than symbolic protest, that their leaders could gain office, and that their demands might be heard.

...Nationally, the Republican success in 1894 led to the passage of antifusion laws by other states in 1895. ...Michigan Republicans, now in complete control of the legislature, reintroduced their anti-fusion bill of the previous session and pushed it into law. Although some judges described it as "unconstitutional" and "revolutionary," the state upheld the measure in the same partisan spirit in which it had been enacted--four Republican judges in the affirmative, one Democrat in dissent. The Ohio legislature, meeting in 1896, concluded this first legislative flurry with the so-called Dana law, an elaborate measure based upon the customary antifusion ballot requirement. In Ohio, the local focus of antifusion legislation seemed particularly evident, at least initially. In the recent Cincinnati mayoral election, the Republican machine of "Boss" George Cox and Joseph B. Foraker had been challenged by a fusion coalition of Populists, Socialists, laborites, and dissident Republicans that had nearly received the Democratic endorsement as well. Some legislative observers regarded the subsequent Dana bill, prepared by a Foraker Republican, as primarily designed to prevent just such unified popular revolts against machine rule in municipal elections.

...The Michigan and Ohio experiences were not lost on Republicans in Indiana, the state in which Mark Hanna feared fusion most. Hoosier Republicans made two unsuccessful efforts during the campaign to secure the effects of antifusion legislation without actually receiving such a law. ...The lessons learned in, and the opportunity presented by, the sweeping 1896 Republican victory led Republican-dominated legislatures in many more states to enact fusion laws quickly. Republican legislatures passed antifusion laws in 1897 in Illinois, Iowa, North Dakota, Pennsylvania, Wisconsin, and Wyoming as well as in Indiana. As Republicans gained sufficient legislative control elsewhere, the law spread further: California and Nebraska in 1899; Kansas, Minnesota, and South Dakota in 1901; Idaho in 1903; and Montana in 1907.

Ending the effective cooperation of Democrats and third party groups was both the primary goal and the major result of these efforts. ...By preventing effective fusion, antifusion laws also brought an end to another major characteristic of late nineteenth-century politics--the importance and even existence of significant third parties.

See also footnote 44:

In addition, the Democratic legislatures of three Southern states also enacted antifusion legislation in the early 1900s, and controversy over the law actually provoked a riot in the Kentucky legislature. Thus, while the focus here has been on the Northern Republicans, the law was obviously regarded as serving the interests of the dominant party wherever it was enacted.