Content uploaded by Tevfik Murat Yildirim
Author content
All content in this area was uploaded by Tevfik Murat Yildirim on Feb 11, 2024
Content may be subject to copyright.
Electoral Studies 88 (2024) 102746
Available online 2 February 2024
0261-3794/© 2024 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Contents lists available at ScienceDirect
Electoral Studies
journal homepage: www.elsevier.com/locate/electstud
Notes on new data sets
The republic at the centennial: A new dataset of the entire population of
legislators and their parliamentary activities in Turkey
Tevfik Murat Yildirim ∗,1
University of Stavanger, Norway
ARTICLE INFO
Dataset link:www.tevfikmuratyildirim.com
Keywords:
Elections
Parties
Legislators
Turkey
Parliamentary activities
ABSTRACT
29 October 2023 marked the centennial of the Republic of Turkey. Over the past hundred years, the country has
undergone major institutional changes, experiencing periods of democracy and non-democracy, bicameralism
and unicameralism, parliamentary and presidential systems, and majoritarian and proportional electoral
systems. This unique range of changes provides excellent opportunities for studying various theoretically
significant research questions. This article introduces a novel dataset encompassing the entire population of
legislators and their parliamentary activities in Turkey over the last century. The dataset includes detailed
information on (i) election results by districts, (ii) the sociodemographic and political backgrounds of 12,500
legislators, (iii) the frequency with which legislators made speeches, asked questions and introduced private
members’ bills, and (iv) the content of these parliamentary activities (i.e., the policy area focus) from the period
of 1920–2023. I demonstrate the usefulness of the data through descriptive analyses in three areas, political
elites across space, political elites across time, and the patterns of policy prioritization in the parliament. I
conclude by suggesting areas of research for future studies.
1. Introduction
29 October 2023 marked the centennial of the Republic of Turkey.
Over the past century, there have been ebbs and flows of democracy
in the country, where sporadic periods of democratization were in-
terrupted by authoritarian political leaders and military interventions.
And the year 2023 represented yet another challenge for Turkey. The
country is now deeply polarized, and the AKP’s authoritarian rule
is under pressure due to economic downturns and social injustice.
Political competition is undermined by the lack of media freedom,
growing oppression, and systematic censorship (Esen and Gumuscu,
2021), and the question of what the country’s political future will look
like remains a subject for debate.
Looking back at the past century, the Grand National Assembly
of Turkey (GNAT) clearly appeared to be one of the most persistent
institutions in the country that survived various political crises and that
managed to generate significant popular support (Kalaycioglu,1980).
Established under difficult financial and logistical conditions in an
underdeveloped Ottoman town by what the imperial government called
a group of disobedient army officers and bureaucrats, the Assembly
has since then played a major role in creating generations of political
∗Correspondence to: University of Stavanger, Kjell Arholms gate 41, 4021 Stavanger, Norway.
E-mail address: murat.yildirim@uis.no.
1Associate professor of political science.
2Turkey is also intriguing when it comes to political parties. Empirical research finds that the traditional left–right divide in party politics is much less clear
in Turkish politics, relative to other country contexts (Aydogan and Slapin,2015).
leaders throughout the episodes of modern Turkish politics. Since its
creation, the GNAT has become home to several thousands of MPs,
whose detailed background information and parliamentary activities
are publicly available online.
This remarkable institution has received relatively scant attention
from scholars of comparative politics, despite providing excellent op-
portunities to study the relationship between political institutions and
elite behavior. In the past century, Turkey has switched between par-
liamentary and presidential systems, majoritarian and proportional
representation electoral systems, bicameralism and unicameralism, just
to name a few major institutional changes in the country’s history.
Furthermore, the country has experienced periods of democratization
interrupted by sudden authoritarian reversals (Marschall et al.,2016;
Bakiner,2023). As such, the case of Turkey presents valuable oppor-
tunities to study elite behavior under different institutional settings
within the same country context.2Despite this unusually rich over-
time variation in political institutions, however, there is a dearth
of empirical research on elite behavior in Turkey. Extant studies on
Turkish politics provided mostly descriptive accounts of political elites
(e.g., Demirel,1994), and focused almost exclusively on political par-
ties or party leaders at the expense of ordinary MPs. Studies that
https://doi.org/10.1016/j.electstud.2024.102746
Received 13 March 2023; Received in revised form 16 January 2024; Accepted 17 January 2024
Electoral Studies 88 (2024) 102746
2
T.M. Yildirim
went beyond providing descriptive accounts of elites covered rather
brief periods in Turkish parliamentary life (Yildirim,2020). Arguably,
the most comprehensive account of Turkish elites was provided by
Frederick W. Frey nearly six decades ago (Frey et al.,1965), which
naturally falls short of explaining the dynamics of political elites in
recent episodes of Turkish politics.
To address these gaps, I present a novel dataset encompassing
the entire population of legislators and their political activities in the
parliament from 1920 to 2023. This dataset combines information on
the sociodemographic and political backgrounds of MPs, details of MPs’
political activities (both counts and content), and electoral provinces.
To the best of my knowledge, this dataset stands as one of the most
extensive sources of data on political elites within a single-country
context. In the following sections, I briefly outline my data collection
efforts and then showcase the utility of the data through descriptive
analyses.
2. Data collection
The dataset of the GNAT includes three sets of information about
parliamentarians. The first set concerns the background characteristics
of individual politicians, including name and last name (as they appear
in parliamentary records), age, gender, place of birth, formal educa-
tion, past parliamentary experience, party affiliation and changes in
affiliation within and across electoral cycles, civil status (i.e., married
or not, and parental status), and foreign language skills. Using data on
past parliamentary experience and background characteristics, I also
classified the MPs in terms of ‘localness’ (whether the MP represents the
district in which she was born), ‘constituency loyalist’ (whether the MP
has ever represented more than one district in his or her parliamentary
career), and ‘party loyalist’ (whether the MP has ever switched parties).
The second set of variables covers legislators’ activities in the par-
liament over the past century. This dataset includes counts of all
legislative speeches, one-minute and response speeches, statements,
oral and written parliamentary questions, motions, and interpellations,
as well as private members’ bills across electoral cycles. I content-
coded the legislative speeches delivered between 1965 and 2011 using
the Comparative Agendas Project (CAP) coding scheme. Additionally, I
coded a subset of parliamentary questions, specifically those from the
period 1991–2003, based on their constituency focus (whether or not
they are ‘constituency-oriented’).
I scraped raw parliamentary activity data from the Grand Assem-
bly’s website. A team of six undergraduate and graduate students
manually addressed issues on incomplete or unidentifiable pages of
parliamentary activities. The background information of 12,500 non-
unique MPs from 27 electoral cycles was collected manually from
published parliamentary records. After intensive training, four of the
six students participated in content analytic coding of parliamentary
activities based on the CAP coding scheme.3Using the CAP scheme,
the coders assigned each parliamentary activity to one of the 21 policy
categories, such as the economy, healthcare, international affairs, and
education4.
The final set of variables pertains to regional characteristics in the
country. Electoral districts in Turkey vary greatly in district magni-
tude, population, and urban/rural dynamics, collectively influencing
3Due to space constraints, I refrain from discussing the coding
scheme of CAP. More information about the codebook can be found at
𝑤𝑤𝑤.𝑐𝑜𝑚𝑝𝑎𝑟𝑎𝑡𝑖𝑣𝑒𝑎𝑔𝑒𝑛𝑑 𝑎𝑠.𝑛𝑒𝑡∕𝑑𝑎𝑡𝑎𝑠𝑒𝑡𝑠𝑐𝑜𝑑 𝑒𝑏𝑜𝑜𝑘𝑠.
4The initial dataset release includes content-coded speeches from 1965
to 2011, totaling nearly 80,000 speeches, and over 9,000 parliamentary
questions from 1991 to 2003. These questions were also coded based on
whether they were constituency-oriented. Two research assistants determined
if the questions referred to the individual’s electoral district or activities and
incidents known to take place in the region (e.g., expos, disasters). Future
releases of the dataset will expand the time span of content-coded activities.
legislators’ policy focus and engagement in parliamentary activities. To
capture this, I collected data on the electoral provinces represented in
the parliament using official records published by the Turkish Statistical
Institute (TUIK). These data encompass information about election re-
sults by districts, population size, the share of urban and rural residents
in the region, and district magnitude. Unfortunately, some data, such
as that on urban/rural population, are not available before 1935. The
resulting dataset comprises 12,500 data points, with the unit of analysis
being individual MPs across electoral cycles. However, as discussed
below, other versions of the dataset (e.g., at the parliamentary activity
level) will be provided to allow for flexibility in aggregations. The list
of variables available in the dataset is presented in Table 1.
In the following section, I will demonstrate the utility of the dataset
through various illustrative cases, focusing on three key aspects: the
spatial and temporal variation in MPs’ characteristics, and the content
of their parliamentary activities. I chose to highlight these aspects as
they are arguably key strengths of the dataset. Importantly, the unique
over-time variation in political institutions (e.g., varying regime and
government types, and electoral systems) in the country offers scholars
of comparative politics a valuable opportunity to study the role of
institutions in shaping elite behavior. Additionally, the spatial variation
within the country in terms of elite characteristics, parliamentary be-
havior, and electoral outcomes further enhances the dataset’s value in
exploring the interrelations among various sociopolitical phenomena.
3. Illustrative cases
3.1. Political elites across space
Turkey has consistently grappled with regional inequalities. The
center-periphery cleavage has been a defining characteristic of Turk-
ish politics since the late Ottoman times (see Bakiner,2018 for an
overview). There are notable disparities in economic development
across the country’s regions. While the question of why such disparities
persist today goes beyond the scope of this study, it is crucial to ac-
knowledge that the substantial variation in the economic development
of regions has significantly influenced the background characteristics of
the political elites emerging from these areas. MPs from rural areas, for
instance, tend to have larger families, less education, and less political
experience, according to the dataset. Such disparities can have signif-
icant consequences for the political behavior of representatives. As an
illustration, the data indicate that education, political experience, and
family size are statistically associated with the number of legislative
speeches one makes in the parliament.
Fig. 1 illustrates the spatial distribution of legislative speeches and
private members’ bills for the period of 1920–2023. Participation in
legislative debates and proposing private members’ bills constitute the
core of lawmaking activities (Bäck and Debus,2016;Bowler,2010).
Engagement in such activities helps MPs establish close ties with the
regional constituency (Bräuninger et al.,2012;Ciftci and Yildirim,
2019), get renominated (Yildirim et al.,2019,2021), and increase
their media visibility (Tresch,2009). That is, the extent to which
MPs are active on the legislative floor has important consequences for
the substantive representation of constituency interests, as well as for
politicians’ career outcomes.
The top half of Fig. 1 illustrates the average number of legislative
speeches (per electoral cycle) made by MPs in each district between
1920 and 2023. The bottom half of the figure illustrates the average
number of private members’ bills (PMBs) by legislators in each district.
It is important to note that some of the most visible social cleavages
in the country are on the west–east axis, with Southeastern Anatolian
provinces being less developed. As seen in the figure, the average
number of PMBs is dwarfed by the average number of speeches, due
in part to the fact that legislative speeches are relatively free from the
party discipline and less costly, compared to PMBs.
Electoral Studies 88 (2024) 102746
3
T.M. Yildirim
Table 1
The list of variables available in the dataset.
Variable Definition Source Coverage
MP-level
Party name MP’s party affiliation GNAT 1920–2023
Party status The governing status of MP’s party; gov’t/opposition TUIK 1920–2023
Seniority Number of terms served in parliament GNAT 1920–2023
Birth place The province in which the MP was born GNAT 1920–2023
Localness Whether the MP was born in the regionashe represents GNAT 1920–2023
Same region Whether the MP always represented the same district; yes/no GNAT 1920–2023
Ottoman parliament Whether the MP previously served in the Ottoman Parliament GNAT 1920–2023
Party switcher Switched parties at least once; yes/no GNAT 1920–2023
Constituency Name of the regional constituency GNAT 1920–2023
Age Age of the MP at the time of election GNAT 1920–2023
Sex Sex of the MP; male/female GNAT 1920–2023
Civil status Civil-status of the MP; married/single GNAT 1920–2023
Number of children Number of children at the time of election GNAT 1920–2023
Formal education MP’s formal educational attainment; college/not GNAT 1920–2023
Foreign languages The foreign languages the MP can speak GNAT 1920–2023
Parliamentary activities Counts of private members’ bills, legislative speeches, GNAT 1920–2023
one-minute speeches, statements, oral and written questions
District-level
District magnitude The number of parliamentary seats allocated to the district TUIK 1920–2011
Vote share Parties’ vote share in the district TUIK 1950–2011
Voting records Total votes, invalid votes and voter turnout in the district TUIK 1950–2011
Urban population Number of people living in urban areas TUIK 1936–2011
Rural population Number of people living in rural areas TUIK 1936–2011
Activity-level
Content of legislative speech Includes CAP-coded legislative speeches Own 1965–2011
Period
Electoral system The electoral system used to elect the MPs TUIK 1950–2023
One-party rule 1 = multiparty elections are not allowed TUIK 1923–2023
Bicameralism 1 = the presence of lower and upper chambers TUIK 1923–2023
Military intervention 1 = Dissolution of parliament due to a military intervention TUIK 1923–2023
Democracy score The average ‘electoral democracy’ score during the term V-Dem 1923–2022
aMy definition of ‘localness’ differs from that of many others in that an MP is coded as ‘local’ if she was born either in
the province she represented or in one of the neighboring provinces. Although this practice required me to hand-code the
‘localness’ variable instead of relying on an automated coding procedure, I believe my approach minimizes measurement
error in the coding of ‘localness’. On numerous occasions, for instance, I found that the MP was born in a town that was
located on the border of or very close to a neighboring province, and coding this MP as having no local roots would be a
mischaracterization of reality.
Another significant observation immediately evident in the figure is
that MPs from Eastern/Southeastern Anatolian provinces make fewer
speeches and propose fewer private members’ bills than MPs elected
from other provinces. In addition to being relatively more underdevel-
oped parts of the country, these provinces are also known to be home
to the largest ethnic minority in the country, citizens of Kurdish origin.
The empirical observation that MPs elected from the Kurdish-populated
areas are far less active than their counterparts from other areas of
the country harms the prospects for the substantive representation of
citizens of Kurdish origin in the parliament.5For example, the dataset
shows that the parliamentary activities of MPs from southeastern re-
gions are more likely to be related to agriculture, health care, and
education, compared to MPs from other regions of the country. As
such, the underrepresentation of these policy areas in the legislative
agenda would surely have important consequences for the political
representation of interests.
3.2. Political elites across time
The data at our disposal allow us to explore longitudinal trends in
various important characteristics of legislators. Although panel data
analysis is fairly uncommon in the study of political elites, partly
5This finding lends support to studies documenting the various representa-
tional challenges facing ethnic minority MPs in the parliament. Past research
showed, for instance, that MPs from Kurdish ethnic minority parties were
more likely to voluntarily retire (i.e., do not seek re-election), and that their
parliamentary questions were less likely to be answered by ministers under
the AKP rule (Yildirim and Overby,2019).
due to lack of comparable data on individual units across time, the
GNAT dataset opens up possibilities for longitudinal and time series
research designs. It is possible, for example, to explore how the leg-
islative behavior of individual MPs changes across electoral cycles.
Such political leaders as Bulent Ecevit, Adnan Menderes, and Deniz
Baykal, to name a few, spent years in the parliament as a legislator
before climbing the career ladder to become ministers, party leaders
and prime ministers. Many other politicians, such as Kamer Genc and
Ismail Camas, served continuously in the parliament over decades,
providing ample opportunities for qualitative and quantitative analyses
of changes in political socialization among individual politicians.
Scholars interested in using the dataset also have the opportunity to
explore the effect of various macro-political phenomena on legislative
politics. In fact, the case of Turkey provides excellent opportunities
to examine the relationship between institutional incentives and elite
behavior. Over the past century, Turkey has experienced different forms
of government, major changes in its electoral system, and phases of
democratization and de-democratization (including multiple military
interventions). These major events have greatly affected the political
environment within which individual politicians, political parties and
voters operated. Changes in electoral systems, as an example, alter
the incentive structures facing politicians in drastic ways; politicians
elected through a closed-list proportional representation system tend
to focus less on developing personal vote-earning attributes, relative
to MPs elected through single-member plurality systems (Carey and
Shugart,1995).
Here I offer some simple exploratory analysis of over-time vari-
ation in politicians’ past parliamentary experience and ‘localness’ to
demonstrate the usefulness of the dataset. Fig. 2 shows the average
number of legislative terms served in the parliament, with the vertical
Electoral Studies 88 (2024) 102746
4
T.M. Yildirim
Fig. 1. Parliamentary activities by MPs across provinces.
reference lines on the 𝑥-axis signifying three major political events that
altered the dynamics of Turkish politics significantly: the introduction
of multiparty elections,6and the military interventions of May 1960
and September 1980. As seen in the figure, the first parliament (1920–
1923) was fairly inexperienced, with an average of 0.4 terms past
experience, though the average experience increased dramatically over
the next two decades. In the first years of the Republic, there were two
main clusters of legislators in the parliament. Most members of the ‘con-
stituent assembly’ in 1920 had no prior political experience, whereas
many members of the dissolved Ottoman Parliament in Istanbul were
given the right to serve in the newly established parliament.
It is also clear from Fig. 2 that the one-party regime in the first
two decades of the Republic largely relied on the same pool of political
elites, as indicated by the sharp and consistent increase in political
experience until the first multiparty elections held in 1946. With the
members of the Democratic Party led by Adnan Menderes replacing
CHP MPs in the parliament, the average political experience declined
sharply. A similar trend was observed in 1960 and 1980. Past political
experience increased gradually from 1950 to 1960, and from 1960 to
1980, though it never reached the experience levels achieved during
the one-party rule. In the aftermath of the 1960 and 1980 military
interventions, the political elites of the day were banned from politics
or forced to leave, which left behind a new generation of politicians.
6Although multiparty elections were first used in the 1946 general
elections, the electoral procedures adopted in these elections were highly
controversial and far from being free and fair. The first reference point in Fig. 2
indicates the first (relatively) free and fair general election, held in 1950.
All in all, the figure shows that these three major events shaped the
pool of political elites in important ways by replacing large segments
of the ruling political elites.
Fig. 3 illustrates the over-time variation in the percentage of MPs
who represented the geographical area they were born in, who are
commonly referred to as ‘locals’ in the literature (Tavits,2010). The
figure also illustrates the variation in ‘constituency loyals’ across time
— those who represented the same district in their parliamentary
career. Recent research maintains that having local roots greatly shapes
citizens’ perceptions toward MPs (Campbell et al.,2019). It is almost
natural to think that MPs elected from provinces they have very weak
connections with find themselves in a relatively disadvantaged position
in representing the policy interests of the regional constituency. Despite
this, however, ‘localness’ was far from being the norm in the first thirty
years of the republic. Nomination decisions in this period were made
exclusively by a very small group of elites leading the party, and it was
not uncommon for MPs to represent districts they are unfamiliar with
Demirel (1994).
As Fig. 3 suggests, representing different districts over the course of
one’s career has become a rare phenomenon in the past fifty years. The
percentage of local MPs increased from 50 during the one-party rule to
around 75 in the second half of the century, and rarely fell below 70 in
the past few decades. The percentage of locals decreased substantially
from the early 1920s to the early 1940s, going as low as 38 percent
in the second half of the 1930s. While I proceed with caution in inter-
preting this finding in the absence of additional analysis, one potential
explanation for this is that the ruling elites in the one-party period
may have sought to rely on the same pool of elites, as indicated by
soaring political experience levels shown in Fig. 2, by making frequent
Electoral Studies 88 (2024) 102746
5
T.M. Yildirim
Fig. 2. The average number of Legislative terms served in the Parliament (i.e., past parliamentary experience).
rotations in the allocation of districts. Izzet Ulvi Aykurt, for instance,
represented Afyonkarahisar, Eskisehir and Hakkari in less than two
decades.7Additional analyses (not shown here) indicate that localness
became a norm in the past six decades in areas where urbanization
remained relatively limited (i.e., districts where >50 percent of the
population live in rural areas). In more urban areas such as cities and
major towns, local MPs were less common in the past decades.
Finally, I turn to Fig. 4, which illustrates the over-time variation
in four parliamentary activity types — parliamentary questions, leg-
islative speeches, response speeches, and one-minute speeches. The
figure shows that there has been considerable variation in the total
number of legislative speeches and parliamentary questions across
electoral cycles. Part of this variation can be explained by the fact
that the length of electoral cycles in the country varied greatly due
to external shocks (e.g. military interventions) and intra-parliamentary
politics (e.g., early elections). Interestingly, parliamentary questions
and legislative speeches follow a similar trend between the period of
1960 and early 2000s, though legislative speeches outnumber parlia-
mentary questions. Interestingly, the sharpest increases in number of
speeches and questions were seen in periods where major structural
changes took place, such as (i) the early years of the Republic, (ii) when
multiparty elections were introduced, and (iii) the military regimes of
1960 and 1980 were replaced by democratically elected governments.
It is also noteworthy that new types of parliamentary activities were
introduced in the second half of the century, and a large number of
MPs increasingly used response speeches and one-minute speeches for
various purposes. The growing use of response speeches is indicative of
brawls associated with deeper political divisions in the parliament.
3.3. Policy agendas across time and space
The desire to examine such questions as which policy problems
enter the agenda and when, why parties and candidates differentially
emphasize issues in election campaigns, and how the media and po-
litical agendas interact led scholars in rather distinct fields of study
7It is useful to note that the distance between Afyonkarahisar and Hakkari
is nearly 1200 km, as the crow flies.
to create content-analytic datasets of parliamentary activities, public
budgets, party manifestos, newspaper stories and TV coverage (Petrocik
et al.,2003;Green-Pedersen,2010;Jennings et al.,2011;Vliegenthart
et al.,2016;Bulut,2017;Baumgartner et al.,2019;Heffington et al.,
2019;Bulut and Yildirim,2020). To join this growing body of research,
I categorized parliamentary activities into the policy area categories
provided by the Comparative Agendas Project (CAP).
The first release of the GNAT dataset includes nearly 80,000 legisla-
tive speeches from the period of 1965–2011, and nearly 10,000 parlia-
mentary questions from the period of 1991–2002 that were content-
coded based on the CAP coding scheme. Each parliamentary speech
and question was assigned to a main topic, such as macroeconomics,
and to a subtopic, such as unemployment, interest rates, or industrial
Policy (under the category of macroeconomics). Unlike the politician-
level dataset introduced above, where the unit of analysis is individual
politicians in each electoral cycle, the content-analytic dataset is in-
troduced in a different format to allow for flexibility. Introduced in
a speech-level format, the dataset allows users to aggregate the data
by individual members of parliament, subgroups of parliamentarians
(e.g., female MPs, senior MPs, MPs from specific regions), political
parties, or legislative years and electoral cycles. By so doing, the
dataset also opens up possibilities for in-depth qualitative analyses
of policy prioritization. For instance, users will be able to focus on
highly specific policy areas by filtering the data (e.g., ‘higher education’
under the category of ‘education’, or ‘foreign aid’ under the category of
‘international affairs’), and then identify the political actors involved in
the discussion of those issue areas for qualitative data collection.
The dataset reveals interesting patterns concerning issue prioritiza-
tion and agenda-setting at the individual-, party-, and legislature-levels.
However, below I focus only on aggregate-level variation in issue
priorities across time due to space constraints. Fig. 5 illustrates how
the content of legislative speeches in the assembly has varied over
time. More specifically, the figure illustrates legislative attention to
relatively salient five CAP policy areas, namely, macroeconomics, law
and order, defense, and government operations. The figure reveals that
while fluctuations in attention to most policy areas were quite common
between 1965 and 2011, the share of some of the most salient issues in
the political agenda has been relatively stable over time. Government
Electoral Studies 88 (2024) 102746
6
T.M. Yildirim
Fig. 3. The percentage of local MPs across electoral cycles.
Fig. 4. Parliamentary activities over time.
operations, macroeconomics, law and order and defense were some of
the most commonly discussed issues on the legislative floor during this
period. Notably, the share of issues related to government operations
has grown considerably over time, starting with the first half of the
1990s.
4. Conclusion
In this brief research note, I introduced the GNAT dataset and
sought to demonstrate the usefulness of the data through descrip-
tive analyses. Specifically, these analyses documented considerable
Electoral Studies 88 (2024) 102746
7
T.M. Yildirim
Fig. 5. The policy agenda in the Grand National Assembly,1965–2011.
variation in representatives’ backgrounds and their parliamentary ac-
tivities across time and space. Despite its comprehensive coverage,
there is still much ground for additional data collection to extend
the dataset. Voting records in the Turkish parliament, for example,
have been underutilized and would make a good addition to the
GNAT dataset. Additionally, since all parliamentary activities are dig-
itized and available from the Assembly’s website, scholars have the
opportunity to use large language models to classify parliamentary
activities in terms of sentiments and emotions, among other things.
Such projects might benefit greatly from the GNAT dataset’s variables
on sociodemographic and political background.
This novel dataset allows scholars to examine a wide range of
research questions on electoral studies, elite behavior, party politics,
and political representation, to name a few. Using original datasets,
recent research in Turkish politics explored various questions regarding
the substantive representation of women (Bektas and Issever-Ekinci,
2019), renomination and promotion in the party rank (Yildirim et al.,
2021), party switching (Kemahlıoğlu and Sayarı,2017). With this
dataset, scholars now have the opportunity to expand on these themes
by making use of the varying institutional contexts within the country
over the past hundred years.
The dataset might be appealing also to scholars of comparative
politics in several ways. Parts of the dataset are entirely compatible
with similar datasets from other countries, which provides opportuni-
ties for comparative research. As an example, a recent edited volume
brought together thirty country-chapters on the determinants of leg-
islative speechmaking, where scholars investigated the role of political
parties and institutions on MPs’ access to floor time (see Back et al.,
2021). Another recent project, the Comparative Legislators Database
(Göbel and Munzert,2022), covers detailed information on legislators’
background from ten countries across decades. Given that the GNAT
dataset includes variables that are similar to those of the above projects,
scholars interested in comparative analysis of elite behavior might find
the GNAT dataset useful.
CRediT authorship contribution statement
Tevfik Murat Yildirim: Conceptualization, Data curation, Inves-
tigation, Methodology, Project administration, Resources, Software,
Writing – original draft, Writing – review & editing.
Data availability
Dataset will be publicly available from www.tevfikmuratyildirim.
com.
Acknowledgments
I thank Emiliano Grossman and Zachary Dickson, as well as par-
ticipants at ECPR’s Standing Group Conference on Parliaments in Paris
(2022). I am grateful to Goker Bilgic, Mehmet Demirsoz, Syeda Ramsha
Gilani, Erman Gorgu, Irem Ozyer, Ali Senyurt, Teodor Syrstad, and Ilay
Yemisci for their research assistance.
References
Aydogan, A., Slapin, J.B., 2015. Left–right reversed: Parties and ideology in modern
Turkey. Party Politics 21 (4), 615–625.
Bäck, H., Debus, M., 2016. Political Parties, Parliaments and Legislative Speechmaking.
Springer.
Back, H., Debus, M., Fernandes, J.M., 2021. The Politics of Legislative Debates. Oxford
University Press.
Bakiner, O., 2018. A key to Turkish politics? The center–periphery framework revisited.
Turkish Stud. 19 (4), 503–522.
Bakiner, O., 2023. Only game in town? The persistence of competitive authoritarian
regimes in modern Turkey. Turkish Stud. 1–29.
Baumgartner, F.R., Breunig, C., Grossman, E., 2019. Comparative Policy Agendas:
Theory, Tools, Data. Oxford University Press.
Bektas, E., Issever-Ekinci, E., 2019. Who represents women in Turkey? An analysis of
gender difference in private bill sponsorship in the 2011–15 Turkish parliament.
Politics Gender 15 (4), 851–881.
Bowler, S., 2010. Private members’ bills in the UK parliament: Is there an ‘electoral
connection’? J. Legislative Stud. 16 (4), 476–494.
Bräuninger, T., Brunner, M., Däubler, T., 2012. Personal vote-seeking in flexible list
systems: How electoral incentives shape Belgian MPs’ bill initiation behaviour. Eur.
J. Political Res. 51 (5), 607–645.
Bulut, A.T., 2017. Measuring political agenda setting and representation in Turkey:
Introducing a new approach and data set. Party Politics 23 (6), 717–730.
Bulut, A.T., Yildirim, T.M., 2020. Political Stability, Democracy and Agenda Dynamics
in Turkey. Springer.
Campbell, R., Cowley, P., Vivyan, N., Wagner, M., 2019. Why friends and neighbors?
Explaining the electoral appeal of local roots. J. Politics 81 (3), 937–951.
Carey, J.M., Shugart, M.S., 1995. Incentives to cultivate a personal vote: A rank
ordering of electoral formulas. Elect. Stud. 14 (4), 417–439.
Electoral Studies 88 (2024) 102746
8
T.M. Yildirim
Ciftci, S., Yildirim, T.M., 2019. Hiding behind the party brand or currying fa-
vor with constituents: Why do representatives engage in different types of
constituency-oriented behavior? Party Politics 25 (3), 369–381.
Demirel, A., 1994. Birinci Meclis’ Te Muhalefet: İKinci Grup: İKinci Grup, vol. 44,
İletişim yayınları.
Esen, B., Gumuscu, S., 2021. Why did Turkish democracy collapse? A political economy
account of AKP’s authoritarianism. Party Politics 27 (6), 1075–1091.
Frey, F.W., et al., 1965. The Turkish Political Elite. Mit Press Cambridge, MA.
Göbel, S., Munzert, S., 2022. The comparative legislators database. Br. J. Political Sci.
52 (3), 1398–1408.
Green-Pedersen, C., 2010. Bringing parties into parliament the development of
parliamentary activities in Western Europe. Party Politics 16 (3), 347–369.
Heffington, C., Park, B.B., Williams, L.K., 2019. The ‘‘most important problem’’ dataset
(MIPD): A new dataset on American issue importance. Conflict Manag. Peace Sci.
36 (3), 312–335.
Jennings, W., Bevan, S., John, P., 2011. The agenda of British government: The speech
from the throne, 1911–2008. Political Stud. 59 (1), 74–98.
Kalaycioglu, E., 1980. Why legislatures persist in developing countries: The case of
Turkey. Legislative Stud. Q. 123–140.
Kemahlıoğlu, Ö., Sayarı, S., 2017. Defecting alone or splitting together? Individual and
collective party switching by legislators. Public Choice 171, 187–206.
Marschall, M., Aydogan, A., Bulut, A., 2016. Does housing create votes? Explaining the
electoral success of the AKP in Turkey. Elect. Stud. 42, 201–212.
Petrocik, J.R., Benoit, W.L., Hansen, G.J., 2003. Issue ownership and presidential
campaigning, 1952–2000. Political Sci. Q. 118 (4), 599–626.
Tavits, M., 2010. Effect of local ties on electoral success and parliamentary behaviour
the case of Estonia. Party Politics 16 (2), 215–235.
Tresch, A., 2009. Politicians in the media: Determinants of legislators’ presence and
prominence in Swiss newspapers. Int. J. Press/Politics 14 (1), 67–90.
Vliegenthart, R., Walgrave, S., Baumgartner, F.R., Bevan, S., Breunig, C., Brouard, S.,
Bonafont, L.C., Grossman, E., Jennings, W., Mortensen, P.B., et al., 2016. Do the
media set the parliamentary agenda? A comparative study in seven countries. Eur.
J. Political Res. 55 (2), 283–301.
Yildirim, T.M., 2020. Politics of constituency representation and legislative ambition
under the glare of camera lights. Legislative Stud. Q. 45 (1), 101–130.
Yildirim, T.M., Kocapınar, G., Ecevit, Y.A., 2019. Staying active and focused? The effect
of parliamentary performance on candidate renomination and promotion. Party
Politics.
Yildirim, T.M., Kocapınar, G., Ecevit, Y.A., 2021. Status incongruity and backlash
against female legislators: How legislative speechmaking benefits men, but harms
women. Political Res. Q. 74 (1), 35–45.
Yildirim, T.M., Overby, L.M., 2019. Abandoning Ankara: Voluntary departures from the
grand national assembly of Turkey. J. Legislative Stud. 25 (4), 597–613.
Tevfik Murat Yildirim is an associate professor of political science at the University
of Stavanger, Norway. His work focuses on political elites and representation, political
behavior, and gender politics, and has appeared in such journals as American Political
Science Review, British Journal of Political Science, European Journal of Political
Research, Journal of European Public Policy, Legislative Studies Quarterly, Party
Politics, Political Behavior, among others. He can be reached at murat.yildirim@uis.no.
Address: Kjell Arholms gate 41, 4021, Stavanger, Norway.
