1. Introduction
Under democratic regimes, elected individuals represent their constituents and deliver their voices to the policy-making process in the legislature. However, the type of representation they provide differs significantly across electoral systems. The extensive amount of research on elections demonstrates that electoral rules create particular incentives for individual candidates seeking office. The electoral systems with intra-party competition particularly motivate individual candidates to develop a close relationship with their constituents in order to gain personalized electoral support (Cain et al., Reference Cain, Ferejohn and Fiorina1990; Carey, Reference Carey2007; Heitshusen et al., Reference Heitshusen, Young and Wood2005; Lancaster and Patterson, Reference Lancaster and Patterson1990; Norton and Wood, Reference Norton and Wood1993; Stratmann and Baur, Reference Stratmann and Baur2002). However, not all legislators equally pursue local interests in parliament after they get elected under such systems. Individual legislators devote themselves to different activities in parliament. For instance, some legislators pursue the opportunities to participate in parliamentary debate that discusses the issues of government spending and authority for special interests, while others spend their time and energy on general public goods provisions for the society as a whole. Why is there such a significant difference among legislators in their engagement in parliamentary activities?
The present study addresses this question by focusing on the personal attributes of individual legislators. The growing literature centered on the cultivation of personal votes pays attention to the personal vote-earning attributes (PVEA) of legislators, such as local-level political experience and local birthplace (Shugart et al., Reference Shugart, Valdini and Suominen2005; Tavits, Reference Tavits2010). These studies argue that the personal attributes of candidates affect their electoral prospects; and candidates with PVEA are more likely to win elections than their counterparts without such attributes because they are able to establish strong support bases in their electoral districts by attracting personal supporters in their constituency. The important underlying assumption of this argument is that voters use candidates’ attributes as informational shortcuts in elections (i.e., voters expect that candidates with strong local ties are better able to represent and promote local interests than those without such attributes). Numerous studies have demonstrated how legislators approach voters and cultivate personalized support under electoral systems with multi-member districts (Ames, Reference Ames1995; Crisp and Desposato, Reference Crisp and Desposato2004; Hirano, Reference Hirano2006; Nemoto and Shugart, Reference Nemoto and Shugart2013; Wood and Young, Reference Wood and Young1997). However, it remains unclear whether legislators exhibit certain behaviors in parliament as is expected from their personal attributes (e.g., whether legislators with PVEA pursue local interests in parliament). This is an important issue because, if legislators with PVEA do not exhibit constituency-oriented behavior in parliament, voters cannot use their attributes as reliable information shortcuts. As a result, the electoral strength of legislators with PVEA may not be derived from the informational value of their attributes.Footnote 1
To demonstrate whether and how the personal attributes of legislators affect their parliamentary behavior, this study examines the individual-level activities of legislators in parliament, such as initiation of private members’ bills and questions at committee meetings. In most parliamentary democracies, individual legislators have the authority to propose their own bills to parliament (Mattson, Reference Mattson and Döring1995). Such types of legislation are called private members’ bills.Footnote 2 Similarly, legislators are also able to pose oral questions to cabinet ministers and government officials at committee meetings. The advantage of analyzing such parliamentary activities at the individual level is that they provide a means to observe legislators’ independent actions in the pursuit of their personal goals after having won an election.Footnote 3 While individual legislators may also engage in casework and constituency services, it is difficult to measure such activities because they are not publicly recorded in a comparative manner.Footnote 4 In contrast, since there are public records for when, where, and how many times each legislator initiated a bill or posed questions in parliament, it is possible to accurately track parliamentary activities across individual legislators, and over time.
Despite these advantages, there have been very few systematic studies on these parliamentary activities among legislators at the individual level.Footnote 5 This is partly because the parliamentary activities of individual legislators have only been considered to play a marginal role in parliamentary democracies, where political parties are the dominant actors in policy formulation. However, individual legislators, both from governing and opposition parties, have initiated numerous private members’ bills (despite the majority of them being dismissed) and posed many questions in committee meetings.Footnote 6 It is thus important to understand why vote-seeking legislators with limited resources bother to engage in such activities in parliament.
When explaining their parliamentary behavior, some studies have emphasized the importance of paying attention to seniority rules and the electoral vulnerability of legislators (e.g., Epstein et al., Reference Epstein, Brady, Kawato and O’Halloran1997; Kellermann, Reference Kellermann2013; Tatebayashi, Reference Tatebayashi2004). However, the diversity of parliamentary activities among legislators is not fully explained by the variation in the electoral strength and seniority of individual legislators. Through drawing on unique data from Japan, this study demonstrates that the personal attributes of legislators also affect their choice of parliamentary activities.
Legislators with different personal attributes can use parliamentary activities as an important means to inform their constituents about what they can do for them and how they differ from other candidates. In elections with intra-party competition, candidates have to adopt different promotional strategies, depending on whichever is the most suitable for them to be elected and further advance their political career. That is, individual candidates have different personal attributes (such as the strength of local ties, and possession of legal knowledge and expertise) as their personal backgrounds differ from one another; and they exploit the best image drawn from their personal attributes during the electoral campaign (e.g., legislators with strong local ties promise their constituents that they will represent the local interests if they are elected). These candidates cultivate personal votes by taking advantage of their personal attributes, and, once elected, they behave in accordance with these attributes in order to maintain their electoral ground for future elections.
The results of the empirical analyses conducted in this study support this argument. There exists a systematic difference among legislators in their engagement in parliamentary activities. Legislators who have strong local ties tend to engage in particularistic pork-barrel politics that will benefit their own support bases, while legislators who have legal knowledge allocate their time to general policy-making activities that will enhance their public image and visibility as a legal expert. These findings suggest that the composition of legislators in parliament has clear policy implications for legislative politics, and that the personal attributes of legislators, which can help predict their parliamentary behavior in the post-election period, will serve as important cues for voters in elections. Furthermore, the results also confirm the clear linkage between the individual-level characteristics of legislators and their local versus national orientations in their types of representation. These results contribute to the literature on the incentives for cultivating personal votes, and enhance our understanding of political activities among individual members in parliament.
2. Theory and hypotheses
Party labels play an important role in elections. However, legislators have individualistic incentives under electoral systems with intra-party competition, whereby they have to compete with other candidates running from the same party. Under such systems, candidates with strong local ties, such as local birthplace and local-level political experience, are able to increase vote returns due to their personal attributes serving as an important signal to voters that they are the candidates who understand local needs and will promote local interests (Shugart et al., Reference Shugart, Valdini and Suominen2005). To garner personal support from the constituents, candidates with strong local ties attempt to take advantage of such an expectation from voters during an electoral campaign. That is, they build strong support bases in the constituency by emphasizing that they will represent the local interests upon becoming elected.
The personal attributes of candidates may help them win elections in this way. Yet, the elected representatives with the help of their personal attributes are not totally free in parliament. Voters monitor and evaluate their representatives to decide whether to provide them with votes in the next election. Therefore, after they become elected, legislators have to behave in accordance with their attributes to maintain their electoral ground. In the case when legislators have PVEA, they will face many demands and expectations from their constituents on their time and attention to local interests in the post-election period, because their electoral advantage is derived from the informational cues that they understand the local needs of their constituency. In order to win reelection, they are required to ‘bring home the pork’ − government subsidies and public works − and satisfy local needs, as they promised during their electoral campaign. In other words, once they are elected, they are required to demonstrate that they care about local needs so as to maintain their support bases (and keep their electorally competitive edge). Hence, they aim to express themselves and establish a distinctive position in the political market through their individual activities in parliament.
For such legislators, drafting legislation and posing questions to cabinet ministers at committee meetings are significant opportunities in parliament to show their understanding of local needs and to separate themselves from other colleagues (see Hill and Hurley, Reference Hill and Hurley2002; Loewen et al., Reference Loewen, Koop, Settle and Fowler2014; Quinn et al., Reference Quinn, Monroe, Colaresi, Crespin and Radev2010). Through legislation and legislative speech in parliament, they are able to send credible signals to their constituents about what they stand for and what they can do to substantiate those claims. In other words, legislators can use these opportunities for advertisement and position taking, as well as credit claiming (Fiorina, Reference Fiorina1989; Mayhew, Reference Mayhew1974). Although many private members’ bills may not pass into law, legislators are able to attract media attention and enhance their public visibility by raising issues in the form of legislation. Furthermore, even if their bills are defeated in parliament, their proposals may have a chance of being incorporated into later government bills (Brazier and Fox, Reference Brazier and Fox2010; Oyama, Reference Oyama1997). More importantly, however, in addition to these benefits, legislators can inform their constituents that they are seriously concerned about local issues, regardless of whether their bills actually become laws (Hyson, Reference Hyson1974).Footnote 7 Similarly, the deliberations of committees may play only a marginal role in policy formulation in parliamentary democracies as the majority party controls the government. However, legislators are able to draw public attention by posing questions to cabinet ministers and government officials at committee meetings due to the fact that committee deliberations are not only broadcast publicly, but also widely reported in newspaper articles. Moreover, legislators can demonstrate their own policy positions and impress their constituents through their questions (Proksch and Slapin, Reference Proksch and Slapin2012; Soroka et al., Reference Soroka, Penner and Blidook2009).
While proposing private members’ bills is attractive to legislators, not all of them equally engage in legislation because it requires time and effort. For instance, in Japan, drafting a bill is time-consuming because it has to be presented in the form of a law with a clear structure of sections and articles. Hence, sending signals by proposing legislation is very costly. The problem they face is that time is a scarce resource for every legislator due to the high volume of competing demands on their time and attention. Legislators always have to sacrifice some of their activities because of the shortage of time (see Ingall and Crisp, Reference Ingall and Crisp2001). Similarly, for individual legislators, asking questions at committee meetings also takes more time and effort than merely attending the meetings because they have to prepare for interrogatory questions. Furthermore, the opportunities given to legislators to pose questions at committee meetings are limited because all members wish to participate in parliamentary debate to appeal to their local voters. Accordingly, legislators want to focus on specific areas that can effectively demonstrate that they represent the interests of their constituents in parliament. When legislators with PVEA initiate private members’ bills or pose questions, they should spend their limited resources on legislation and committee meetings that directly deal with particularistic pork-related issues, such as government subsidies and public works, in order to demonstrate their representation of local interests. This line of argument leads to the following hypothesis about the link between legislators’ personal attributes and their individual-level activities in parliament:
Hypothesis 1:Legislators with strong local ties are more extensively involved in parliamentary activities related to particularistic pork-barrel politics than legislators without such an attribute.
In elections, candidates with strong local ties are better able to cultivate personal votes than those without such attributes (Shugart et al., Reference Shugart, Valdini and Suominen2005; Tavits, Reference Tavits2010). Nonetheless, a significant number of candidates without such attributes win elections in electoral systems which feature strong intra-party competition, even though candidates with strong local ties are electorally advantageous. Local origin and prior local-level political experience are not necessarily the only attributes that make candidates distinct from others in the political market. Legislators are recruited from various sectors, and they have diverse attributes. One of the attributes that marks a stark difference from PVEA is prior experience of legal work. In many parliamentary democracies, the legal profession is also a major source of recruitment (De Winter, Reference De Winter, Blondel and Thiébault1991). In contrast to candidates with local ties, candidates with legal-work experience may not be able to earn personal votes solely by emphasizing their personal attributes because their experience cannot make them claim that they have a good understanding of local needs. However, they are able to take advantage of alternative skills in order to differentiate themselves from other candidates. If personal attributes of legislators affect their choice of parliamentary activities, such candidates should choose different parliamentary activities from those with PVEA.
Candidates with legal-work experience are able to emphasize that they will contribute to the pursuit of national programs and policies, such as foreign affairs and national security issues, through participating in general law-making activities. It is possible for them to send their constituents such a signal, different from that of legislators with local ties, because they understand the language of statutes and possess the capacity to discuss legislation technically. Therefore, in order to maintain their competitive electoral edge under candidate-centered electoral systems, legislators with prior legal work experience should particularly spend more time and effort on making national programs that will enhance their public image as legal experts than their colleagues who do not possess such an attribute.Footnote 8 This line of argument generates the following hypothesis:
Hypothesis 2:Legislators with legal skills are more extensively involved in parliamentary activities related to general public goods provisions than legislators without such an attribute.
3. Data and measurement
These theoretical arguments lead to the above two hypotheses, which provide the following observable implications respectively. First, Hypothesis 1 implies that, compared to other legislators, legislators with strong local ties, specifically those with prior local-level political experience, should introduce more private members’ bills to pork-related committees; however, it should not be the case for the bills referred to other committees (Hypothesis 1a). Furthermore, they should seek to more frequently pose questions at pork-related committees than other legislators, but they should not do so at other committees, especially at committees that deal with broad national programs, such as foreign affairs and national security issues (Hypothesis 1b). Second, Hypothesis 2 implies that, in contrast to the cases of legislators with local-level political experience, legislators with legal-work experience should introduce more private members’ bills − especially the bills referred to committees excluding pork-related committees − than other legislators (Hypothesis 2a). Moreover, they should also more frequently ask questions at committees related to national programs, compared to other legislators (Hypothesis 2b).
These hypotheses are tested through the use of data drawn from parliamentary activities among legislators in Japan between 1990 and 1996. This period covers two electoral cycles under the single non-transferable vote (SNTV) system, where candidates running from the same party compete against each other in their electoral districts. Under such an electoral system, candidates have incentives to cultivate personal votes because personal reputation is important for winning elections (Carey and Shugart, Reference Carey and Shugart1995; Cox and Thies, Reference Cox and Thies1998). Therefore, the Japanese case during this period provides an excellent case to test the hypotheses about parliamentary behavior of individual legislators under pressure to cultivate personal votes. One of the other advantages of using data from Japan during this period is that it provides a unique opportunity to examine the effect of legislators being in government and in opposition, because legislators during this time period experienced multiple rotations of political power during their terms in parliament; sometimes in governing parties, and at other times in opposition parties.
The Diet − the national parliament of Japan − convenes multiple sessions every year. In the period of study, 20 sessions (the 118th to the 137th session) were held in total.Footnote 9 The unit of analysis in the present study is individual legislators in each session. For each session, I collected information about all the legislators who had a seat in the Lower House of the Diet.Footnote 10 The total number of legislators in the dataset is 682.
Dependent variables
In this study, two types of parliamentary activities are employed as dependent variables to test the hypotheses: (1) initiations of private members’ bills and (2) participations in committee deliberation. Specifically, this study uses the frequencies of bill initiations and questions at committee meetings to measure how much effort each legislator puts into parliamentary activities. These activities are further classified into policy areas. The models are estimated separately across policy areas in order to examine how legislators behave differently from one another in parliament.
The variable on the initiations of private members’ bills is measured using the number of private member's bills, per session, that each legislator introduced to the parliament.Footnote 11 During the period of study, individual legislators, excluding those who simultaneously assumed any government positions, drafted 2.89 bills on average; and in total 125 bills were proposed to the parliament as private members’ bills.Footnote 12 The bills are usually referred to one of the committees with primary jurisdiction over their policy issues once they are introduced by legislators. In the Japanese Diet, around 20 standing committees have been created, directly corresponding to the government's ministry functions. These committees are relatively strong, partly because legislators design such institutions to cultivate personal votes (see Martin, Reference Martin2011a). However, committees are not equally useful to legislators in the pursuit of local constituency interests. In the Diet, the following three committees in particular deal with the government's spending projects: Agriculture, Construction, and Commerce (see Inoguchi and Iwai, Reference Inoguchi and Iwai1987). Thus, in order to examine whether legislators with different personal attributes concentrate their law-making resources on different policy areas (particularistic pork-barrel or general public goods), this study analyzes the initiations of private members’ bills separately between those that were referred to any one of these three committees and those that were referred to the other committees.
Another dependent variable − committee participation − is measured by the number of committee meetings that each legislator attended to ask questions during the session.Footnote 13 In parliament, party leaders control the access to policy-making positions, including committee assignment (see Epstein et al., Reference Epstein, Brady, Kawato and O’Halloran1997; Kam, Reference Kam2006). However, even when legislators are assigned to undesirable committees, they are often given opportunities to ask questions at their favorite committees by temporally substituting committee members. Instead of observing the committee membership of individual legislators, therefore, this study examines their involvement to committee deliberation. During the period of study, 3,360 committee meetings were held at the lower house of the Diet. In the sample, the average number of committee meetings that each legislator attended to pose questions in an ordinary session is 3.98.Footnote 14
In the Diet, standing committees correspond to the government's ministry functions, as mentioned above. While some committees primarily discuss issues related to the government's spending projects, other committees deal with national programs that will benefit society more broadly. To examine how legislators with different personal attributes participate differently in committee meetings, I separately analyze legislators’ participation in committee meetings by categorizing the committees into two groups depending on their jurisdictions over policy issues: (1) standing committees that deal with the government's spending projects (Agriculture, Constructions, and Commerce committees), and (2) other standing committees. In addition, I further analyze the participation in standing committees that mainly handle national policy issues − External Affairs and National Security committees.
Independent variables
The primary independent variables to test this study's hypotheses are the following personal attributes of legislators: local-level political experience and legal-work experience.Footnote 15 These are dummy variables where the value of one was assigned to the legislators who have prior experience of local-level politics—such as local assembly members, mayors, and governors—or legal professionals, respectively, and the value of zero was assigned to the others. The strength of local ties varies among legislators depending on whether they had previously served as a local politician. The level of legal knowledge also varies among legislators depending on whether they had worked as lawyers and legal professionals. In Japan, these professions are major sources of recruitment to the parliament. Of the legislators in the sample, 232 (34%) have local-level political experience and 45 (6.6%) have legal work experience before they joined the parliament. Moreover, they are almost mutually exclusive.Footnote 16 Legislators with such personal attributes are able to send important signals to their constituents about what they can do for them in parliament, and how they differ from others.
Having discussed this study's primary independent variables, I proceed to explain other independent variables used in the analyses. Individual legislators have various personal attributes and characteristics. Moreover, they face different political conditions. To thoroughly control for these factors that might also affect their parliamentary behavior, the following categories of variables are included in the empirical model: (1) political attributes, (2) political positions, (3) characteristics of electoral districts, (4) election results, and (5) biological attributes. There exist multiple variables under each category.
(1) Political attributes
First, legislators have different political attributes regarding their political experience (i.e., tenure in office) and party affiliations. Their individual-level activities may differ depending on the level of political experience, because they are at different stages in their political career. For instance, legislators should have more resources (e.g., time, money, staff members, and information) to engage in parliamentary activities as they gain more experience in national politics.Footnote 17 In contrast, junior legislators have to be more attentive to their constituents than do their senior colleagues due to their lack of sufficient name recognition. They especially need to spend their time on casework and constituency services in order to develop personalized support bases (Nemoto et al., Reference Nemoto, Krauss and Pekkanen2008). Accordingly, junior legislators spend less time in parliament (and more time in their electoral districts) compared to senior legislators (see Inoguchi and Iwai, Reference Inoguchi and Iwai1987). Political experience is measured using the number of terms a legislator has served as a member of parliament. Legislators with political experience, however, may not necessarily increase or decrease their engagement in parliamentary activities in a monotonic fashion. The effect of political experience is expected to be curvilinear rather than linear. In order to capture the diminishing effect of political experience, the empirical model also includes a squared variable − political experience 2.
Party affiliations also may affect the individual-level activities of legislators in parliament. Almost all legislators affiliate with any one of the established political parties to strengthen their bargaining power in the legislature (Aldrich, Reference Aldrich1995; Cox and McCubbins, Reference Cox and McCubbins1993). Political parties, however, have different levels of bargaining power and leverage at the bargaining table due to their size and status in the parliament. Therefore, political parties offer different opportunities to their members, such as question time at committee meetings and direct access to the government.Footnote 18 In order to control for such differences among legislators across parties, ruling party affiliation and party size are included in the model. The former is a dummy variable indicating whether or not a legislator is affiliated with a ruling party, and the latter is the share of seats that a legislator's party holds in the parliament.
(2) Political positions
Second, legislators play different roles in parliament because they have different positions and status in their party or in the parliament. Legislators with high-ranking positions, such as committee chairs and party leaders, may participate in committee meetings more frequently than do backbenchers. However, they simply engage in these activities as a part of their duties and do not necessarily seek local interests or other personal benefit through their engagement in parliamentary activities. To control for the difference between their behavior and rank-and-file members, the model includes the following dummy variables, which indicate whether or not they are high-ranking position holders: party leadership, committee chair, and committee director. They are coded as 1 if they hold such positions, respectively, and 0 otherwise.
(3) Characteristics of electoral districts
Third, legislators may exhibit different behavior from one another in parliament, because they represent different locations and demographics of population. In order to control for the characteristics of their electoral districts, the following three variables are included in the model: urban district, elderly population, and district magnitude. Urban district − the degree of urbanization of an electoral district − is captured by an urbanization index based on the Densely Inhabited Districts (DIDs), which has been complied by the Statistics Bureau of Japan. This index indicates the ratio of the population residing in urban areas with a high population density. This study controls for the population density characteristics of a legislator's electoral district, since there may be a significant difference between urban and rural politicians in their behavior in parliament. Elderly population is the share of senior people over 65 years old in a legislator's electoral district. The turnout rate of senior citizens tends to be higher than other cohort groups. Furthermore, retired senior citizens have different policy demands of their representatives, compared to the labor force population (Campbell, Reference Campbell2005; Plutzer, Reference Plutzer2002). District magnitude indicates the number of legislative seats assigned to the electoral district. This variable is included in the model because the importance of personal votes increases with district magnitudes (Carey and Shugart, Reference Carey and Shugart1995). Moreover, the value of information shortcuts is expected to increase with district magnitudes because voters have to choose one candidate among a pool of multiple candidates.Footnote 19
(4) Election results
Fourth, the electoral strength of legislators may affect their parliamentary behavior. Electorally vulnerable legislators need to improve their reputation and convince voters to support them (Kellermann, Reference Kellermann2013). Such legislators may want to focus on casework activities that directly appeal to their constituents (see Heitshusen et al., Reference Heitshusen, Young and Wood2005; Hirano, Reference Hirano2011), rather than disclosing their own personal policy positions that may diverge from the median voter (see Burden, Reference Burden2004). In order to control for the effect of electoral strength on parliamentary behavior, this study incorporates two variables that describe a legislator's election outcome: TK index and RS index. These variables, offered in the JED-M dataset, are comparable across electoral districts in Japan.Footnote 20 The former variable indicates the legislator's chance of winning. The latter variable indicates the regional balance of votes gathered by legislators at the municipal level in their electoral districts.Footnote 21
(5) Biological attributes
Finally, the empirical model also controls for the biological attributes of legislators, such as age and gender. Age is the legislator's age at the time of convening session. There exists a positive correlation between age and political experience. I control for age in the model so as to single out the impact of political experience, because the omission of age may bias the impact of political experience. The engagement of parliamentary activities might be different between male and female legislators (see Chattopadhyay and Duflo, Reference Chattopadhyay and Duflo2004; Dollar et al., Reference Dollar, Fisman and Gatti2001). Gender is coded 1 for male legislators and 0 for female legislators.Footnote 22 The coefficient will be positive if male legislators dominate parliamentary activities.
4. Analysis and findings
A series of empirical analyses are conducted using the following dependent variables: (1) initiations of private members’ bills and (2) participations in committee meetings. The two distinct types of parliamentary activities − legislation and committee participations − are not mutually connected; they are independent. Table 1 shows the correlation coefficients between these activities among legislators. These coefficients have very low values (0.191 at the highest). This implies that a legislator who is active in pork-related legislation may not be so salient in committee meetings dealing with pork issues, and vice versa. Therefore, it is important to analyze these activities separately in examining whether legislators with certain personal attributes exhibit a particular pattern of behavior in parliament. In order to test the hypotheses, it is inappropriate to use OLS regression models because dependent variables are bounded at 0. This study instead employs the negative binomial model, which is suitable for analyzing count data.
Table 1. Correlation coefficients between legislative activities and committee participations
Initiations of private members’ bills
First, I discuss the results regarding the engagement in parliamentary law-making activities among legislators. Table 2 shows the estimation results, in which the dependent variable is the number of private members’ bills initiated by each legislator. The models include a battery of session dummies to control for various factors in parliamentary sessions, such as the session length and changing nature of political conditions, which commonly affect every legislator in the same manner. The inclusion of the session dummies implies that the coefficient estimates of independent variables capture only a variation within each parliamentary session. However, the estimates of session dummies have been excluded from the table to save space. There are three types of models in Table 2. The dependent variable in Model 1 includes all private members’ bills initiated by a legislator. On the other hand, the dependent variable used in Model 2 and Model 3 is restricted to only those referred to pork-related committees and other committees, respectively. The results of these models are compared to test the observable implications (Hypothesis 1a and Hypothesis 2a) drawn from the hypotheses.
Table 2. The effect of legislators’ personal attributes on legislative activities
* p < 0.1 ** p < 0.05 *** p < 0.01.
Standard errors (clustered by each legislator) are in parentheses.
According to the results shown in Table 2, there exists a stark difference in the pattern of law-making activities between legislators with local-level political experience and legislators with legal-work experience. The results show that, in Model 2, the coefficient estimate of local-level political experience is positive and statistically significant; however, the coefficient estimate of legal-work experience is negative and not statistically significant in the full model (Model 2c). In contrast, and more importantly, in Model 3, the coefficient estimate of local-level political experience is negative and statistically significant; on the other hand, the coefficient estimate of legal-work experience is positive and statistically significant. These results suggest that legislators with strong local ties introduce more private members’ bills than other legislators in the field of pork-barrel politics, while doing the opposite in other fields.Footnote 23 In contrast, legislators with legal skills are prominent in general law-making activities that broadly serve the public interest. Thus, the data support Hypotheses 1a and 2a in terms of the effect of legislators’ personal attributes on their legislative activities in parliament.Footnote 24
Participation in committee meetings
Second, a similar difference is observed among legislators in their participations in committee meetings. Table 3 shows the results of empirical analyses using the number of participations in committee meetings as a dependent variable. Similar to Table 2, the models also include a battery of session dummies that control for various factors in parliamentary sessions; and their estimates have been excluded from the table to save space. In Table 3, there are four types of models. The dependent variable used in Model 4 includes all committee meetings that a legislator attended to pose any questions to cabinet ministers and government officials during the session. The committee meetings are further separated between the ones dealing with pork-related issues and others. They are used for Model 5 and Model 6, respectively. In addition, Table 3 further presents Model 7, in which the dependent variable exclusively includes committees that handle national policy programs.
Table 3. The effect of legislators’ personal attributes on committee participations
* p < 0.1 ** p < 0.05 *** p < 0.01.
Standard errors (clustered by each legislator) are in parentheses.
The results demonstrate that the coefficient estimate of local-level political experience in Model 5 is positive and statistically significant; however, the coefficient estimate of legal-work experience is not statistically significant. In contrast, in Model 6, while the coefficient estimate of local-level political experience is not statistically significant, the coefficient estimate of legal-work experience is positive and statistically significant. Furthermore, the coefficient estimates of these variables in Model 7 have mutually opposite signs with statistical significance.Footnote 25 These results suggest that legislators with strong local ties more frequently ask questions in the pork-related committees compared to other legislators; however, they are invisible in other committees, especially in committees dealing with national programs (i.e., they avoid posing questions at committee meetings that do not help them promote local interests). The opposite pattern is observed among legislators with legal skills. Their active involvement stands out in the latter committee meetings, which go beyond local interests. Hence, Hypotheses 1b and 2b about the effect of legislators’ personal attributes on their committee participations are supported by the data.
The empirical results so far have shown a clear relationship between legislators’ personal attributes and committee participations. However, some might argue that this is observed because their party leaders assign tasks to the legislators according to their previous experience, and legislators are just following their party leaders. In order to examine whether the personal attributes of legislators also affect parliamentary activities that are well beyond the control of party leaders, this study further employs the content of questions that individual legislators asked at committee meetings as a dependent variable. When legislators ask questions at committee meetings in parliament, they are able to formulate questions more or less at their discretion. Hence, their speech often reveals their personal interests and policy positions (see Hill and Hurley, Reference Hill and Hurley2002; Martin, Reference Martin2011b; Soroka et al., Reference Soroka, Penner and Blidook2009; Quinn et al., Reference Quinn, Monroe, Colaresi, Crespin and Radev2010).
The variable of content is captured by a dummy variable that describes whether a legislator referred to the local community in his/her oral questions during the parliamentary session when he/she was given the opportunities to participate in committee meetings. More specifically, the value of 1 is assigned to the legislators who mentioned the name of their local prefecture in their questions; otherwise, the value of 0 is given. When legislators mention the name of their local prefecture in their parliamentary speech, they may not always attempt to promote local interests; they may instead simply introduce themselves. However, they can at least present an impression that they represent their local constituents by mentioning the name of their home districts in their speech. Therefore, this variable can serve as a proxy for the intention of legislators to express their particular attention to local interests in their participation in committee meetings. I ran a logistic regression to examine the content of questions, because the dependent variable is a dummy variable; other variables in the model remain the same. Table 4 below shows the results.
Table 4. The effect of legislators’ personal attributes on the content of their speech
Notes: * p < 0.1 ** p < 0.05 *** p < 0.01.
Standard errors (clustered by each legislators) are in parentheses.
Despite the fact that this proxy may lack accuracy as a record of legislations’ local orientation, the results show that there still exists a significant difference in their pattern of behavior between legislators with strong local ties and legislators with legal skills. The coefficient estimate of local-level political experience is positive and statistically significant; and its value implies that, holding all other variables in the model constant, the odds of mentioning their local prefecture in their questions is 1.25 times greater for legislators with strong local ties than for those without such an attribute. In contrast, however, while the coefficient estimate of legislators with legal-work experience is positive, it is not statistically significant at the conventional level. This suggests that, having a prior experience of legal work does not necessarily increase or decrease the likelihood that legislators refer to local community in their speech, compared to other legislators. These results corroborate the observable implications drawn from this study's hypotheses.
In summary, while individual legislators have the authority to initiate their own bills and ask questions at committee meetings, they do not equally engage in these activities across policy areas. The results of empirical analyses show that there exists a significant difference in the pattern of behavior between legislators with strong local ties and legislators with legal skills, even after controlling for other personal attributes and political conditions, such as electoral strength and seniority. In the Diet, legislators with strong local ties initiate more private members’ bills and pose more questions at committees that deal with pork spending projects than do other legislators. By contrast, legislators with legal skills initiate more bills and pose more questions in other committees, compared to those without such an attribute. These results are consistent with this study's theoretical predictions that legislators with different personal attributes engage in parliamentary activities differently in accordance with their attributes. Moreover, the results are robust to a wide range of specifications and controls for member-specific factors as well as time-varying factors.
5. Conclusion
This study explored the individual-level activities of legislators in parliament, which have been largely ignored in the literature on parliamentary democracies. There exists a significant difference among legislators in their activities in parliament. Individual legislators are extensively involved in parliamentary activities, such as drafting private members’ bills and posing questions at committee meetings, even though these individual activities are not considered to be important for policy formulation in parliamentary democracies. Moreover, their engagement varies significantly. While representative-centered electoral systems tend to promote localism among legislators, not all legislators equally pursue local interests in parliament.
The difference in parliamentary activities among legislators under such systems is derived from their incentives to establish a distinctive position in the political market, whereby they need to cultivate personalized support in their constituency. This study has argued that individual legislators use parliamentary activities as an important means to send clear signals about their policy positions and policy-making ability to secure support from the constituents. In evaluating candidates, voters rely on the personal attributes of candidates; and candidates take advantage of this to succeed in elections. Once elected, legislators have to satisfy their constituents; therefore, they choose to spend their time and resources on parliamentary activities that will enhance the credibility of their personal attributes in order to maintain their electoral ground. Accordingly, there exists a systematic difference among legislators in their engagement in parliamentary activities. In other words, personal attributes motivate individual legislators to exhibit different behaviors in parliament.
The results of the empirical analyses using data from Japan support this argument. They demonstrate that the personal attributes of legislators affect the choice of their parliamentary activities. By leveraging the resources and opportunities they have, legislators choose to engage in more effective activities among the various means available for their career advancement. In the analyses, the individual-level activities in parliament are compared and contrasted between legislators with local-level political experience and legislators with legal-work experience. The findings confirm the expectations drawn from this study's argument: (1) legislators with strong local ties are more prominent in particularistic pork-related fields, which help them promote local interests, than other legislators; (2) in contrast, legislators with legal skills take a leading role in activities dealing with nationally oriented issues, rather than pork-related ones, to influence general policy-making in parliament.
The findings of this study enhance our understanding of the significant linkage between legislators’ personal attributes and their parliamentary activities. The individual-level analysis of parliamentary activities reveals the importance of paying attention to the personal attributes of legislators to understand the independent actions of individual legislators in parliament. Political parties have been considered very strong in parliamentary democracies because legislators in each party usually vote in a unified manner in parliament. However, individual legislators engage in various activities in parliament.Footnote 26 Legislators, who are recruited from various sectors not only represent different interests but also bring different knowledge and expertise to the legislature. Since their backgrounds and personal attributes affect their behavior in parliament, the composition of parliament can exert significant influence on politics and policy outcomes within parliament as well as within political parties.
The results of this study also have important implications for understanding the issue of delegation and accountability in parliament. In parliamentary democracies, cabinet ministers are mostly chosen from among legislators (Blondel, Reference Blondel, Blondel and Thiébault1991). Those cabinet ministers make policies on behalf of the government and oversee the implementation of policy as heads of administrative agencies (see Strøm, Reference Strøm, Strøm, Müller and Bergman2003). Thus, the personal attributes of cabinet ministers can have a significant effect not only on policy, but also on bureaucratic delegation, through the utilization of the knowledge and expertise they have gained for their own political activities. Although significant variation exists in the personal attributes of cabinet ministers among countries, cabinet minister posts in many parliamentary democracies are often assumed by experienced senior members with legal knowledge who were recruited from among legal professionals (see De Winter, Reference De Winter, Blondel and Thiébault1991). Modern democratic governments rely heavily on their bureaucracies, but there is a concern that powerful bureaucracies may limit the capacity of democratic political systems due to the lack of efficient mechanisms to provide accountability to the public. The appointment of such members to cabinet minister posts may be one of the solutions taken by political actors, as they are able to discuss the details of government policies drafted in legislation without completely delegating their task to bureaucrats in their agencies.Footnote 27 Taking personal attributes of individual members into account enriches research on parliamentary democracy.
About the author
Yoshikuni Ono is Associate Professor of political science at Tohoku University in Japan. He received his Ph.D. from the University of Michigan in 2010. His research focuses on the comparative study of political parties and legislative politics in parliamentary democracies. His recent work has appeared in American Journal of Political Science.
Appendix
The effect of legislators’ personal attributes on their involvement in parliamentary activities can be further explored by using the number of private members bills drafted by legislators across party lines as a dependent variable. In the Japanese Diet, some private members’ bills are jointly initiated across party lines. During the period of this study (1990−96), 125 bills were proposed to the Diet as private members’ bills, excluding the ones initiated by committee chairs. Among them, 35 bills were jointly drafted by legislators across party lines. Interestingly, 15 out of 35 bills (43% of such multi-party bills) passed the legislature and became law. This appendix shows the results of analyses using the initiations of these bills to the parliament.
The present study argues that legislators with strong local ties are more extensively involved in parliamentary activities that allow for the pursuit of local interests than others (Hypothesis 1), while legislators with legal skills, in contrast, are more active in parliamentary activities related to general public goods provisions compared to other legislators (Hypothesis 2). The multi-party bills drafted by legislators across party lines tend to be linked to general public goods provisions (see Bräuninger and Debus, Reference Bräuninger and Debus2009). Therefore, these hypotheses imply that legislators with local-level political experience should initiate fewer multi-party bills compared to other legislators without such an attribute (Hypothesis 1c); in contrast, legislators with legal-work experience should initiate more multi-party bills than others (Hypothesis 2c). To test these hypotheses, I ran a negative binomial regression, in which the dependent variable is the number of private members’ bills initiated by legislators across party lines. Other variables in the model remain the same as the ones shown in Table 2.
Table 5 shows the results. The coefficient estimate of local-political experience is negative and statistically significant. Its value suggests that legislators with strong local ties initiate 28% fewer multi-party bills compared to those without such an attribute. In contrast, the coefficient estimate of legal-work experience is positive and statistically significant. Its value indicates that legislators with legal skills initiate 56% more multi-party bills compared to other legislators. In sum, legislators with legal skills are more salient than those with strong local ties in their submissions of multi-party bills, which tend to have a feature of public goods provisions. These results support both hypotheses (Hypothesis 1c and Hypothesis 2c).
Table 5. The effect of legislators' personal attributes on legislative activities across party lines
