Historical background

The roots of the woman suffrage movement in the United States can be traced to the

Seneca Falls Convention in 1848. In Seneca Falls, New York, a group of men and

women met to discuss injustices against women and the elimination of gender

inequality. The topics included property rights, educational opportunities, wages, and

woman suffrage. From Seneca Falls the woman suffrage movement spanned 72 years,

culminating in the ratification of the Nineteenth Amendment in 1920.

In addition to being the birthplace of the woman suffrage movement in the U. S.,

the state of New York was also headquarters for the National Woman Suffrage

Association (NWSA). The NWSA was organized in 1869, and was founded by

Elizabeth Cady Stanton and Susan B. Anthony. It was organized to secure voting rights

for women by focusing on changing the federal constitution. During the same year that

the NWSA was established, Lucy Stone and her husband Henry Blackwell established

the American Woman Suffrage Association (AWSA) which was headquartered in

Boston. The AWSA differed ideologically from the NWSA. For instance, The AWSA

supported the right of blacks to vote while at the same time working for the vote for

women (Hannam et al. 2000: 12). Another point of difference was that the AWSA

focused primarily on securing woman suffrage at the state level. Other than voting

rights for blacks and focusing on state-level campaigns rather than a federal amendment,

issues separating

these organizations included the NWSA s acceptance of free love and their disapproval of

androcentrism within the religious establishment (Banaszak 1996). As time went on,

however, the NWSA became more ideologically moderate which set the stage for the two

organizations to merge.

In 1890, the AWSA and NWSA merged to form the National American Woman

Suffrage Association (NAWSA) (Hannam et al 2000). This further centralized the

organizational structure of the suffrage movement in the eastern U. S. (Giele 1995).

Since the national suffrage organization was located in the East, most of the suffrage

leaders came from the East and the majority of the local chapters were located in the

East before later spreading to the Midwest (Banaszak 1996). Therefore, many of the

suffragists early efforts were concentrated in the East. Suffrage activity took place in the

West but it was not as well organized due to fewer financial resources, fewer members

of suffrage organizations, fewer suffrage leaders, as well as geographic isolation from

other local suffrage organizations (Flexner 1970).

Despite the leadership and the organizational activity in the eastern U. S., woman

suffrage was first adopted in western states and territories. Table 1 lists the states and

territories that granted full woman suffrage before the ratification of the Nineteenth

Amendment. Wyoming Territory passed woman suffrage as early as 1869 despite a

small number of woman suffragists in the state (Larson 1965). Wyoming was also the

first state to give women the vote when it achieved statehood in 1890. One year after

Wyoming Territory passed woman suffrage, Utah Territory followed; though woman

suffrage was later revoked by the Edmunds-Tucker Act which was passed by Congress

in 1887 (Beeton 1986). In 1893, Colorado became the second state to grant woman

suffrage, followed by Idaho and Utah in 1896. Although Illinois and Rhode Island

granted presidential suffrage in 1913 and 1917, respectively, New York was the first

state east of the Mississippi to grant full woman suffrage in 1917. In 1918, Michigan

became the second eastern state to grant full woman suffrage. New York and Michigan

were the only eastern states to pass full woman suffrage before the Nineteenth

Amendment.

Woman suffrage research

Few comparative state-level studies have addressed woman suffrage and the political

process, especially in the West. Most studies are limited to individual state campaigns

or to the national movement. T. A. Larson (1965, 1970, 1972, 1976) and Beverly

Beeton (1986) have done the most extensive work on woman suffrage in the West.

Their research examines social, economic, and political factors that influenced the

adoption of woman suffrage. However, their research is limited mainly to the

Intermountain West. Additionally, they do little comparative analysis of state-level

government characteristics.

Research examining the national suffrage movement (Stanton et al. 1881;

Flexner 1970; Giele 1995; Banaszak 1996) focuses on the movement s organization and

leadership in the eastern states. It underestimates the significance of suffrage success in

the western states and their contribution to the national woman suffrage movement.

Although sociologists are beginning to explore the significance of the woman suffrage

movement in the West (McCammon and Campbell 2001; Schiffman 1998), no study

systematically examines characteristics of state governments that may have impeded

success or alternatively produced a conducive political environment.

McCammon and Campbell s (2001) study of woman suffrage in the West

attributes success to the West s unique culture which increased opportunities for women.

They maintain that sentiments of politicians concerning the appropriate roles of women

in society shifted as gender norms changed, increasing support for woman suffrage. In

another study, McCammon et al. (2001) identify gendered opportunities that changed

expectations concerning voting rights. Two of the variables they identify as gendered

opportunities were significant in their model: the rise of the new woman and World War

I. The rise of the new woman resulted from increased education, working outside the

home, professional careers, divorce, involvement in the public sphere, and fewer

children. Women s contributions in World War I working in factories and farms, fund

raising for oversea hospitals, and organizing student nurses changed gender expectations

since women filled roles previously filled by men. These changes also opened doors in

the political arena by challenging traditional gender norms that were previously limited

to men.

Woman suffrage scholars also attribute woman suffrage success in the West to

the unique political culture. The Western frontier closed in 1912 and until that time

territorial governments provided a fertile testing ground for legislative innovation.

Beeton (1986) argues that territories could experiment with woman suffrage while not

threatening the political stability of the established states or the national political scene

since territorial voters could not vote for their own governors or for the President (xi).

Territorial governors were appointed by the federal government and territorial

governments were largely controlled by Congress. Territories were viewed as

nonthreatening, temporary, and alterable. As a result, territorial governments had fewer

procedural constraints and were more easily altered than states. For example, amending

territorial constitutions did not require a popular referendum. In addition, if any political

problems occurred, Congress could simply modify the territorial government (Beeton

1986; Dye 1985).

Clemens (1997) extends Beeton s argument that territories were a testing ground

for new legislation to the newer western states in the nineteenth century. Clemens calls

newly formed western states an experiment station because they provided opportunities

for political innovation. New states were less constrained by traditional party systems

and were more liberated from traditional elites. Consequently, institutional innovation

was more common. Popular groups were better able to initiate and institute alternatives

in the polity. This is manifest by the legislative reforms adopted by western states.

Clemens demonstrates that states in the West were more innovative by showing that they

were more likely to adopt reforms like the Australian ballot, the mandatory direct

primary, the initiative and referendum, and woman suffrage. These reforms opened the

door to political innovation and loosened the grip of traditional parties on the political

status quo. Consequently, some suggest that third party challenges encouraged woman

suffrage success. However, McCammon et al. (2001) find that third party challenges

(measure as the percentage of seats in the state legislature held by third parties) did not

significantly increase the likelihood of suffrage success.

A survey of gubernatorial elections reveals that one particular political party did

not dominate states in which woman suffrage was adopted. Table 2 compares the states

that granted full suffrage before the Nineteenth Amendment and lists which party s

gubernatorial candidate received the most votes in the election prior to the year or

during the same year that suffrage was adopted. Seven of the fifteen states elected a

Republican governor, five of the states elected Democratic governors, and three elected

governors from other parties. Since one party did not dominate the elections in these

states, partisan support by one particular party did not seem to be a determining factor.

Table 2 demonstrates that what mattered most was not which party supported woman

suffrage but, more likely, that it was a platform of a political party s campaign.

Other reasons scholars give for early suffrage success in the West are listed by

Schiffman (1998). These include smaller populations, smaller legislatures to convince,

greater equality and democracy. And, because of the scarcity of women in the West,

states passed woman suffrage to encourage the immigration of women (see also Beeton

1986; Flexner 1970; Larson 1970). Schiffman (1998) asserts, however, that while some

of these factors may have contributed to suffrage in the West, these claims have not been

empirically tested (6). While several studies of woman suffrage mention suffrage

activity in the West, they do not underscore its significance and they are biased toward

eastern states due to the lack of available data and due to the lack of organization in the

West (Banaszak 1996; Stanton et al. 1881).

Another problem with this literature is that is does not address specific and

systematic political processes of state government. State suffrage campaigns were not

exclusively dependent on cultural changes. State building was also a factor. In his

monograph on the modification of practices concerning the ordaining of women in

religious denominations, Chaves (1997) states that changing practices were not only a

result of cultural changes, but also internal organizational processes. Chaves (1996)

maintains that if changes in cultural norms were all that mattered, denominational

change would have been relatively uniform. However, denominations did not uniformly

change their practices. Not only was there variation but some denominations did not

ordain women ministers at all. Therefore, Chaves contends that internal organizational

dynamics must also be studied to understand the variation in the timing of organizational

change. Similarly, I argue that cultural change does not tell the entire story of woman

suffrage success. Internal organizational dynamics of state government also played an

important role in the adoption of woman suffrage.

Social movement theory

Sociological research on the woman suffrage movement is often influenced by theories

of social movements. Resource mobilization, political opportunity structure, and

cultural framing processes have animated studies of woman suffrage (Banaszak 1996;

Giele 1995; King 2001; McCammon et al. 2001; Schiffman 1998). Unfortunately, these

theories are insufficient for explaining state-level legislative outcomes.

Most of the research on social movements from the resource mobilization

perspective emphasizes the rational organization of social movements. According to

resource mobilization theorists, social movements result from rational calculated

behavior by its members and leaders. Their goals and strategies are clear. Social

movements form because of long-term changes in resources, organizations, and

opportunities (Jenkins 1995). For example, the civil rights movement emerged in the

1950s due to changes in resources among blacks. Changes taking place during this time

include blacks migrating from the poverty-stricken South, black churches in the urban

south becoming more affluent, expanding black middle- and working-classes, and

growing college enrollments among blacks. All of these changes provided resources for

blacks and increased their ability to mobilize (McAdam 1982).

Aldon Morris (1995) studies the black sit-in movement in the South during the

early 1960s and concludes that its success came by drawing on existing resources

provided by the social networks of black churches and colleges. Many of the sit-in

movement leaders were school teachers and church leaders. Consequently, leaders had a

pool of participants from which to draw. Morris says, The spread of the sit-ins followed

the networks of the preexisting institutional networks (363). By accessing existing

social networks black sit-in leaders increased the ability of the black sit-in movement to

organize and mobilize members.

Other studies confirm the impact of already existing social networks on social

movements and they also emphasize the importance of accessing already existing

organizations (McAdam 1988; Gould 1995). Giele (1995) argues that the woman

suffrage movement developed with the help of already existing organizations. She

identifies many organizations in which women participated during the first half of the

nineteenth century before the principal suffrage movement; namely missionary and

moral reform societies. From these arose local temperance groups and eventually the

Woman s Christian Temperance Union (WCTU). Giele contends that the temperance

movement and the WCTU were crucial to the success of the woman suffrage movement.

The WCTU constituted a strong social movement organization that reached into local

communities and linked individual women to the larger feminist movement. . . .(It)

developed a reform strategy linking temperance and woman s rights that mobilized

immense support nationally as well as locally (64).

This literature explains the importance of resources, especially organizational

resources and social networks, to the vitality of movement emergence and to eventual

success. However, resource mobilization is a not sufficient condition for social

movement success. This is evidenced by the woman suffrage movement since women

were enfranchised in the Intermountain West before they were enfranchised in the East.

In his analysis of the woman suffrage movement in the western U. S., King (2001)

shows that 14 out of 22 states introduced their first suffrage bill in the state or territorial

legislature before their first suffrage organization existed. Contrary theories of resource

mobilization theorists, Table 3 reveals that woman suffrage activity often took place in

states before movement leaders had mobilized organizational resources.

Another concept used by social movement theorists emphasizes political process

and identifies elements of the political opportunity structure that influence social

movement outcomes. Skocpol and Campbell (1995) assert that after a social movement

emerges, political opportunity structure becomes important. Political opportunity

structure is used to understand the political environment of a social movement. The

common feature of these studies is an emphasis on the legal and administrative rules of

the political system. Scholars use this approach to understand the timing and fate of

movements in the context of opportunities afforded by the political structure.

Most studies of political opportunity structure fall into one of two categories. The

first seeks to explain movement emergence in relation to the changing political structure.

Changes in the political structure create opportunities for social movements to emerge

and to increase their legitimacy and support. The second category seeks to explain the

success of cross-national, comparable social movements (McAdam et al. 1996). These

studies usually focus on the different characteristics of international governments and

their impact on the social movement within their country (Kitschelt 1995; Kriesi et al.

1995).

One problem with the concept of political opportunity structure is that social

movement scholars frequently disagree on its definition. It is often defined too broadly

and it is sometimes used as a catch-all. It has been applied to the political structure as

well as almost any feature of a social movement s environment including cultural

changes. McAdam (1996) notes that although political and cultural aspects of the

environment can be closely related, changes in the political structure should not be

mistaken for cultural processes through which political elements are frequently framed.

In addition to issues of conceptual clarity, another problem exists. Political

opportunity structure has been confused with resource mobilization. Theorists have

maintained that political opportunities were resources to be used by social movements.

However, Jasper (1997) criticizes political process theorists who use the same label for

both political opportunity structure and resource mobilization because these broad

definitions deprive each perspective of their clarity and explanatory value.

Beyond social movement theory: institutional politics

To understand why some states adopted woman suffrage while others did not, I begin by

drawing from institutional politics theory. Institutional politics theory integrates

concepts from institutional and political theories of social policy. Amenta and Halfmann

(2000) use institutional politics theory to comprehend differences in social spending in

states at the end of the New Deal (see also Amenta 1998; Amenta and Poulsen 1996). I

apply aspects of institutional politics theory that relate directly to state law-making:

democratization, reform-oriented regimes, and the social movement s political presence.

Democratization

Amenta and Halfmann (2000) describe a democratized polity as one with high levels of

political participation. Political participation is characterized by greater voting rights.

Amenta and Poulsen (1996) find that spending on Old-Age Assistance pensions is

influenced by increased voting rights. I expect that democratization also increases the

success of woman suffrage legislation. For my study I characterize democratization as

elements of political openness identified by Clemens (1997). Clemens demonstrates

that these elements the initiative and referendum, the Australian ballot or the secret

ballot, and the mandatory direct primary (Clemens also includes woman suffrage as a

measure of political openness) increased political participation by weakening party

control, thereby permitting greater participation among the electorate. Clemens also

shows that these political innovations were widespread throughout the Western states.

I combine these elements as an index of democratization because they directly

impacted political participation by weakening traditional party ties and party control in

state law-making. The Australian ballot weakened party control by replacing party

ballots. The Australian ballot listed political candidates by the office for which they

were running, rather than listing candidates strictly by political party. Furthermore, the

Australian ballot was distributed by the local government. On the other hand, party

ballots listed all candidates from one party and were distributed by the party directly to

the voters. After receiving the ballots from the party, voters took the ballots to the polls

to cast their votes. The practice of using party ballots was often poorly regulated and

corrupt. Party ballots were intended to expedite straight party voting over selecting

particular candidates from the other party (Clemens 1997).

States use of mandatory direct primaries also increased voter participation. Direct

primaries allowed voters, not parties, to select political candidates in party conventions.

Prior to the implementation of direct primaries, party leadership selected candidates thus

limiting the electorates choices for political office (Merriam and Overacher 1928).

The final element of political openness is the initiative and referendum. The

initiative and referendum increased democratization by allowing voters to amend the

constitution by completely bypassing the state legislature. This is carried out when a

percentage of voters petitions the government. The amount of signatures required is

specified by each state s constitution. When the requisite number of signatures is

secured the amendment is submitted to the electorate at the next general election

(Ranney 1978). In the end, two states granted woman suffrage via the initiative and

referendum: Arizona and Oregon, both in 1912. Since these elements gave voters more

choices and more chances to participate in the political process, I expect democratization

to increase the opportunities for women to participate in the political process. I expect

that states with a greater degree of democratization will be more successful passing

woman suffrage legislation.

Reform-oriented regimes

Institutional politics theory argues that reform-oriented regimes are more likely to

support social spending. Third parties are characteristic of reform-oriented regimes and

they generally supported more progressive social policy. Their presence increased social

spending at the time of the New Deal (Amenta and Halfmann 2000).

Third parties were also generally more supportive of woman suffrage. Following

the lead of previous third parties, the Progressive Party was especially supportive of

woman suffrage and included it on their political platform in 1912, their first year in

existence. This resolution was an attempt to garner support for their presidential

candidate, Theodore Roosevelt, in the western states where women were already

enfranchised (Wolbrect 2000). The Progressive party s platform spurred woman

suffrage debates in other states as well. Edwards (1997) contends that the appearance of

the new Progressive Party played an influential role in a number of states. The

momentum of state-level victories resurrected old debates within the suffrage movement

(165). Thus, I expect the presence of third parties to increase the activity and success of

woman suffrage legislation.

Married women s property rights as another indicator of a reform-oriented

regime. Before the passage of married women s property rights, common law transferred

control of a woman s property to her husband upon marriage. With the passage of

separate property rights women gained (to varying degrees depending on the state) the

right to own the property they brought to the marriage as well as the property and

earnings they acquired during their marriage (Conway et al. 1999). While wealthy

families often created special contracts or trusts to protect their property from a

prospective husband, common law was especially hard on middle- and lower-class

women who did not have the legal means to protect their property (Matsuda 1985). Kahn

(1996) also finds distinct regional differences between property rights acts passed by

states. She finds that despite the early success of many southern states which passed

property rights, they interpreted the statutes conservatively.

Beginning in 1838, states began to pass legislation to protect women s property

from their husbands creditors (Zeigler 1996). The intent of these laws was to secure the

property of a married woman from her husband s creditors to protect family assets

during economic problems, especially the economic downturn of the late 1830s (Kahn

1996). Although these changes marked the beginning of legislation to protect married

women s property rights, the statutes were often narrowly defined and ambiguous,

leaving the courts freedom for interpretation (Chused 1985). Methods used by judges to

limit the rights of wives included refusal to apply new statutes to property rights

acquired before the passage of the statute, limiting contractual powers of women,

resolving conflicts in the rights of husband and wife in favor of the husband, and

maintaining that property acts did not include property purchased with the earnings of

married women. Married women s property rights acts often resulted in continued

inequality. As a result, I focus on feme sole or sole trader rights. Sole trader rights

granted married women the right to engage in business on the same basis as single

women. Sole trader rights include the right to make contractual agreements and file law

suits (Kahn 1996). In my view, states that granted sole trader rights is an indication that

a state legislature is more sympathetic to progressive feminist legislation. Granting sole

trader rights demonstrates a propensity to reform prevailing political measures. I expect

that states passing sole trader rights will be more likely to pass woman suffrage

legislation.

Political presence

The third aspect of institutional politics theory I draw from is political presence.

Amenta (1998) argues that social policy advocates gain support by establishing a

presence in the political arena. When this occurs, states are more inclined to increase

social spending. Other scholars also emphasize the importance of support for a social

movement in the political arena (see McAdam 1996; Tarrow 1998). In my study I

consider whether or not woman suffrage was endorsed by either the Democratic or

Republican parties. I expect support by either party to be more important to legislative

success than which particular political party supported suffrage.

Beyond institutional politics: organizational characteristics

In institutional politics theory the political structure and the party system are crucial for

explaining social policy outcomes. However, institutional politics theory does not

adequately identify structural constraints or internal organizational characteristics that

restrain legislative success. For that reason, this study goes one step further than past

studies of woman suffrage by analyzing the impact of organizational characteristics of

state government on legislative success. Consistent with Chaves (1997) findings that

internal organizational dynamics were crucial to the timing of changes in the Catholic

church, analyzing organizational characteristics of state government should reveal

specific structural characteristics that influenced early suffrage adoption in the

Intermountain West.

Gorden Bakken (1987) discusses the distinct structural characteristics of state

governments in the Intermountain West. According to Bakken, constitutional

convention delegates who had the task of creating constitutions for Rocky Mountain

states were concerned with the scope of power of the new state legislatures. The

constitutions and legislative procedures reflected their sentiments. For example,

Wyoming addressed the issue of limiting legislative power in their debates regarding

legislative apportionment. Liberals argued that apportionment should be based on

population to give legislative power to the people and not to the economic interests.

Conversely, cattle ranchers and others from counties dominated by cattle interests,

argued that representatives should be apportioned according to county lines. This debate

illustrates one way that the newly formed states in the West sought to break with past

political conventions and create a stronger tie between popular sentiment and legislation.

As a result, suffragists faced fewer institutional constraints in newly formed states in the

West.

Another organizational characteristic, state founding date, is also related to

woman suffrage success. States achieving statehood in earlier years were less likely to

pass woman suffrage. Table 4 compares states that achieved statehood from 1776 to

1925 and whether they granted full suffrage to women. Overall, the proportion of

suffrage success increased in states founded in later time periods. In the first time

period, only one state, New York, eventually passed woman suffrage. None of the states

founded during the next time period granted woman suffrage. The proportion of states

that passed suffrage increased during the remaining time periods through 1876-1901.

During the final time period, New Mexico was the only state that did not pass full

suffrage. However, New Mexico did pass partial suffrage the year it achieved statehood

in 1912.

One reason New Mexico s did not grant full suffrage to women was that its

requirements for constitution revision were especially difficult. New Mexico s 1912

constitution specified that woman suffrage could only be granted by three-fourths

majority vote in each house of the legislature. The proposal needed to carry a

threefourths majority at a state-held referendum with a two thirds majority of those

voting in each county. These provisions were more stringent for woman suffrage than

for other constitutional amendments in New Mexico. These requirements were designed

to prevent Spanish speakers, who had traditionally held positions of importance in the

government, from losing political power.

Imprinting

Organizational imprinting describes the impact of environmental conditions during

particular time periods on newly formed organizations. Specifically, imprinting is the

idea that conditions at the time of founding are reflected in an organization s structure.

According to Stinchcome (1965) organizations founded during the same time period are

typically similar in form. An organization s environment at one point in time provides

certain resources and political constraints which imprint organizations and these

conditions have a lasting effect on organizations ( Hannan and Freeman 1989; Meyer

and Rowan 1977; Scott 1992).

Organizational imprinting provides a framework for understanding the timing of

woman suffrage adoption in states founded during different time periods. Imprinting

effects are illustrated in Table 4. Governments founded in similar time periods exhibit

similar structural patterns concerning the adoption of woman suffrage. The effects of

imprinting are exhibited by the differences in the timing of woman suffrage adoption by

state governments.

Which conditions were imprinted in the newly formed western states that made

them receptive to changes concerning citizenship? Before answering this question it is

necessary to describe the environmental processes that constrain organizations generally,

and state governments specifically. Scott (1992) delineates two types of environments:

technical and institutional. Technical environments are characterized by markets that

exchange products or services produced by organizations which are compensated for

their efficiency. On the other hand, institutional environments support organizations

through rules and regulations to which the organizations must comply. Institutional

environments exist for economic organizations but are more prevalent for political

organizations. Political organizations, including state governments, are subject to voter

preferences and fiscal limitations; but the most stringent constraints come from already

established laws concerning citizens rights (March and Olsen 1999).

Federalism

Federalism is one way in which state governments were imprinted. Although there are

many different definitions of federalism (Gray and Eisinger 1991), Elazar (1984) defines

federalism as the mode of political organization that unites separate polities within an

overarching political system by distributing power among general and constituent

governments in a manner designed to protect the existence and authority of both (2).

Elazar points out that federalism ties state and federal government to one another. Both

are required to share in the creation of and in the implementation of laws, and power is

diffused between them as they share in political activities. Federalism requires state and

federal polities to be interdependent while simultaneously allowing enough autonomy to

maintain separate identities as policy-making institutions.

In the case of woman suffrage, the federal government provided a political

environment for and imprinted newly formed states and territories. Since states had to

petition Congress for admittance into the Union, state constitutions needed to conform to

federal policy and to other conditions mandated by Congress. Conformity to federal

laws by territorial governments was expedient or admission to the Union was denied.

Federalism illustrates the interdependent relationship between state and federal

government and also explains the timing of woman suffrage activity in state legislatures.

For instance, in 1869, Senator Young proposed a joint resolution recommending the U.

S. constitution be amended to extend suffrage to women. The reasoning was that while

the Constitution of the United States is being amended, extending the right of suffrage to

all men over twenty-one years of age, that the same right be granted, at least, to all

women over twenty-one years of age. . . . (Journal of the Senate for the State of West

Virginia 1869: 56). At this time, Congress was taking up the issue of enfranchising

black men and the federal amendment granting the vote to black men was ratified in

1870. The language and reasoning used by Senator Young illustrates the impact of

federal policy on state policy, despite the absence of a direct mandate by the federal

government to which state governments needed to conform.

After the Civil War boundaries concerning citizenship, previously limited to

white men, expanded. The Fifteenth Amendment, ratified February 7, 1870, provided

that citizens should not be precluded from voting on account of race and provided

universal suffrage to men. These changes in the definition and boundaries of

citizenship at the federal level provided a political context for states founded in the

West during and after this time period. The federal renegotiation of boundaries in

citizenship are reflected in the discussion and adoption of woman suffrage and

underscores the importance of organizational imprinting. The implication for this study

is that political conditions present at the founding of newly formed states will be

reflected in their structure.

It is no coincidence that Wyoming Territory was admitted to the Union in 1868

and passed suffrage one year later in light of the environment in which it was founded.

Wyoming Territory was influenced by the Civil War amendments which expanded

citizenship and citizens rights to black males. It was a natural leap to extend the

discussion of voting rights to women, especially since many women advocating woman

suffrage were heavily involved in the abolition movement in an attempt to secure the

franchise for blacks and themselves (Banaszak 1996). Coupled with the previous

discussion of territorial governments as arenas of political experimentation, federalism

provides an explanation for the timing of the enfranchisement of woman suffrage in the

Wyoming Territory.

The concept of imprinting demands attention to historical time period. States

founded after the Civil War amendments should reflect changes concerning citizenship

in the political landscape. Consequently, I expect states founded after the Civil War

amendments to be more likely to pass woman suffrage legislation.

Structural inertia

Structural inertia holds important implications for political innovation and for the

adoption of woman suffrage. Structural inertia is defined by Hannan and Carroll (1995)

as resistence to change. Since organizational resources are absorbed by facilities,

personnel, equipment, training, and costs of gathering and distributing information,

changing organizational structures and procedures involves time and money. Innovation

is not only difficult, but reform efforts often fail. Despite tremendous effort and

legislative activity, the suffrage movement in Iowa demonstrates the difficulty with

which change is achieved. Iowa s state legislature introduced 42 full suffrage bills, more

than any other state. Yet, Iowa did not grant woman suffrage.

The implications of structural inertia for state legislatures are obvious. States in

existence for longer periods of time will exhibit greater procedural constraints and less

policy innovation. Organizational learning processes, which encode inferences from

history into routines that guide behavior, shape an organization s future by encouraging

the homogenization of procedures and organizational action (March and Levitt 1999:

75). As an organization gains experience, learning processes tend to be self-limiting.

That is, learning reduces diversity of beliefs and behavior and it reduces variation in

performance resulting in the institutionalization of procedures (Levinthal 1991).

Through learning processes, procedures become informally and formally inscribed;

informally through social norms and formally through manuals, goals, and reports.

Although change is often prompted by changes in the environment, the organization s

structure mediates future environmental uncertainty. And political institutions are no

different. As in the case of Iowa, the legislative bodies of older states will be less able to

respond to cultural changes and popular opinion. I surmise that older states will be less

effective passing woman suffrage legislation and that newer states will be more

successful introducing, voting on, and passing woman suffrage bills.

Although my expectations are similar concerning structural inertia and the

previous discussion of organizational imprinting (i.e., that new states are more likely to

pass suffrage legislation), these organizational processes are conceptually disparate and

they are measured differently. Inertia describes the propensity to change as a function of

the newness of the state. Imprinting, on the other hand, pertains to the time period in

which the state was founded. That is, different time periods produce different

organizational forms.

Organizational size

The third organizational characteristic of state governments I examine is organizational

size. Scott (1992) reports that organizational structure tends to become more complex

over time, especially as its tasks and size increase. As organizational size increases and

tasks become more complex, coordination problems arise (Blau 1970; Blau 1972; Hall

1999; Scott 1992). It follows that larger state legislatures will more constrained by

coordination problems.

Most scholars who study organizational size and structure emphasize the effect

of the number of employees. However, as Scott (1992) points out, using the number of

employees as a variable has disadvantages. First, the boundaries between participants

and non-participants of organizations are often difficult to determine. Second,

participants differ from one another in the intensity of their work. This is problematic

when studying participants among different types of organizations because making

comparisons is difficult. However, the disadvantages of using the number of

organizational participants cited by Scott are negligible when examining state

governments since it is easy to determine the number of legislators from the legislative

journals and from other sources. In addition, my study does not deal with different types

of organizations.

Based on the conclusions of Scott (1992), that organizational structure becomes

more complex over time, smaller state legislatures of the West should be more amenable

to change and less constrained by problems of communication and coordination. It

seems more likely that a handful of legislators advocating woman suffrage would be

more successful in smaller states. Therefore, I expect legislative success in states with

smaller state legislatures.

Mimetic isomorphism

One challenge facing any organization is environmental change and uncertainty.

Although organizational fields (groups of interrelated organizations such as suppliers,

buyers, and regulatory agencies) show remarkable diversity early on, Dimaggio and

Powell (1983) argue that well-established organizational fields are characterized by

organizations that copy procedures and outcomes. Organizations often innovate by

copying rules and procedures of other organizations and they are often motivated to

imitate in order to improve their efficiency.

Dimaggio and Powell identify three types of institutional isomorphism which

occur as organizations seek cultural and political legitimacy: mimetic, coercive, and

normative. In my study I focus on mimetic isomorphism as a relevant practice of state

legislatures. Mimetic isomorphism occurs amidst high levels of uncertainty and

ambiguity. When dealing with problems of ambiguity, organizations seek legitimate

solutions despite any concrete evidence that the adopted models enhance efficiency

(70).

Mimicry is a common practice among state legislatures. Walker (1969)

maintains that states look to other states for direction in legislative matters. Often, state

governments imitate others who have successfully taken the lead in legislative matters.

For example, the 1911 legislative handbook for New Mexico contains a summary of the

various voter qualifications in other states. Walker also finds that regional clustering

occurs. Some states act as regional leaders while others only follow after the leaders

have enacted the legislation. As a result, I expect mimetic processes to influence woman

suffrage legislation and success. As more states within a particular region adopt

suffrage, other states will be more likely to follow.

Legislative process

The legislative process varies by state and each bill has to pass many stages to make its

way through the legislature. First, a bill needs to be introduced by a legislator who is

willing to sponsor it. After a bill is introduced in either house it is sent to a committee.

Committees can impede a bill s progress by recommending that a bill not pass, by taking

no action, by discussing the bill for an inordinate amount of time, or by sending it to

another committee altogether. For example, a full woman suffrage bill, introduced New

York in 1892, was referred to the committee on the judiciary and no action was taken.

The bill died in committee. Another bill, introduced in New York in 1878, was reported

to the committee on the judiciary and received a favorable recommendation.

Nevertheless, it was sent to another committee, the committee on the whole, and no

further action was taken. It should be noted that if a bill does not receive a favorable

recommendation by the committee, the bill can still be voted on.

After a bill is reported by the committee, it is read a second and third time before

it is voted on. At this stage a bill can also be tabled; meaning it will not be considered

further. When a bill is voted on it must pass by a constitutional majority. This varies by

states and is stipulated in the state constitution. A constitutional majority consist of

votes in the affirmative by one half, two thirds, three fifths, or sometimes three fourths.

After one of the houses passes a bill it is sent to the other house where it is required to

undergo through the entire process again in order to pass. Some state constitutions, such

as Nevada s, specified that bills must pass two consecutive legislative sessions. In fact,

Nevada s legislature passed woman suffrage bills on three different occasions (1869,

1883, and 1895) but the bills did not pass a second session and they were required to

begin the process anew. Once a bill passes both houses it is either signed or vetoed by

the governor. If the bill is signed by the governor, it is required to be ratified in a

referendum in which the electorate votes on whether or not the bill should become law.

Another aspect of the legislative process that influenced bill success is the frequency

with which legislative sessions are held. Some state legislatures met yearly but others

met less frequently, every other year. States which required two consecutive legislatures

to pass a bill may also be less likely to pass a bill.

The legislative process is arduous and its multiple stages can hinder legislative

success. I expect that states with more stringent legislative requirements larger

constitutional majorities and requiring two consecutive legislatures will obstruct the

success of woman suffrage bills. I expect that the frequency with which the legislature

meets will also influence the legislative outcome. Increased time between legislative

sessions should allow additional time for obstacles to arise. More obstacles might

include problems concerning coordination required to maintain support for suffrage

legislation and also procedural problems which could impede the success of suffrage

initiatives. Therefore, legislatures that meet more often should be more successful

passing woman suffrage bills.

Legislative outcome

Considering the difficulty of the legislative process and its importance to the timing of

woman suffrage success, I measure the effects of institutional politics theory and

organizational characteristics on different stages of the legislative process. I do this by

examining outcomes at various stages in the legislative process. The legislative stages

that I examine are (1) whether or not a bill was introduced in the house or in the senate,

(2) whether or not a bill was voted on in the house or in the senate, and (3) whether or not

a bill was passed in the house or in the senate.

Summary and hypotheses

This study uses data from state governments to understand why woman suffrage was

first adopted in western states even though most of the movement s resources were in

the East. To determine the effect of state building on the adoption of woman suffrage I

draw from institutional politics theory by looking at party support, the presence of third

parties, and political openness. I also draw from organizational theory by examining the

processes imprinting, structural inertia, and organizational size, mimetic isomorphism,

and the legislative process.

Since no literature exists on legislative success at various stages, I hypothesize

that the effects of institutional politics and organizational characteristics will effect

each stage of the legislative process. The following are the formally stated

hypotheses: H1: More democratized states increase the likelihood that woman suffrage

legislation is introduced, voted on, and passed.

H2: States with a larger number of political parties are more likely to introduce, vote on,

and pass suffrage legislation.

H3: States in which woman suffrage has a greater political presence (party endorsement

of woman suffrage and sole trader rights) have a greater probability of introducing,

voting on, and passing suffrage legislation.

H4: States founded after the Civil War Amendments (1870) are more likely to

introduce, vote on, and pass suffrage legislation.

H5: States with greater structural inertia are less likely to introduce, vote on, and pass

suffrage legislation.

H6: States with fewer legislators are more likely to introduce, vote on, and pass suffrage

legislation.

H7: States in regions with states that have passed woman suffrage are more likely to

introduce, vote on, and pass suffrage legislation.

H8: States with greater legislative constraints larger constitutional majorities, require

two consecutive sessions, and met less often are less likely to introduce, vote on, and

pass suffrage legislation.

Chapter 2 Method

This study examines state-level panel data to determine the impact of state

government on legislative outcome. The population under investigation includes all 45

states in which state legislatures could amend their constitutions between 1848 and

1918. The advantage of studying a population is that a population, versus a sample,

justifies interpretation of coefficients regardless of statistical significance. Nevertheless,

statistical significance is preferred to demonstrate the more stable effects and also to

emphasize which variables are generalizable.

Three of the contiguous states are excluded from this study: Utah, Wyoming, and

New Hampshire. Utah and Wyoming are omitted because woman suffrage was included

in their first state constitutions. In Utah and Wyoming, woman suffrage was proposed in

a constitutional convention and was approved by the electorate prior to statehood. New

Hampshire is excluded because their state constitution prevented the legislature from

amending the constitution. Their constitution could only be amended through a

constitutional convention. Several states constitutions similarly prevented the

legislature from amending the constitution in various years. These include Florida from

1865 to 1867, Indiana from 1848 to 1856, Iowa from 1848 to 1856, Kentucky from 1848

to 1890, Nebraska from 1867 to 1874, Ohio from 1848 to 1850, and Virginia from 1848

to 1869. These states years are also omitted from the analysis. Otherwise, states are

included beginning in 1848 or year of statehood. States are right censored when they

successfully amended their constitution to include woman suffrage or after the year

1918.

The analysis begins in 1848 since this is the year of the Seneca Falls Convention.

The Seneca Falls Convention is most often identified as the beginning of the feminist

movement in the United States. 1918 is the last year of the analysis since Congress

passed the Nineteenth Amendment in 1919, and state legislatures began voting on the

Federal Amendment at this time. The passing of the Nineteenth Amendment would have

biased state legislative action on woman suffrage legislation and also resulted in several

state suffrage campaigns that only focused on the Federal Amendment.

Data

Many studies of woman suffrage examine the movement from the perspective of the

eastern United States. This is a result of the data that are available. Most scholars use

the six-volume History of Woman Suffrage (Stanton et al. 1881) for information

concerning the woman suffrage movement. Unfortunately, Stanton et al. primarily focus

on the national movement. Their sources of information about state suffrage campaigns

are provided mainly by women affiliated with the NAWSA. Information concerning

suffrage activities independent of the NAWSA may or may not be included in this

record. Additionally, affiliations to the NAWSA were not as strong in the West as they

were in the East and not all suffragists were associated with the NAWSA (Beeton 1986).

In fact, Abigail Dunaway, the most prominent suffragist in the Pacific Northeast, refused

to be affiliated with the national organization (Duniway 1971). As a result of these

biases, Stanton et al. underestimate suffrage activity in the West. For instance, only 473

of the total 5,555 pages of the History of Woman Suffrage volumes are dedicated to the

movement in the western states (Schiffman 1998). In addition, information reported by

Stanton et al. concerning many of the western states, especially concerning legislation, is

incomplete and sometimes inaccurate.

For complete information I use data that I gathered with Professor Cornwall, Dr.

Torres, Kendra Schiffman, and Brayden King on woman suffrage legislation. This

research was funded by two grants from the National Science Foundation. The data on

woman suffrage legislation was gathered from state libraries and archives. At state

libraries we consulted legislative records (House and Senate journals) and found every

full suffrage bill that was introduced prior to 1919. Each bill was coded by state, year,

whether it was considered in the house or the senate, the date it was introduced in either

house, by whom it was introduced, the date it was referred to committee, the date it was

voted on in either house, whether or not it passed either house, and the number of votes

for and against the bill. After a bill passed one house the same information was coded

for the other house as it went through the same process. On some occasions the house

and senate journals were not indexed. In these instances the process of finding bills

included using clues from secondary sources such as dates that bills were introduced,

voted on, passed, and legislators who were mentioned that introduced bills and then we

searched page by page for the woman suffrage legislation.

Information on state structure was collected by consulting state constitutions,

statutes, secondary sources, the Office of the Secretary of State for many states, and

state librarians and archivists. I was assisted in this process by other research assistants,

Bruce

Lott and Liz Mykla. To find the years in which elements of democratization were

granted I began with Clemens (1997) and the sources she cites concerning elements of

political openness (see Heckelman 1995; Kettleborough 1923; Merriam and Overacker

1928; Ranney 1978). These sources provided data for states until 1913. To find the

years for the remaining states I looked through their statutes and constitutions. To find

the years in which sole trader rights were granted, I began with Kahn (1996). Her data

provided years for states in which sole trader rights were granted until 1900. To find the

remaining states I again consulted statutes and constitutions.

Another aspect of state structure is the number of state legislators in each state by

decade. This was gathered by contacting every Office of the Secretary of State (or state

librarian or archivist) in the contiguous U. S. via email or phone. Most of them emailed,

faxed, or mailed the desired information. Information for the remaining states were

found by searching state histories, Blue Books, and state constitutions. Data concerning

the legislative process were collected by searching every new state constitution during

the time period of this study. From state constitutions I identified changes in the

proportion of legislators required to pass a bill in the House and in the Senate, the

number of sessions required to pass a bill before it could be submitted to the electorate

for approval, and the frequency with which sessions were held (see Table 5).

Information concerning political parties were collected from Gubernatorial

Elections, 1787-1987 (1998). This source includes votes for gubernatorial candidates

from all parties receiving at least five percent of the vote cast in an election. I also use

one variable, party support, from McCammon and Campbell (2001).

Dependent variable: legislative outcome

There are three dependent binary variables in this analysis. The first variable is coded 0

or 1 indicating whether or not a woman suffrage bill was introduced in a legislative

session in either the House or in the Senate (n=516). The second variable is coded 0 or

1 indicating whether or not a woman suffrage bill was voted on in a legislative session in

either the House or in the Senate (n=236). The third variable is coded 0 or 1 indicating

whether or not a woman suffrage bill was passed in a legislative session in either the

House or in the Senate (n=126). Only full woman suffrage bills are included in this

analysis. Full suffrage bills enfranchised all women in a state as opposed to partial

suffrage bills that grant suffrage in certain matters: school elections, presidential

elections, primary elections, etc. Table 6 shows the distribution of legislative sessions in

which a full woman suffrage bill was introduced, voted on, and passed by decade from

1848-1918. Because other studies may report different numbers of full suffrage bills, it

should be noted that I do not include territorial suffrage bills (n=49). In addition, my

analysis counts the sessions in which a bill was introduced, voted on, or passed; not the

total number of bills by outcome. In other words, I do not count the number of bills in a

legislative session. I only count whether or not a bill was introduced, voted on, or

passed.

Independent variables

Four variables measure aspects of institutional politics. First, democratization is an

index indicating the number of elements of progressive politics that a state has adopted

for any given year. These elements are the constitutional initiative and referendum, the

Australian ballot, and the mandatory direct primary. Second, total parties is the number

of political parties represented in a gubernatorial election (up to four). This variable

was created by coding elections in which one party s gubernatorial candidate received

85 percent or more of the vote as one party. Elections in which two candidates received

votes were coded as two parties. Elections in which three or four candidates received

votes were coded as three and four parties respectively. For the years in which there

was no gubernatorial election the number is attained from the last election that took

place. Third, party support measures political presence and is a dichotomous variable,

coded 1 or 0, indicating whether or not woman suffrage was included in a Democratic

or Republican party platform for any given year. One problem with this variable is that

it begins in 1866. Nevertheless, it is probable that woman suffrage was not endorsed by

the Republican or Democratic parties before this date since it was not a prominent

political issue for politicians prior to the Civil War. According to the data the first year

that a party endorsed suffrage was in Ohio in 1873. Fourth, sole trader rights also

measures political presence and is a dichotomous variable. It is coded 1 after sole

trader rights have been passed in a state and 0 for the years that they have not been

passed.

I use five variables to measure organization characteristics. Imprinting is a

dichotomous variable, coded 1 or 0, indicating whether or not a state was founded in

1870 or later; after the Civil War Amendments were passed. Structural inertia is the age

of the state. Organizational size is the total number of legislators in the state legislators.

Mimetic isomorphism is the number of states in the same region that adopted woman

suffrage. The regions are based on the geographic regions utilized by Elazar (1984). I

measure legislative process by using three different three variables. First, constitutional

majority is the proportion of legislators required to pass a bill. Second, sessions

required is the number of legislative sessions required to pass a bill. Third, session

frequency is the frequency (measured in years) with which a legislature meets. Table 7

shows the means and standard deviations of the independent variables.

I use a random-effects regression model for binary data (Maximum Likelihood

Estimation). The model is formally stated as

yit = xNit $ + "i + eit

where yit is the binary probability of an event for the ith individual and the tth wave.

The expression xNit indicates the index function, where $ = ($0, . . . , $K)N is the K + 1

column vector of regression coefficents. The unit-specific effects (the unobserved

heterogeneity) are expressed by "i and xNit = (xit0, . . . , xitK) is the K + 1 row vector of

independent variables corresponding to the ith individual and the tth wave. The error

term, eit, represents the time-varying stochastic error. These are assumed to be normally

distributed with mean zero. . . . Additionally, the error term is assumed to be

uncorrelated the independent variables and with "i (Powers and Xie 2000, pp. 60, 183).

All statistical analyses are performed using Intercooled Stata Version 7 (StataCorp

2001b).

I use three models in my analysis. Each model represents a stage in the legislative

process: whether or not a bill was introduced in either house, whether or not a bill was

voted on in either house, and whether or not a bill was passed in either house.

Since bills in each stage are conditional on and are a subset of the previous outcome, I

use a sequential logit model (Mare 1980; Powers and Xie 2000). For instance, in the

first stage I model the log odds of a bill being introduced for the years in which a

legislative session was held. In the second stage, I model the log odds of a bill being

voted on only in sessions in which a bill was introduced. This is accomplished by

selecting only the sessions in which a legislature introduced a bill. In the third stage, I

model the log odds of a bill being passed in the instances in which a bill was voted on.

This is accomplished by selecting only the sessions in which a legislature voted on a

bill. I report the log odds for all of the independent variables. In order to emphasize the

findings that are generalizable to other populations, I report the probabilities for only the

significant independent variables.

The log odds can be interpreted as odds ratios or probabilities. I calculated the

probabilities using an algorithm employed by Powers and Xie (2000). The baseline

probability is calculated by first multiplying the log odds of the random-effects logistic

regression model and the mean for each independent variable. The dummy variables are

multiplied by their mean. This is done to prevent the baseline models from being equal

when the dummy variables are set at zero. Next, the log odds of the constant is

multiplied by one. The products, consisting of the log odds multiplied by the mean, for

each independent variable and for the constant are summed. The baseline probability is

determined by dividing the exponent of the sum by 1 + the exponent of the sum. This is

formally stated as

exp ($0) / [1 + exp ($0)

The probability for an independent variable at any given value is calculated by

first multiplying the log odds by that value instead of by the mean. Then, holding all

other independent variables constant at their mean and multiplying the log odds of the

constant by one, the products for each independent variable and for the constant are

summed. As in the baseline probability, the exponent of the sum is divided by 1 + the

exponent of the sum.

Chapter 3 Results

From 1848 to 1918, 516 legislative sessions introduced a woman suffrage bill in

either house, 236 sessions voted on a bill, and 126 sessions passed a bill. Suffrage bills

that were introduced in either house had a 46 percent chance of being voted on (236 out

of 516). Of these, 53 percent were passed (126 out of 516). The overall success rate was

25 percent (126 out of 516). The Senate, which is generally considered the more

conservative legislative body, was 10 percent more successful passing bills than the

House (Table 6).

The sessions that were held between 1910 and 1918 introduced 40 percent of all

woman suffrage bills. This finding emphasizes the increasing activity in suffrage

legislation. It also shows the effectiveness of early adopters since fewer bills were

introduced in the earlier decades. It is important to note the decline in suffrage activity

from 1900 to 1910. This decade is known as the doldrums in the woman suffrage

literature because no states granted suffrage. The doldrums were a backlash following

the success of the prior decade.

A correlation matrix reveals that several independent variables are moderately

correlated (Table 8). The strongest correlation is between structural inertia and

democratization (0.491). This is a function of time since both increase as year increases.

Structural inertia and legislative size are also correlated (0.424). These variables are

correlated because newer western states had smaller legislatures while state legislatures

in the East were generally larger. In addition, imprinting and structural inertia are

negatively correlated (-0.392) as are session frequency and sessions required (-0.368).

These correlations are also a function of time. Sole trader and democratization are

correlated (0.344). This stands to reason since more democratized states passed sole

trader rights.

For each legislative stage a separate model is run for each group of variables

institutional politics and organizational characteristics. In the first model I estimate the

log odds of institutional politics and in the second model I estimate the log odds of

organizational characteristics. The third model combines all of the variables from both

groups. By analyzing the effects of the groups of variables separately in the initial

models and by comparing them to the third model, I can determine whether one group of

variables has an effect on the size and sign of the coefficients in the other group. In the

fourth model I drop the insignificant variables in order to emphasize the most important

variables and also for parsimony. I report the effects of all independent variables since

the data come from the population of state legislatures that are capable of amending

their constitutions. However, I report and interpret only the probabilities for the

significant variables to emphasize the more stable and the generalizable effects. In all of

the models I control for temporal effects by including year as an independent variable.

Bill introduced

Table 9 shows the log odds of the independent variables for whether not a woman

suffrage bill is introduced in a legislative session. In each model the effect of year is

positive and significant. Holding the other variables constant at their mean, the

probability of introducing a bill in 1854 (the first year in which a full suffrage bill was

introduced) is almost one half of one percent while the probability of introducing a bill

in 1918 is 43% (Table 12). This is one of the strongest effects in the model and indicates

that bills are more likely to be introduced over time.

Three institutional politics variables are significant. Democratization is

significant but the coefficient is negative. This finding is surprising and indicates that

democratized states are less likely to introduce a woman suffrage bill. However, the

effect is small. The probability of introducing a bill decreases from 8% for a state that

has not passed any element of political openness to 5%, 3%, and 2% as states pass one,

two, or three elements respectively. Democratization is positively correlated with

structural inertia meaning that newer western states were more democratized. This

suggests that more democratized states introduced fewer bills because the they were

more efficient in the legislative process. These states ended up needing fewer legislative

sessions to pass suffrage which is the case in the Intermountain West. The effects of sole

trader rights are negative and nonsignificant.

The positive, significant coefficient for total parties indicates that the presence of

third parties increased bill introduction. Nonetheless, the effect is small. The

probability of bill introduction increases from 4% in states represented by two parties in

the prior gubernatorial election to 10% when four parties were represented. The effect

of party support is also significant and positive. The significant and positive effect

shows, as predicted, that woman suffrage bills are more likely to be introduced in years

in which the Democratic or Republican party supports woman suffrage in its political

campaign.

Party support more than doubles the probability of bill introduction from 4% to 11%.

Of the organizational characteristics, imprinting, the number of legislative

sessions required to pass an amendment, and the frequency of legislative sessions are

not significant. The other legislative variable, constitutional majority, is negative and

significant. Therefore, states with smaller constitutional majorities were more likely to

introduce bills. The probability that a state requiring a one-half majority as opposed to a

two-thirds majority slightly increases the probability of bill introduction from 3% to 7%.

The other organizational variables are statistically significant and their impact

confirms my hypotheses. Structural inertia s negative coefficient indicates that younger

states were more likely to introduce suffrage bills. An increase in structural inertia from

its value at the 25th percentile (38 years) to its value at the 75th percentile (94 years)

decreases the probability of introducing a bill from 5% to 3%. However, the size of this

effect is negligible. The positive effects of organizational size and mimetic isomorphism

indicate that larger state legislatures and more states in the region with suffrage increase

the likelihood that a state will introduce a woman suffrage bill. The change in

probability for organizational size at the 25th percentile (103 legislators) to the 75th

percentile (173 legislators) is minimal. It increases from 4% to 5%. The probability of a

bill being introduced increases from 5% to 32% when the number of states in the region

with woman suffrage increases from zero to four states. States requiring a larger

constitutional majority slightly decreases the likelihood of introducing a bill from 8% to

3%.

The rho statistic is used to ensure that the model is properly specified by

comparing the pooled estimator (logit) with the panel estimator (StataCorp 2001a: 380).

If rho is significantly different from zero, a random-effects model is properly specified.

The rho statistic in Table 9 is significantly different from zero for each model.

In sum, mimetic isomorphism has the greatest effect on bill introduction. The

other significant institutional politics variables and organizational variables have only a

slight impact. Despite the small effects of most of the variables, the direction of the

effects of total parties, party support, inertia, and mimetic isomorphism confirm my

hypotheses. My hypotheses concerning democratization and organizational size were

not confirmed.

Bill voted

Table 10 lists the log odds for the models in the next stage of the legislative process. It

is important to note the number of cases decreases from the previous model, 1683 to

347. The models in Table 10 are a subset of the models in Table 9. The number of cases

decrease since a bill could only be voted on in sessions in which a bill was introduced.

347 is the number of sessions in which a bill was introduced and then voted on in either

the House or Senate.

As expected, the significant institutional politics variables, sole trader, total

parties, and party support, have a positive effect. These findings show that states

granting sole trader rights are more likely to vote on a suffrage bill. After a state passes

sole trader rights the probability of voting on a bill increases a considerable extent, from

34% to 62%. States in which more parties are represented in gubernatorial elections

also increases the likelihood of voting on a bill. When the number of total parties

represented increases from two to four, the probability of voting on a bill increases

moderately from 53% to 65%. The last significant institutional politics variable, party

support, increases the probability of voting on a bill when the bill is endorsed by a

political party. The probability increases from 49% to 70%; a relatively strong effect.

The other institutional politics variable, democratization, is positive though

nonsignificant.

Of the organizational characteristics, imprinting is not significant and it is

negative. The findings concerning structural inertia are similarly nonsignificant but

organizational size is significant. States with smaller legislatures are more likely to vote

on a bill. An increase in the legislative size from the 25th percentile to the 75th percentile

decreases the probability of voting on a bill from 61% to 51%. The effects of mimetic

isomophism and constitutional majority are positive but not significant.

The last two variables, sessions required and session frequency, are positive and

significant. These findings indicate that an increased number of session required to pass

a bill increases the likelihood that a bill will be voted on. An increase in the number of

sessions required from one to two increases the probability of voting on a bill from 40%

to 70%. This is a large effect but does not support the idea that more stringent

legislative procedures reduce the likelihood of voting on a bill. As for session

frequency, the less often a legislative session is held, the less likely a bill will be voted

on. The probability of bills being voted on increases from 38%, if sessions are held

every year, to 63%, if sessions are held every other year.

The rho statistic is not significant. This means that the time variant

heterogeneity of the overall model is not significantly different from zero. This is likely

a result of the decline in the number of cases.

To summarize, the large effects for sole trader, total parties, and party support

indicate that institutional politics are more important for voting on a bill than for

introducing a bill. The effects of sole trader and party support become substantial while

the effects of total parties are modest. Compared with the last stage mimetic

isomorphism becomes less important. The only significant organizational variables are

organizational size, sessions required, and session frequency. Organizational size has a

slight effect, but sessions required and session frequency have large effects and are

important factors for whether or not a bill is considered. While the impact of

institutional politics was expected and its importance increased over the previous

legislative stage, the direction of the effects of sessions required and session frequency

was unexpected.

Bill passed

Table 11 shows the log odds of the independent variables for passing a bill in at least one

house of the legislature. The cases decrease in number from the previous set of models

from 347 to 206. The cases decrease to include only the sessions in which a bill was

passed in at least one house as a subset of sessions in which a bill was voted on.

In this stage democratization is positive and significant. As democratization

increases, state legislatures are more likely to pass suffrage bills. The probability of a

state legislature passing a bill increases from 24% to 78% when the number of elements

of democratization increases from zero to three. This increase represents a tremendous

effect on the success of a suffrage bill. Sole trader becomes negative and nonsignificant.

The log odds of the other institutional politics variables are positive, significant, and

have substantial effects. An increase in total parties and party support increases the

likelihood that a bill will passed. These findings confirm my hypotheses. The

probability that a bill will be passed increases from 43% for a state represented by two

parties to 89% for a state represented by four parties. The probability of a bill passing in

a state with party support increases from 39% to 61%.

For the organizational variables, the effect of imprinting is positive but

insignificant. This effect is in the opposite direction for the models in Tables 10 and 11.

The significant negative effect of structural inertia shows, as predicted, that newer states

are more likely to pass a suffrage bill. The probability of passing a bill decreases from

62% percent for newer states with values in the 25th percentile to 33% for older states

with values in the 75th percentile. Compared with the first stage, this effect is

substantially larger. The log odds for organizational size and sessions required are

positive, though nonsignificant. The log odds for session frequency are negative and

also nonsignificant. As expected, constitutional majority is significant and decreases the

likelihood of passing a bill. An increase in constitutional majority from one half to two

thirds decreases the probability of a bill passing from 55% to 36%.

The rho statistic is not significantly different from zero. As in the previous

legislative stage, this is likely a result of the decline in the number of cases.

Compared with the last two stages of the legislative process, the effects of the

significant variables have increased to a considerable extent. Democratization and total

parties were significant in Table 10 but their effects were minimal. However, their

effects have become noticeably larger in this stage. Party support, which was

significant in the preceding models, is a consistent predictor at each legislative stage

with its effects tripling between the first and second stages. Its effects remain constant

between the second and third stages. The effects of structural inertia on passing a bill

substantially increase compared with introducing a bill and its effects were not

significant in the previous stage. Similarly, constitutional majority was previously not

significant but its effect is large in this legislative stage.

Chapter 4 Discussion and conclusion

The results demonstrate the importance of examining characteristics of state

government at different legislative stages. Institutional politics are significant

throughout each stage but have the greatest impact on bills passing. The majority of

organizational characteristics are significant in the initial stage but only structural inertia

is consequential in the final stage. In addition, sessions required and session frequency

are important for whether or not a bill is voted on. The findings paint a picture of

innovative states that are smaller and newer, reside in regions granting suffrage, are

more democratic, have multiple parties vying for power, and have a party that endorses

woman suffrage. It is not surprising that these characteristics were widespread in the

western states.

Institutional politics

Democratization increases the likelihood of passing a bill, but has a negative effect on

introducing a bill. Although the latter finding is unexpected, the effects are minimal.

The former finding, however, supports my hypothesis and its effect is substantial.

Democratized states were more willing to increase voter participation by extending the

vote to women by passing woman suffrage legislation. This confirms Clemens (1997)

findings that western states were more democratic.

Sole trader rights is one indication of states support for women s rights in general.

However, sole trader rights is only predictive in the voting stage. This suggests that

sole trader rights were not necessarily an indication of gender equality in a state. States

granting sole trader rights were only more likely to consider the issue. In fact, though

the effect is not significant, sole trader rights had a negative impact on bill introduction

and on bill passing. Sole trader rights may have been one step toward equality but it

did not guarantee further political opportunities for women. Similar dynamics took

place relative to women s employment during this time period. Although more women

were moving into nonagricultural employment at the turn of the century, women were

becoming teachers, nurses, and social workers; all traditional women s occupations.

The West might have provided different conditions and expanding opportunities for

women in some regards but conditions were also gendered and continued to reinforce

traditional gender norms. States granting sole trader rights were not as fertile arenas

for gender politics as one might think.

The number of political parties was predictive of whether or not a bill was

introduced, voted on, and passed. The size of the effect increases with each legislative

stage demonstrating that the importance of third parties gets stronger at each stage.

These findings suggest that pluralism in the political arena is indicative of the extent to

which political minorities can access the political system. In the case of woman

suffrage, third parties, particularly the Progressive Party, were more supportive of

woman suffrage. These findings also provide insight into the late adopting states,

Michigan and New York. The Progressive Party s attempt to gain presidential votes in

the West by endorsing woman suffrage would have been cause for alarm among

politicians in the East. This is reflected in New York in the 1917 election campaign. To

gain women s support, Republican and Democratic district leaders sought women for

positions in the parties previously occupied by men (Dinkin 1995). The New York

Times also recognized the important role women would play in the election campaign of

1912. The Times reported that the 1912 campaign had become feminized because the

Progressive Party supported woman suffrage and involved women in the party

(Gustafson 2001). Progressive support not only helped the woman suffrage movement

but its threat prompted the Republicans and Democrats to include a woman suffrage

platform in the 1916 presidential campaign. In fact, this was the first year the

Democrats included a woman suffrage platform in their campaign (Wolbrecht 2000).

To test whether or not the data show that the presence of third parties in state

politics prompted Republican or Democratic support for suffrage, I first ran Model 4 for

bills voted on and bills passed and dropped party support. In the new model, the effect

and significance level of total parties increases. Since party support s inclusion in the

model decreases the effect of total parties, party support accounts for some of the

variance of total parties. Next, I ran a random-effects logistic regression model with

party support as the dependent variable and total parties as the independent variable.

The log odds were positive and significant (p<.001). This demonstrates that third party

presence prompted party support for suffrage from Democrats and Republicans.

Party support for woman suffrage was predictive of suffrage success at all three

stages in the legislative process. The effect was weak for bills introduced but increased

for bills voted on and bills passed. These findings emphasize the importance of

legitimation in adopting policy innovation benefitting political minorities, particularly

women. These findings indicate that more in depth analysis is required to discern the

implications of party politics for woman suffrage and other policy issues.

Organizational characteristics

Although organizational imprinting is not significant, it is significant when it is the only

independent variable in the model for each stage. The coefficient in each model is

positive. So, states founded after 1870 were more likely to introduce, vote on, and pass

a suffrage bill. However, when age and legislative size are added to the model the

effects of imprinting disappear. One reason for imprinting s weak effect might be that it

is better measured by state age and legislative size. Stinchcombe s (1965) imprinting

argument is that historical time period influences organizational form and produces

organizations that exhibit similar characteristics. State legislatures founded after the

Civil War

Amendments were newer and had smaller legislatures. Consequently, the organizational

characteristics used in this study are picking up the effects of imprinting on legislative

activity and success.

States with less structural inertia were more successful in all three phases of the

legislative process but the effects were only consequential for passing a bill. This is not

surprising and shows that newer states were more successful passing suffrage bills.

What about states that granted woman suffrage although they were founded early, such

as New York, which granted woman suffrage in 1917? The effects of structural inertia

still help to explain late adoption. Despite being the birthplace of the woman suffrage

movement and despite being home to the NWSA, New York took 69 years from the

Seneca Falls Convention to grant suffrage. A woman suffrage bill did not pass both

houses of the state legislature for two consecutive legislative sessions until 1915. That

is 67 years after Seneca Falls. Structural inertia was an element that prevented New

York, the nexus of suffrage activity and ideology, from adopting earlier. Accordingly,

inertia helps predict the timing of success among the late adopters as well as the early

adopters. That is, older states are more likely to take longer to adopt innovations, if they

adopt at all.

The size of the state legislature has an effect on introducing a bill and voting on a

bill but the effects are minimal. Larger state legislatures are slightly more likely to

introduce suffrage bills. It makes sense that larger legislatures represent a larger range

of interests and interest groups. Smaller legislatures, on the other hand, were more

willing to vote on suffrage bills but were not willing to pass them. One explanation is

that smaller legislatures may have been easier to coordinate and convince to at least

consider the issue. However, passing the bill was another matter altogether.

The effects of organizational size raise the issue of innovation in relation to large

and small organizations. Small legislatures might be more easily persuaded by a strong

leader. This was the case in Wyoming whose first governor was largely responsible for

the success of woman suffrage. At the other end of the spectrum, large organizations

can be extremely complex and, I think, are susceptible to rapid, even unexpected

change. A survey of studies on organizational complexity and innovation cites studies

with contradicting findings. Some studies find that innovation takes place in smaller,

simple organizations while other studies find that innovation occurs in large, complex

organizations (Damanpour 1996). These disparate findings can be reconciled if a

curvilinear relationship exists between organizational complexity and innovation. That

is, I surmise that organization innovation occurs most frequently in newer, simple

organizations and also in older, larger, complex organizations. This issue deserves

greater attention by organizational scholars and is an area that should be in the future.

Mimetic isomorphism is only significant in the first stage of the legislative

process but the effects are large. Walker (1969) argues that states look to one another

when legislating policy, and it seems reasonable that isomorphic processes would

influence bill introduction. However, other states in the same region with suffrage is

not a sufficient condition to motivate legislators to pass a woman suffrage bill.

Legislators must answer to their constituency, especially if they want to be reelected.

And, other states granting suffrage in the region would not make a difference in passing

suffrage bills if woman suffrage is unpopular; that is, if voting for suffrage would

diminish a legislator s chance for reelection. It may also be that I have used the wrong

measure of mimetic isomorphism. The number of states in the region may not be as

important as the number of states in the West that have passed woman suffrage. State

legislatures were aware of the actions of each state that passed suffrage. The more

states in the West that passed suffrage, the more salient the issue became as is evidenced

during the time of the Progressive Party.

The last three variables concern legislative rules and procedures and the effects

are inconsistent between each stage. The effects of constitutional majority are

pronounced for passing a bill. Passing a bill when a smaller majority is required

increased the probability by approximately 20%. This finding should not be

underestimated considering many votes for woman suffrage were close and lost by a

small margin. This finding confirms the idea that stringent amendment procedures

impact legislative outcome.

It is interesting that the number of sessions required does not significantly

influence a bill passing but it does influence whether or not the legislature votes on a

bill. In other words, requiring two consecutive sessions to amend the constitution

increases the odds that a bill will be voted on. One explanation is that since the bill has

to be introduced and passed in an additional session before it goes to a referendum,

legislators are not threatened by simply voting on the issue. Passing an additional

legislative session requires additional steps and the status quo may not be threatened by

a vote under these circumstances.

Session frequency is also significant for whether or not a bill is voted on. The

less frequently the legislature meets, the more likely they are to vote on a bill. It may be

that the same set of legislators are more likely to be aware of the outcome and they are

more likely to be sure that the status quo will be upheld. Further research should take

into account the extent of legislative turnover between sessions. Though specific effects

need to be explored in greater detail, the overall impact of legislative procedure on

voting on a bill and passing a bill is substantial.

Conclusion

The primary contribution of this study is a greater understanding of the extent to which

state government influences legislative outcomes. Although innovation is often

prompted by environmental changes, the structure of state government mediates these

changes (Amenta and Halfmann 2000). Institutional politics theory and organizational

theory provide a framework for understanding the processes that constrain

environmental changes at different stages of the legislative process.

Institutional politics theory has a substantial effect on legislative outcome. More

democratized states have a large impact on passing a bill. The findings show that party

politics are also important to the success of woman suffrage. Third party presence was

not only important in state politics but during the 1910s the threat of the Progressive

Party stimulated suffrage support in the national campaigns of the Republicans and

Democrats as well. The literature on the woman suffrage movement has largely ignored

the dynamics of party politics. Party politics have been mostly studied in individual

state histories of woman suffrage. However, this study shows, contrary to other findings

(see McCammon et al. 2001), that third party politics were consequential and deserve

further attention. McCammon et al. determined third party strength by measuring the

percent of third party seats in the legislature. This study differs from McCammon et al.

by counting the number of third parties represented in the state elections. The results

indicate that my measure is a more accurate measure of the influence of third parties on

woman suffrage legislation. My findings suggests that a more accurate measure for

McCammon et al.

would be a count of the number of third parties represented in the state legislature.

The overall effects of organizational characteristics were small relative to the

effects of institutional politics. Despite the small effects, the impact of organizational

characteristics on bills introduced shows that the organizational structure can act as a

filter that constrains changing cultural norms and political agendas. The organizational

characteristics seemed to curbed the extent to which suffrage bills were introduced. In

this way the organizational structure mediates cultural changes, preferences of state

legislators, and preferences of interest groups. As the legislative process proceeded,

organizational characteristics grew less important but their effects became increasingly

substantial. As a result, this study identifies several organizational characteristics that

were consequential to bill passage. Specifically, structural inertia and legislative

majority are robust predictors of whether or not a bill will pass.

Up to this point research on woman suffrage has focused primarily on the

national movement or on cultural changes that led to woman suffrage adoption. These

studies do not, however, help us understand the impact of state building on policy

adoption and do not adequately explain the timing of changes in specific states,

especially in the West. This study demonstrates that understanding the legislative

process and party politics provides insight into the timing of woman suffrage success at

the state level. For instance, New Mexico was the only state in the Intermountain West

that did not pass full woman suffrage. However, their constitutional amendment

requirements for granting woman suffrage were extraordinarily stringent and their party

politics were marked by racial issues involving Spanish speakers and their attempt to

maintain political power that were not as prevalent in other western states.

This utility of this study is not only its contribution to the study of woman

suffrage but also to other state-policy issues. The model presented here is not restricted

to historical state-level contexts, such as temperance or civil rights, but also to

contemporary issues such as same-sex unions and felon disfranchisement. For instance,

Vermont and Hawaii have passed legislation enabling gays and lesbians to obtain

samesex unions. Alternatively, California has defeated similar proposals. Why did

Vermont and Hawaii pass same-sex legislation while California did not? I propose that

this question can be answered by examining a state legislature s capacity for reform by

implementing the same model I use to study woman suffrage success. Such an analysis

would include understanding whether or not a political party supported the issue, the

number of third party challengers, and other organizational characteristics. Some

organizational characteristics not included in my model as a result of a lack of data

should also be considered such as the number and size of legislative committees and

legislators political affiliation.

In addition, a study of felon disfranchisement reveals that only two states

currently have no restrictions on felons voting rights (Manza and Uggen 2002). These

two states are Maine and Vermont. Thirteen states have the next fewest restrictions.

These restrictions prohibit inmates from voting who are currently serving prison terms.

Hawaii is one of these states thirteen states. It should not be surprising that two of the

same states are progressive in the cases of same-sex unions and felon disfranchisement:

Vermont and Hawaii. A study of the characteristics of state government in Vermont and

Hawaii should reveal less stringent legislative requirements and also a more accessible

political system.

Understanding the organizational characteristics and the party politics of state

government from 1848 to 1918 demonstrate why some states were late adopters of

woman suffrage and why some were early adopters. Despite organizational resources,

New York was one of the last states to adopt woman suffrage. This is attributed to

legislative constraints and also partisan politics preceding the 1917 suffrage campaign

which prompted the inclusion of women as party workers and as electors. This study

helps us understand why some states, particularly in the West, were early adopters of

woman suffrage; why some states were late adopters of woman suffrage; and why some

states never adopted woman suffrage.

Students also viewed