The right of citizens of the United States to vote shall not be denied or abridged by the United States or by any State on account of sex.
Nineteenth Amendment to the Constitution of the United States
The ratification of the 19th amendment in August of 1920 was a pivotal moment, the culmination of a more than 70-year struggle to gain voting rights for women. But what happened after ratification? In order to translate this new right into actual votes by women, local and state governments, political parties, advocacy groups, and individual women needed to learn how to navigate a new legal order. For women, winning the vote gave way to a long-term effort to overcome social norms that discouraged participation and lack of experience with voting itself. Parties and interest groups with a long history reaching out to and mobilizing male voters had to learn what, if any, changes in strategies or appeals would be required to reach women. State and local governments had to add staff, resources, and equipment to accommodate the influx of new voters, in general and in time for the November 1920 presidential election only three months removed from the ratification.
The initial verdict and much of the early scholarship concluded that woman suffrage was a failure, as turnout was low and the addition of women voters failed to shake up the two-party balance of power. While the observers were right in important ways—women’s turnout was indeed lower than men’s—we now know that there were a host of legal and organizational factors that conspired to block or slow the mobilization of women. The incorporation of women as full equals in the electoral process would take decades, and understanding why this is the case helps us better understand the challenges facing efforts to make voting rights a reality for traditionally marginalized groups.
What is the government’s role?
The ratification of said [Nineteenth) amendment placed additional burden upon this department … The time outside registration in wards was extended by direction of the Mayor … 3 evenings before the presidential election, and 2 additional registrars were added to each ward registration place. Ten additional registrars were employed in the central office … [for] many days and evenings for this registration.
Annual Report of the Boston Election Department for the Year 1920
The U.S. Constitution itself is famously silent on the issue of voting rights; the framers did not expect most people to vote and left the practice of elections almost entirely to the states. More than 200 years later, the Supreme Court reaffirmed this in Bush v. Gore (2000): “The individual citizen has no federal constitutional right to vote for electors for the President of the United States.” While modern state and local election departments engage in get-out-the-vote (GOTV) efforts and public relations campaigns in the U.S. today, no state actor (federal, state, or local) is charged with ensuring that each person can and does cast a ballot in every election.
Before the 1920 general election, some state and local governments chose to accommodate new women voters, although they may have (like the head of the Boston Election Department above) complained about the “additional burden.” In Bridgeport, Connecticut, for example, the “veteran registrar of voters” claimed women’s suffrage required “a great deal of extra work, but can be done nevertheless.” In anticipation, Bridgeport ordered fourteen new voting machines, enough to handle a doubling of the electorate, although the registrar doubted “it will go that high.”
After ratification, the state of Connecticut called a special legislative session to “provide the state with sufficient legal basis for receiving women as voters,” including increasing the number of registrars. Connecticut and other states also automatically rolled women registered for school elections (permitted by a number of states prior to 1920) over to general election lists. In another example of accommodation, the Quincy (MA) City Council voted in special session to hold primaries in the smaller precincts, rather than wards, to accommodate the expected influx of women voters in 1920.
Not all states chose to be so accommodating of new women voters. In the most extreme examples—Arkansas, Georgia, Mississippi, and South Carolina—women were prohibited from voting in the 1920 presidential election because ratification in August took place after deadlines to register or pay poll taxes. To be clear, this was a choice; other states with similar provisions found a way to legislate accommodations for new women voters, such as temporary legal changes and registration days for women only.
More generally, state election laws contained a range of restrictions and requirements at the time of women’s enfranchisement, and permitted considerable bureaucratic discretion in the enforcement of those laws. Black women in the South were particular targets. In Richmond, Virginia, for example, black and white women seeking to register to vote in advance of the 1920 presidential election overwhelmed registration offices. The city responded by appointing three additional deputies for white women, but multiple requests to make similar accommodations for black women were ignored. The result was a long line of black women outside registration offices, due to both the small number of registrars and the more frequent challenges to the black woman vote. Even states that accommodated women often had restrictive election laws that created barriers for women. Despite extending registration opportunities for women, both Connecticut and Massachusetts required a literacy test. Massachusetts added a poll tax, while Connecticut piled on a morals clause and a long residency requirement.
If the state does not mobilize, who does?
Given states’ lack of responsibility to get voters to the polls, the onus for voter mobilization falls to individuals themselves, as well as to political parties and civic, labor, religious, and other organizations. Not surprisingly, suffrage advocates were eager to meet this need. Suffragists reached out to the mayor of Bridgeport the day the Nineteenth Amendment was ratified. Their offer of assistance “was gladly accepted” and space in City Hall offered to them. The leading national suffrage organization, the National American Woman Suffrage Association (NAWSA), voted in 1919 to dissolve their organization when the fight was won, and to create a new organization, the League of Women Voters, to continue their work on behalf of a political voice for women. The LWV chapters engaged in a range of GOTV activities, including citizenship schools, regular radio broadcasts, and practical demonstrations of election machinery at fairs and other locations in the years following suffrage: The St. Paul Dispatch reported “Women Learn How to Vote at Fair” (September 6, 1920).
Anticipating suffrage success, both major parties created national women’s committees to mobilize women voters. For the most part, actual GOTV work was the province of state and local party organizations. Women’s clubs established by both candidates and parties held teas, presented speakers, and made the case for the appeal of the candidate on the issues that they believed mattered most to female voters. In Massachusetts, for example, “Women Taught How to Run An Election: Republicans of Fair Sex Open Headquarters,” and, in Minnesota, “Coolidge Club Plans to Hold Precinct Teas; Women Voters to Hear Speakers at a Series of Meetings.”
While these efforts were substantial and often effective, leaving voter mobilization to non-state actors has limits and bias. A state-led mobilization effort would ostensibly be required to mobilize all women, regardless of preferences or positions. Efforts led by parties and interested groups sought to mobilize only those women most likely to support the party, interest, or group.
What happened?
While states now maintain records of who votes in which elections, in the 1920s this type of information was either not recorded, not preserved, or did not include the sex of the voter. Today we also track who votes through exit polls and public opinion surveys, but those tools either did not exist or were not reliable in the 1920s. We figured out what proportion of women and men voted using a statistical tool known as ecological inference. Ecological inference lets us combine information from the U.S. census about the population (in this case, the number of voting-age women and men) with information from the voting record (such as the number of votes cast for each party) to estimate the percentage of women and men who voted and for which parties in a sample of 10 states. (If you’re interested, our 2016 book, Counting Women’s Ballots, goes into details about the data, methods, and detailed findings).
Our estimates of women’s turnout immediately after suffrage suggest that women’s turnout initially lagged men’s by a wide margin, as Figure 1 shows. While the gap slowly closed, even by 1936, nearly 20 years after ratification, women’s turnout was still about 20 percentage points behind men. Dominant narratives after suffrage, in popular and academic conversation, put the blame for women’s lower turnout squarely on the shoulders of women themselves. As early as 1924, headlines from Good Housekeeping (Is Woman’s Suffrage a Failure?) to Harper’s (Are Women a Failure in Politics?) and beyond declared women’s suffrage a failure because women chose not to vote (and when they did, voted like their husbands). Lamenting women’s failure, as women, to embrace their role as voters remained a theme for decades. A long story on women voters in the Los Angeles Times in June 1960 (“Femme Bloc Could Run U.S.”) concludes:
“Today women of voting age outnumber men. Yet, offsetting this to some extent has been the fact that women have generally stayed in the kitchen in unhealthy numbers on Election Day.”
Figure 1. Women’s turnout in presidential elections lags men’s by a considerable margin after suffrage (ecological inference estimates, 1920-36)
The effects of disenfranchisement were long-lasting, but not permanent. While it took decades for the gap between men and women to fully close, today women are more likely to vote than men. Data collected as part of a long-running survey of American voters reveals how the gap slowly closed after 1960 (Figure 2).
Figure 2. Today, women are as or more likely to vote than men (American National Election Studies, 1948-2016)
Where did women vote and where did women stay home?
If women are not inherently or always less likely to turn out than men, why do we observe such a large gender gap in the first few elections after suffrage was extended? Whether women voted appears to be more a function of where they lived, and the electoral institutions within which they could exercise their right to vote (i.e., the legal structure of American voting rights), and less about gender alone. What national averages obscure is that women’s turnout varied considerably from state to state in the first presidential elections after suffrage. Figure 3 reports our estimates of 1920 turnout in ten US states.
Figure 3. Women’s turnout highest in states with few ballot access restrictions and competitive elections (ecological inference estimates, 1920)
Why did fewer than 5% of women turn out in Virginia compared to more than 50% in Kentucky? States with more voting restrictions (especially Virginia, Massachusetts, and Connecticut) had considerably lower turnout overall, but especially among women. In states where political competition was high (Missouri and Kentucky), parties and other organizations had the incentive (and presumably the resources) to engage in more extensive voter mobilization. For this and other reasons (salience, excitement, stakes), turnout was higher in those states, and again, particularly among women. Because states do not have a responsibility to ensure that citizens vote, mobilization is a function of individual capacity to negotiate regulations and barriers as well as the incentives for non-state institutions to mobilize voters when they perceive the need and then among those from whom they expect support. A consequence is that women living in some states were far more likely to convert their right to ballots than were women in other states.
Women of color faced additional challenges
Well after the formal incorporation of women in the 1920s, the presence of state (and other) barriers to voting created particular burdens for women of color. State restrictions on voting in the South were designed explicitly to block black voters from the polls. When state action fails to affirmatively support a right to vote, parties and citizen organizations often step into the breach. In the South, however, the dominant Democratic political party viewed its interests as best served by engaging in extensive and violent efforts to block or slow black voter mobilization. The leading African-American newspaper of the twentieth century, The Chicago Defender, reports that those who sought to register and educate black women voters in 1920 alone experienced murder, kidnap and lynching, threats of arson against homes and businesses, and in one town, 500 warrants against black women charged with “registering illegally.”
Civil rights activism and particularly the passage of the Voting Rights Act (1965) narrowed the gap between white and black electoral participation in the 1960s. Self-reported turnout of whites and minorities is very similar today, but entirely closing the gap took decades (see Figure 4). Unlike the Nineteenth Amendment, the Voting Rights Act put the onus on state governments to meet certain standards for electoral participation, with mechanisms for evaluation and enforcement. Yet even in the Voting Rights Act, provisions largely focused on stopping state action that created barriers to voting, rather than requiring states to take affirmative steps to facilitate voter registration or turnout.
Figure 4. Minority turnout lags white considerably before 1964, but racial differences in turnout are very small by 2016 (ANES, 1948-2016)
Mobilizing traditionally-excluded groups requires resources and time
That voter mobilization in the U.S. is largely left to parties and advocacy groups means that gender differences in organizational capacity had consequences for the representation of women’s interests after suffrage. The newly-created League of Women Voters had virtually no experience with voter mobilization and their energies were divided between GOTV efforts and the work of studying, recommending, and advocating for policy proposals. In comparison, political parties, labor unions, and other male-dominated organizations already had extensive experience with voter mobilization. While men were mobilized by parties and by organizations representing their interests, women were sometimes mobilized by neither. This is particularly a problem in the states that made voting more difficult. When political, labor, and civic organizations and activists did devote energy to challenge state policies that discourage turnout and mobilizing their members in elections, this necessarily meant (and means!) fewer resources are available to understand and advocate for the interests of their constituencies in other ways. This dynamic remains important today as groups or individuals respond to efforts by some states to ramp up ID requirements or resist efforts to permit mail-in or early in-person voting.
Newly-enfranchised and marginalized groups face a long-term struggle to convert voting rights into political equality in the absence of an affirmative right to vote and in the presence of additional burdens on potential voters—the poor were confronted with the poll tax, African-Americans in the South were met with state-sanctioned violence, and immigrants and people of color were required to pass literacy tests. All of these burdens hampered women in particular. Families faced with a poll tax tended to prioritize male voting, physical risks dissuaded the most vulnerable, and immigrant women had fewer opportunities to learn English. Without the state (and often despite resistance from the state), newly-enfranchised groups must do the work—individually and collectively—of mobilizing resources to translate the “right” to vote into an actual ballot.
Voting may be an individual responsibility, but the actions of parties and other organizations make a big difference in getting Americans to the polls. And that mobilization is most likely to happen in places where party competition is close—precisely the kinds of contests that are again increasingly scarce in U.S. politics. Further, laws that make it harder to vote weigh most heavily on those people who are already marginalized. Rather than asking if potential voters—women in 1920 or other groups today—are failing to fulfill their civic responsibilities, we might do better to ask if the political system is failing its citizens.