Nelda

Overview

Note: This website and dataset remain under construction.

The National Elections across Democracy and Autocracy (NELDA) dataset provides detailed information on all election events from 1960-2006. To be included, elections must be for a national executive figure, such as a president, or for a national legislative body, such as a parliament, legislature, constituent assembly, or other directly elected representative bodies. In order for an election to be included, voters must directly elect the person or persons appearing on the ballot to the national post in question. Voting must also be direct, or "by the people" in the sense that mass voting takes place. That voting is "by the people" does not imply anything about the extent of the franchise: some regimes may construe this to mean a small portion of the population. However, when voting takes place by committee, institution or a coterie, the "election" is not included. By-elections are not counted as elections for the purpose of this project, unless they take the form of midterm elections occurring within a pre-established schedule. In federal systems, only elections to national-level bodies are included. Cases in which any portion of the seats in a national legislative body are filled through voting are included.

Beyond these basic requirements, elections may or may not be competitive, and may have any number of other ostensible flaws. In fact, this last feature of the dataset is what separates NELDA most clearly from other available datasets on elections.

The unit of observation is the election round. All rounds of an election are coded, regardless of the number of seats remaining to be filled. In the 1994 Ukrainian legislative elections, six rounds were held, with fewer than 10% of the legislative seats up for grabs after 4 rounds of balloting- all six rounds appear in the dataset.

When deciding what counts as a new election, and what is a follow-up round to an election already underway, we ask (1) whether the regime calls the election a new one; (2) whether candidates are allowed and/or required to register again. Positive answers imply a new election. By this rule, the rerun of the second round of the Ukrainian Presidential election of 2004 counts as a third round: the set of continuing candidates was based on rounds one and two. In contrast, when the results of the November 2, 2003 legislative election in Georgia were cancelled, the legislative election of March of the following year is coded as a new election: new registration lists and candidates were allowed.

Indirect elections are not included. Thus, in the 1983 presidential election in the Arab Republic of Yemen, the President was re-elected for a second 5-year term at a meeting of the Constituent People's Assembly - causing us to drop this election. We do not include Chinese legislative elections because the people do not directly vote on deputies. However, presidential elections which involve an electoral college such as those in the U.S. and South Korea are included because the electoral college mechanically implements the outcome of a popular vote. An example of a special case is the 1970 presidential election in Chile, where the legislature elected Salvador Allende to the presidency after he won a plurality of the popular vote. The Chilean constitution required that if no candidate won a majority of the popular vote, the legislature would select between the top two vote-getting candidates, and tradition dictated that the legislature would select the candidate who received the largest number of votes. The 1970 Chilean election is included because a popular vote took place, and the indirect election within the legislature was determined by the popular vote. Another borderline case is Kenya in the 1970s, where voters cast a ballot for a deputy to parliament knowing that each deputy supports a particular presidential candidate, and that the presidential candidate supported by a majority of elected parliamentarians would be confirmed as president. We counted this peculiar system as a direct election of both the President and the legislature, and we code the two as two separate events (in part, to distinguish the different choices voters may have had with respect to the lists of parliamentary and presidential candidates.)

Most referenda are not included as elections, with one important exception. Some referenda on continued rule are functionally equivalent to presidential elections in single-party regimes, which are included in this study. Therefore, we include referenda when they are direct votes on candidates, most commonly referenda that ask voters whether they would like the incumbent to continue in office or not. If any referendum is a direct vote on the incumbent candidate's continued rule, it is included as an election. The 1988 referendum on the continued rule of Augusto Pinochet, therefore, counts as an election. Referenda that extend a leader's term in office but that are not leader-specific are not included as national elections. For example, we do not include constitutional referenda that change the length of term limits or that make some parties illegal. Thus, although there was a referendum on a new constitution in Equatorial Guinea on August 15, 1982 that also provided for an extension of the President's term in office, the event was not counted as a direct election in the NELDA data.

Sometimes, elections are cancelled immediately before, during or after election day. We include elections if and only if voting on election day has, in fact, commenced. This decision rule holds regardless of whether the balloting was not completed or was eventually cancelled, whether the results were never announced, or whether there were no "consequences" from the voting (such as a power succession).

At present, all independent countries with a population above half a million are covered, with the exception of Western countries (this excludes almost exactly the set of members of the OECD), from 1960 to 2006. See the countries covered section below.

Contact and Data Access

A subset of the data, including electionIDs and basic attributes of each election event are available in a beta version.

For questions, please contact either Susan Hyde (susan.hyde at yale.edu) or Nikolay Marinov (nikolay.marinov at yale.edu).

Variables

In addition to basic attributes, there are 58 variables coded for each electionid. For all but three of the 58 variables, there are only four possible values: "yes", "no", "N/A", "unclear." Answering "unclear" means there is not enough information to answer the question, either because sources are unavailable, sparse, or the available information seems to conflict in ways thatundermine both "yes" and "no" answers. "N/A" represents not applicable, and is used when the question does not make sense given the context of the election event, such as asking whether the incumbent won when the incumbent did not run. The 58 variables are labeled numerically as "NELDA1", "NELDA2", "NELDA2", and so on. For each variable, there is another associated variable allowing a free-text note clarifying the coding decision, with the note clarifying the variable NELDA1 named "NELDAn1".

Only three variables allow for free-text answers: NELDA43, NELDA44, and NELDA54.

In multiple round elections, results from the very last round of that election are never coded in an previous round's variables. The variables coded for each election round should reflect only the status of the election at the point of that particular round. For example: NELDA39, "Was the incumbent replaced?" should reflect the status of the incumbent at the end of that particular round, not whether the incumbent was replaced in subsequent rounds. Thus, the answer is usually "no" until last round of the election unless the incumbent has been formally succeeded by someone else at the conclusion of a non-final round. Some attributes will not change between rounds and will be the same for all rounds. For example, in elections in which the government harasses the opposition, this variable is unlikely to change significantly between rounds, but the coding rules allow for this possibility.

ccode: ccode refers to the three digit country code assigned by the Correlates of War (COW) dataset

cname: cname gives the name of the country

electionid: Unique for every observation in the dataset, electionid is the 16-character variable that identifies each election event by country, date, type, and round. It has four subcomponents, separated by a dash (-), defined as XXX-YYYY-MMDD-TR, where "XXX" represents the COW three digit country code, "YYYY" represents the year in which voting started for the entire election event, including all rounds, "MMDD" represents the month and day that the election event started, and does not change for multiple round elections. Note that for election events with two or more rounds, the MMDD and year remain the same as those of the first round. "TR" denotes a letter-number combination, with T representing the type of national office being elected, and "R" represents the round number, with "1" representing the first round, "2" representing the second round, etc. The letter 'P' stands for presidential elections, 'L' stands for legislative or parliamentary elections, and 'A' stands for constituent assembly elections.

For example, for electionids 100-1990-0311-A1 and 100-1990-0311-A2, the number 100 identifies the country as Colombia, the first day of the election event was March 11, 1990, the 'A' indicates that it was an election for constitutent assembly, and the final digit indicates that it had two rounds. The day of voting for the first round took place on March 11, 1990, but note that the date of the second round cannot be extracted from the electionid variable (see the variable mmdd below).

In election events spanning multiple days, we take the first day of voting as the month and day for the electionid. For example, the 1985 election in Burma began on October 6 and ended on October 20. The electionid as 775-1985-1006-L1, indicating that the first day of voting was October 6th.

Note that data can be collapsed in various ways using the electionid variable, depending on which format is most appropriate for the research question. Common formats may include combining same-day election events into one observation, combining multiple rounds into one election event, or collapsing the data into country-year format.

year: year indicates the year in which the election round took place

type: type codes the type of election that took place. All elections in this dataset are for national offices, including Presidential, Legislative, and Parliamentary (the last two are often referred to as Legislative), and Constituent Assembly. Some researchers may want to combine same-day legislative and executive elections, which are referred to as General Elections. However, in the base dataset, even when executive and legislative elections occur on the same date, they are coded as two separate events, allowing variation in coding for specific national offices.

mmdd: mmdd provides the month and day on which each election round took place. For first round elections mmdd and year are redundant to the information provided in the electionid. For elections with two or more rounds, year and mmdd provide the actual date that voting starting for the round of the election.

notes: notes is a free text field allowing for general remarks on the election being coded. Examples of such notes include "massive human rights abuses" and "first election in 20 years."

NELDA1: Were regular elections suspended before this election? Details...

In democracies, elections take place at regular intervals or within a specified period of time. This question is specifically asking if regular elections were suspended preceding the election at hand, not if elections have ever been suspended. Therefore, if the previous round of elections had been suspended, or sometime after the last round an announcement was made that elections would not be held pending further notice, the answer to NELDA1 would be yes. Even if a regime disbands the elected legislature and says it is paving the way for fresh elections, the answer would be yes unless they follow through on the promise in a relatively short period of time (typically less than a year). Whether elections had been suspended one or more times in a country's history does not lead to a yes answer. What counts is the immediate past, i.e., the aftermath of the last balloting. It does not matter whether elections are held somewhat earlier or later than usual for this question, nor whether there was a regime change that affected the timing of the elections, nor does it matter whether the freedom of elections changed. Suspending elections typically means that the legislative body is also disbanded, and that the regime legislates by decree. If a regime had never had elections, the answer is yes.

Examples of Yes:
Dominican Republic - 42-1962-1220-L1-- Following the death of the authoritarian leader, a transitional leadership was established to run elections.
Niger - 436-1989-1212-L1 and 436-1989-1212-P1; Following a 1974 military coup, Niger held its first legislative and presidential elections in 1989.
Comoros - 581-1990-0304-P1 and P2 581-1992-1122-L1 and L2; These were the first contested elections in Comoros. It previously had a one party state.

Examples of No:
India 750-1991-0520-L1, L2, and L3; India has held elections since independence from Britain in 1950.
Zambia - 551-1991-1031-L1 and P1; The 1991 elections marked Zambia's transition to multiparty democracy. However "No" was coded because single-party elections had been held in Zambia since 1972. While the 1991 elections were early, previous elections had not been suspended.

Special Cases:
India - 750-1991-0520-L4; India had suspended elections in one state in this round. However, since elections had not been suspended nationally, the case was coded as no.

NELDA2: Were these the first multiparty elections? Details...

This indicates when countries are newly independent and are having their first elections, when countries hold the first multiparty elections after a significant period of non-democratic rule, or when countries transition from single-party elections to multi-party elections. Multiparty means that more than one party is allowed to contest the election, and that at least some of the parties are both nominally and effectively independent of the ruling actors.

Examples of Yes:
Fiji-950-1992-0523-L1; In 1992, Fiji held its first legislative elections since a military coup in 1987.
Dominican Republic - 42-1962-1220-L1; An authoritarian regime was previously in place. After the death of the leader, multi-party elections with opposition were allowed.
Egypt - 651-2005-0907-P1; The President initiated reforms and allowed multi-candidate elections that were decided by the people as opposed to solely by referendum as in the past.

Examples of No:
Albania - 339-1974-1006-L1; Only one party, the Communist Democratic Front party, was legal, and thus these were not the first multiparty elections.

Special Cases:
Mongolia - 712-1990-0729-L1 and 712-1993-0606-P1; Mongolia held its first multi-party legislative elections in 1990, but its first multi-party presidential elections did not take place until 1993. Both cases are coded as "yes."

NELDA3: Was opposition allowed? Details...

This variable indicates whether at least one opposition political party existed to contest the election. Some countries have multiple government parties but no opposition political party. An opposition party is one that is not in the government, meaning it is not affiliated with the incumbent party in power. Note that if No is coded, all other opposition questions (NELDA13, NELDA14, and NELDA15) should be coded "N/A."

Examples of Yes:
Egypt - 651-2005-0907-P1; The President initiated reforms and allowed opposition contestation; whereas past elections were uncontested and just reaffirmed the legislative assembly's choice for president by referendum, many candidates could run in this election.
Dominican Republic - 42-1962-1220-L1; Following the death of an authoritarian leader, opposition parties were allowed to contest the election.

Examples of No:
Syria - 652-1990-0522-L1; Opposition was not allowed and candidates were only permitted to run through parties associated with the National Progressive Front.
Laos - 812-1992-1220-L1; No opposition was allowed in the one-party state, and many opposition leaders fled into voluntary exile.
Vietnam - 816-1992-0719-L1; Only one party was officially allowed, although a few "independent" candidates were permitted to run.

Niger - 436-1965-0930-P1, 436-1965-1021-L1, 436-1970-1001-P1; In these elections, only the military-backed incumbent party was considered legal even though an opposition party was organized.

Tunisia - 616-1994-0320-P1; There were no candidates other than incumbent Zine al-Abidine ben Ali. The other parties could not meet the required threshold for entry, which included the support of 30 mayors. The police effectively stopped opposition parties from meeting the legal requirements to participate in the election.

Special Cases:
Iran - 630-1997-0523-P1; In cases of limited opposition party competition, but where some opposition parties are allowed to compete, the variable is coded as "yes". In this case, other political parties existed and some were allowed to run; however, it is worth noting that candidates for any election have to be approved by the country's Council of Guardians before they are allowed to run, and the Council usually only approves a very small fraction of applicants.
Portugal - 235-1958-0608-P1; An independent candidate affiliated with the political opposition movement ran against the dictatorial regime's chosen candidate. However, opposition parties were offically banned. Thus, though there was an oppositon movement, partially underground, sometimes tolerated by the regime, "No" was coded because all opposition parties were banned.

NELDA4: Was more than one party legal? Details...

This variable indicates whether multiple political parties were technically legal. The legalization of multiple parties need not necessarily mean the existence of a functioning opposition party, as there may be other non-legal barriers to the development of an opposition party. Similarly, a well organized opposition party may exist but may not be legal.

Examples of Yes:
Mexico - 70-1994-0821-P1; There were several legal parties in addition to the PRI, including the PAN.
German Democratic Republic - 265-1971-1114-L1; There was no effective political opposition in the communist country, but some small "alliance" parties were permitted. They did not compete with the ruling Communist Party.

Examples of No:
Haiti - 41-1979-0211-L1; Parti de l'Unite' Nationale was the sole legal party in Haiti.
Sudan - 625-1996-0306-L1; Political parties were officially banned.

NELDA5: Was there a choice of candidates on the ballot? Details...

This variable indicates whether the voters were allowed to make a choice between candidates on the ballot, which is possible when the number of candidates competing for a slot exceeds the number of slots to be filled.

Examples of Yes:
Guatemala - 90-1966-0306-P1; Several candidates competed in the election to transfer power to a civilian president.

Examples of No:
Egypt - 651-1993-1004-P1; Voters were only permitted to vote for or against the individual nominated by the legislative assembly for president.
Niger - 436-1965-0930-P1, 436-1965-1021-L1, 436-1970-1001-P1; In these elections, only the military-backed incumbent party was considered legal even though an opposition party existed. The leader of the incumbent party was the only presidential candidate on the ballot.

NELDA6: If regular, were these elections early or late relative to the date they were supposed to be held per established procedure? Details...

If elections are regular, this variable indicates whether they were early or late compared to the date they were originally scheduled for. "Yes" means that elections were either delayed, or took place earlier than scheduled. "No" means that elections took place according to their scheduled date. "N/A" codes for elections which are not regular in that domestic political actors have no shared expectation about when elections will be held. If NELDA1 is coded as "yes" then NELDA6 must be coded as an "N/A."

Examples of Yes:
Fiji - 950-1994-0218-L1; Fiji's 1994 legislative elections were held three years earlier than scheduled.
Haiti - 41-1987-1129-L1; With the fall of the Duvalier regime, elections happened earlier than usual in order to establish a government.
Turkey - 640-1995-1224-L1; The 1995 elections were held one year early because the ruling coalition had collapsed and a new government needed to be formed.
India - 750-1991-0520-L2; The election was held several days after it was originally scheduled because the leader of one of the major parties was assassinated.
Comoros - 581-1993-1212-L1, 581-1993-1212-L2; The President delayed elections, ostensibly because of financial and logistical reasons. However, international commentators said it was a ploy to give the government more time to harass the opposition.

Examples of No:
Bulgaria - 355-2001-0617-L1; This parliamentary election took place 4 years after the previous parliamentary election, and its timing was consistent with the widely-known constitutional requirements. "

Special Cases:
Examples of "N/A":
Zambia - 551-1991-1031-L1 and P1; These were election held during the transition to multiparty democracy. NELDA1 had been coded as a "No," noting that elections had been regularly held before 1991. However, this questionwas coded as an "N/A" because the 1991 transition elections themselves were not regular. NELDA1 need not always be "Yes" to produce an "N/A" in NELDA6.
East Timor - 860-2002-0414-P1; These were the first elections in the newly independent state.
Honduras - 91-1971-0328-L1, 91-1971-0328-P1, 91-1980-0420-A1; These were the first elections of a new regime.
South Africa: 560-1984-0510-L1; Under the apartheid regime, elections were held separately for non-whites for set-aside seats in the legislature. Therefore, it is not a regular election, rather a special election for a subset of the population. "

NELDA7: Before elections, were there clear indications that the incumbent had made a prior decision to give up power? Details...

If, prior to an election that involves the incumbent, the incumbent signals that he or she will be giving up power by not running or stepping down, then a "yes" is coded - this includes cases where the leader abides by a constitutional ban on further terms. If the incumbent declares his or her intention to run for re-election, the answer will be "no." "N/A" is coded for elections that do not involve contests for the chief executive (where NELDA20 is coded "no."), but otherwise this question should be answered "yes" or "no."

Examples of Yes:
Bangladesh - 771-1991-0227-L1; The elections were held and administered by a neutral interim government. The interim Prime Minister banned all members of the interim leadership from running in the elections. The ousted leader of the previous military government ran in five districts. Because the incumbent is considered the leader of the government immediately preceding and during the elections, and the interim prime minister formally banned members of the interim government, including himself, from running, the variable was coded as "yes."

Examples of No:
Fiji - 950-1992-0523-L1; Although Fiji held legislative elections for the first time after the 1987 military coup, the military leader did not step down, and ran as a candidate with the intent of regaining office through the elections. "

Special Note:
For some elections, there is some ambiguity about who should be considered the incumbent leader for the purposes of this dataset. One special example is Iran, where presidential elections take place but the Ayatollah is considered the incumbent executive leader, as defined by the Archigos dataset (see NELDA43 and NELDA44 below). Although the Iranian presidential elections are included in the NELDA data, the Ayatollah is used in reference to the incumbent questions.
Another example of this ambiguity is Panama during General Manual Noriega's regime. In 1989, the Noriega government allowed elections for the presidency while keeping de facto presidential powers (95-1989-0507-P1).

NELDA8: Did the incumbent reach their term limit? Details...

This variable indicates whether the incumbent executive had reached the end of their constitutional or legal term in office. In theory, once these leaders have reached the maximum number of terms allowed in office, they are no longer eligible to run and must step down. A "yes" is coded when the incumbent reaches their term limit. If the executive is allowed multiple terms in office and reaches the end of one of them, the answer is "no". Parliamentary systems where the leader is a member of the parliament typically allow unlimited terms in office, and are coded as "no". If the position of the leader is not contested (NELDA20 is "no"), this question is coded "N/A," but is otherwise coded on a "yes"/"no" basis.

Examples of Yes:
Iran - 630-1997-0523-P1; A "yes" was coded for the 1997 presidential election in Iran because President Rafsanjani had already been in power for two terms and was barred by the constitution from seeking a third term.

Guatemala - 90-1970-0301-P1; Under the existing Guatemalan constitution, presidents were limited to one four-year term, so the incumbent president was ineligible to run for re-election.

Examples of No:
Belarus - 370-2001-0909-P1; According to the 1994 Constitution, Lukashenko's term should have ended in 1999. However, in 1996 Lukashenko pushed through a new constitution which stated that his term would not expire until 2001, and so he had not reached his term limit.

NELDA9: Had the incumbent extended his or her term in office or eligibility to run in elections at any point in the past? Details...

The variable is coded as "yes" if research indicates that the incumbent extended his or her term in office or eligibility to run again in an election (see NELDA43 on definition of the incumbent). A "no" indicates that the incumbent has not extended his or her term in office, while a "N/A" is coded when elections do not involve the executive office. For systems which impose no limit on the number of terms in office (e.g., most parliamentary systems), or that do not limit eligibility to run, the answer is "no." If the position of the leader is not contested (NELDA20 is "no"), this question is coded "N/A," but is otherwise coded on a "yes"/"no" basis

Examples of Yes:
- 616-2004-1024-P1; Zine al-Abidine Ben Ali won presidential elections after he had amended the constitution such that he could run for an additional term.

Examples of No:
Cyprus - 352-1993-0207-P2; the incumbent had not extended his term in office or eligibility to run in elections

NELDA10: Was the country ruled by "transitional leadership" tasked with "holding elections?" Details...

Yes indicates that an officially "transitional" leadership was appointed in order to organize and hold elections. An existing government cannot become "transitional leadership" simply by proclaiming itself to be transitional. Rather, major political actors must share the expectation that the main reason the interim government is appointed is to organize and hold an election.

Examples of Yes:
Honduras - 91-1981-1129-L1 and 91-1981-1129-P1; Prior to these elections, Honduras had a provisional Constituent Assembly that wrote the new constitution and a provisional president appointed by the assembly to organize elections and govern the country until elections were held.

Dominican Republic - 42-1962-1220-P1; With the fall of the previous authoritarian regime, a transitional leadership was established and tasked with holding elections.

Ecuador -130-1968-0602-P1; The military had overthrown the president in 1961 and was managing a transition to civilian rule following a national referendum on whether elections should be held."

Examples of No:
Costa Rica - 094-1994-0206-P1; these elections were regular and the country was not ruled by a transitional leadership.

Special Cases:
Greece - 350-1964-0216-L1; "Yes" was coded, though officially Greece was ruled by a caretaker government, not a transitional government. In the Greek case, the government would be taken over by a non-political caretaker governmnet charged with preparing the nation for the election three months before the election-date. Though offically the leader of Greece was caretaker-head Parakevopoulos, a banker, the election race was viewed by all Greeks as a race between the incumbent party Centre Union and the opposition party National Radical Union and its allies. Thus, aside from this question (NELDA10) and the question regarding who was the leader in office before the election (NELDA43), the caretaker government is not coded as the incumbent. Rather, the incumbent related questions were all answered with respect to the Centre Union and its head, Panpandreou.

NELDA11: Before elections, are there significant concerns that elections will not be free and fair? Details...

A "yes" indicates that there was evidence of domestic or international concern that the election process was not going to be free or fair. A "yes" is also coded when the elections were widely perceived to lack basic criteria for competitive elections, such as more than one political party.

Examples of Yes:
Fiji - 950-1992-0523-L1; In the 1992 legislative election, there were widespread domestic concerns about the electoral process relating to the reservation of 37 of the 50 seats in Parliament for ethnic Fijians that ensured they would always have a majority in parliament.

Examples of No:
Botswana - 571-2004-1030-L1; Botswana consistently had free and fair elections and there were no concerns that this election would be different.

Special Cases:
During the 1970s, Kenya allowed a degree of competition in elections within the context of a one-party state. Because competition was overtly restricted, the question is still coded as a "yes."

NELDA12: Was the incumbent or ruling party confident of victory before elections? Details...

A "yes" indicates that the incumbent or ruling party made public statements expressing their confidence that they would win. A "yes" may also indicate that the opposition parties stated that they were not likely to win. If the political situation in the country is such that the opposition had no chance of winning, a "yes" is also coded. When a competitive election outcome is anticipated, a "no" is coded.

Examples of Yes:
Guatemala - 90-1974-0303-P1; The ruling MLN party was confident of victory prior to the election.

Algeria - 615-1997-0605-L1, 615-2002-0530-L1; The ruling party was assured of victory. The opposition parties boycotted, and many of the citizens who would have voted for the parties boycotting did not vote.

Examples of No:
Slovakia - 317-2002-0920-L1; Pre-election polls showed the incumbent party trailing.

Special Cases:

NELDA13: Were opposition leaders prevented from running? Details...

A "yes" was coded when at least some opposition leaders were prevented from running and contesting the elections. A decision to boycott the election was coded "yes" here only if it was in response to the government preventing opposition figures from running. Cases where opposition was not allowed were also coded as "yes." Note that this question is similar to NELDA3 (was opposition allowed?), but is distinct in that it should be coded as "yes" if any specific opposition party candidates are explicitly prevented from running. If NELDA3 is coded "no" this question is coded "N/A." If NELDA3 is "yes" then is question is coded on a "yes"/"no" basis.

Examples of Yes:
Guatemala -90-1970-0301-P1; Communist and revolutionary groups were prevented from running candidates for president.

Iran - 630-1997-0523-P1; In the 1997 presidential election, only 4 out of 238 potential candidates were allowed to run by the Council of Guardians.

Vietnam - 816-2002-0519-L1; The ruling party screened all candidates before they were put on the ballot, and prevented many candidates from running.

Algeria - 615-1997-0605-L1; In the Algerian legislative elections of 1997, all three main opposition parties were banned.

Examples of No:
Equatorial Guinea - 411-1996-0225-P1; Althoughopposition candidates were harassed during the campaign and many of their supporters were detained or tortured, they were not prevented from running.

NELDA14: Did some opposition leaders boycott the election? Details...

If at least some opposition leaders announced and carried out a public boycott of the election, a "yes" was coded. If not, a "no" was coded. A boycott implies an overt decision by a political party not to contest the election. Typically, these leaders also encourage their supporters to boycott the election by not voting. If opposition was banned, or if there was no opposition, (if NELDA3 is "no") then an "N/A" was coded.

Examples of Yes:
El Salvador - 92-1962-0429-P1; All opposition parties boycotted the presidential election after complaining of fraud in the 1961 elections for the Constituent Assembly that wrote the new constitution and transformed itself into a legislative body.

Sudan - 625-2000-1213-P1; In the Sudanese presidential election of 2000, all the main opposition parties boycotted the election.

Sri Lanka - 780-1994-0816-L1; Election turnout was high with the exception of the area around Jaffna, where the LTTE opposition party called an electoral boycott.

Morocco - 600-1993-0625-L1; Many parties boycotted Morocco's 1993 legislative elections. These were mostly leftist parties who were afraid that the government was going to manipulate the elections. The government had not held free elections since 1984 and elections were also postponed from 1989 to 1993 by the leader.

Examples of No:
El Salvador - 092-1991-0310-L1; Opposition leaders did not boycott the election, in fact, this was the first election since 1979 in which the FMLN (main guerrilla group) did not boycott the elections.

Special Cases:
Examples of N/A: Laos - 812-1992-1220-L1; "N/A" was coded because opposition was not allowed in the one-party state.

NELDA15: Is there evidence that the government harassed the opposition? Details...

If there was evidence of intentional government harassment of the opposition, a "yes" was coded. Harassment may include detaining opposition leaders, disrupting opposition political rallies with state forces, and shutting down opposition newspapers and offices. If opposition was banned, or if there was no opposition, (if NELDA3 is "no") then an "N/A" was coded.

Examples of Yes:
Guatemala - 90-1974-0303-P1; The opposition National Opposition Front was the target of violence perpetrated by the government and forces allied with the government. Immediately following the election, the government closed down three radio stations and occupied the headquarters of the Christian Democrats.
Malaysia - 820-1999-1129-L1; The government monitored the opposition's speeches and filed lawsuits against them for slander.
Philippines - 840-2001-0514-P1; Former President Estrada and his son were arrested on TV when they challenged President Arroyo's legitimacy.

Examples of No:
Fiji - 950-2006-0506-L1; there is no evidence the government harassed the opposition.

Special Cases:

NELDA16: In the run-up to the election, were there allegations of media bias in favor of the incumbent? Details...

If there were reports by either domestic or outside actors of media bias in favor of the incumbent or ruling party, it is coded as a "yes." In cases where the media is totally controlled by the government, and/or no opposition is allowed, the answer is "yes." It is possible that the answer is "no" even if the political system is tightly controlled.

Examples of Yes:
Singapore - 830-2001-1103-L1; The ruling party is widely understood to influence media coverage of politics, and opposition parties complain of media bias.
Mauritius -590-1991-0915-L1, 590-1995-1220-L1, 590-2000-0911-L1; In the 1991, 1995, and 2001 legislative elections, the ruling party/government manipulated media coverage, and the media was accused of self-censorship.

Examples of No:
Hungary - 310-2006-0409-L2; There were no allegations of media bias in favor of the incumbent.

NELDA17: Is economic growth in country said to be good? Details...

The answer to this question is based on perception, as reported by media sources. The coding is not based on data on economic growth, which may or may not correspond to perceptions of economic growth. If perceptions were not reported, "unclear" was coded.

Examples of Yes:
Mexico - 70-1991-0818-L1; The Mexican economy was widely viewed as growing fast prior to the 1991 elections.

Examples of No:
Malaysia - 820-1999-1129-L1; The country was described in the media as recovering from the Asian financial crisis.
Algeria - 615-1999-0415-P1; In the period surrounding this election, he economy worsened, inflation was high, and unemployment was at 30%. This only exacerbated the civil war and public discontent.

NELDA18: Is country said to be in an economic crisis? Details...

Examples of Yes:
Guatemala - 90-1985-1103-L1; Several sources described Guatemala's economy as being in crisis.
Laos - 812-1997-1221-L1; Laos' currency was greatly affected by the Asian financial crisis in 1997, especially because Thailand was one of its major trading partners.
Indonesia - 850-1992-0609-L1; The economy slowed, foreign aid dropped, and the country had a large level of debt.

Sudan - 625-2000-1213-P1; Sudan in 2000 had a real GDP growth rate of around 5%, which suggests good economic growth, however, the country had been in a civil war for years that was especially bad in 2003. At this time, the country was perceived as being in an economic crisis, which included widespread starvation.

Examples of No:
Algeria - 615-1999-0415-P1; In 1999 the Algerian economy was not considered to be in a crisis. The economic situation was poor, however, and it was cited as one of the reasons why the government cracked down on opposition groups and manipulated the election.

NELDA19: Is country said to be a large recipient of outside economic aid? Details...

A "yes" is coded if the country is a large recipient of outside economic aid, from either other countries or from international organizations.

Examples of Yes:
Mexico - 70-1985-0707-L1; Mexico was a large recipient of IMF and World Bank loans.
Vietnam - 816-2002-0519-L1; Vietnam had one of the largest aid programs in the world, with economic aid from Western countries, Japan, IMF, and the World Bank.

Examples of No:
Mexico - 070-2006-0702-L1; Mexico was not a large recipient of outside economic aid.

Special Cases:

NELDA20: Was the office of the incumbent leader contested in this election? Details...

If the office of the incumbent leader was contested in the election, such as in a presidential election or in a parliamentary election where the leader is the Prime Minister, then a "yes" was coded. This question focuses on the de facto or genuine leader of the country, rather than whether there is an elected executive. The identity of the incumbent leader is based on Archigos 2.0, a dataset of political leaders by Hein Goemans, Kristian Gleditsch, and Giacomo Chiozza. Where Archigos says the incumbent leader is someone other than the Prime-Minister or President (such as a king or religious figure), the answer is "no".

Examples of Yes:
Mexico - 70-1994-0821-P1; The election was to replace the incumbent president.
Turkey - 640-1999-0418-L1 ; Yes is coded because legislative elections can lead to a change in prime minister, who is considered the incumbent.

Examples of No:
Congo - 484-1973-0624-L1; this was a legislative election and the office of the incumbent leader was not contested.

Special Cases:
Togo 461-2005-0424-P1: The son of the recently deceased president attempted to assume power without an election, but was compelled by international pressure to step down and contest a constitutionally-demanded presidential election. During the election, the acting President was the former vice-president of the National Assembly and a puppet of the dead President's son. We coded that "yes", the office of the Presidency was contested, even though the person holding the office at that point of time was not necessarily the true leader of the country. This decision reflects that contenders perceived the Presidency as the paramount executive institution in the country.

Lebanon--1960-2006. The Archigos dataset codes the office of the president as the incumbent leader, which is not directly elected. However, in this case, because the president is selected by the parliament after elections, we consider the office of the incumbent to be contested in all Lebanese parliamentary elections.

NELDA21: Did the incumbent run? Details...

If the incumbent contested the election, a "yes" was coded. If he or she did not, a "no" was coded. If this is an election that did not involve the leadership position, then "N/A" was coded.

Examples of Yes:
Uganda - 500-2001-0312-P1; The incumbent, Yoweri Museveni, ran for re-election.

Examples of No:
Bolivia - 145-1997-0601-P1; the incumbent president had reached his term limit and did not run in the election.

NELDA22: If no (NELDA21): was there a chosen successor? Details...

If the incumbent leader did not contest the election, but there was a clearly designated or "anointed" successor (who was understood to be the choice of the departing incumbent or ruling group), a "yes" was coded. If the departing incumbent leader did not have a chosen successor, a "no" was coded. If the incumbent leader did run, or the election does not involve the leadership position, "N/A" was coded.

Examples of Yes:
Mexico - 70-1994-0821-P1; The long-ruling PRI's chosen presidential successor was Ernesto Zedillo.
Togo - 461-2005-0424-P1: The son of the recently deceased president attempted to assume power without an election, but was compelled by international pressure to step down and contest a Constitutionally-demanded presidential election.During the election, the acting president was the former vice-president of the National Assembly and a puppet of the dead President's son.This case was coded as "yes" to reflect that the election was meant to arrange for the transfer of power from the interim leader to the son of the former President.

Examples of No:
Liberia - 450-2005-1011-P2; the incumbent government was a transitional government, and no members of the government ran nor was there a chosen successor.

Special Cases:

NELDA23: If yes (NELDA22): did a successor assume power as a result of the elections? Details...

If there was a chosen successor (if a "yes" was coded for NELDA22), and if the successor assumed power due to the elections, then a "yes" was coded. If the successor did not assume power, then a "no" was coded. If the answer to NELDA22 was anything other than "yes", then "N/A" was coded here.

Examples of Yes:
Mexico - 70-1994-0821-P1; The long-ruling PRI's chosen presidential successor, Ernesto Zedillo, won the election.

Examples of No:
Costa Rica - 094-1970-0201-P1; Senor Mario Echandi Jimenez, candidate of the incumbent National Unification Party, lost to candidate of National Liberation Party.

Special Cases:
Ukraine - 364-2004-1031-P2; "No" was coded though Yanukovych, the favored government candidate, won rigged elections in the second round of the Ukrainian presidential election. After the Ukrainian Election Commission declared Yanukovych the victor of the presidential race, international pressure and domestic protests, in tandem with the Ukrainian Supreme Court's invalidation of the presidential results, forced a reround of the runoff election. Though Yanukovych won the presidential election, he did not take power and later lost the rerun election. Opposition candidate Yushchenko took power after the rerun (369-2004-1031-P3).

NELDA24: Did the incumbent's party lose? Details...

If the incumbent or ruling party lost, then a "yes" is coded. If no party is associated with the incumbent, or if the regime allows no parties, then "N/A" was coded. If the election is for the executive office, and the incumbent president is not reelected, then the answer is "yes."

Examples of Yes:
Guatemala - 90-1978-0305-P1; The incumbent's MLN party lost the election.

Examples of No:
Vietnam - 816-2002-0519-L1; The country has only one official party.

Special Cases:
"Mongolia - 1993-0606-P1; No was coded. The incumbent ran for re-election under a different party than the traditional ruling party, but still won. Although he ran under a new party, "no" is coded because there was no meaningful change or transfer in power.

Thailand - 800-1988-0724-L1; In the legislative election the primary party of the old coalition lost seats vis-a-vis other parties, notably the second largest party, in its coalition. Even though the incumbent party was no longer the largest party in the coalition, this election was widely seen as a victory for the incumbent coalition, so the answer coded was "no".

Bangladesh - 771-1991-0227-L1; The incumbent was the interim PM and he did not have a party since he was a politically neutral agent sent to restore political and economic stability. The pre-interim government leader Ershad's Jatiya party still won seats, but two opposition parties won a far greater number. However, because the incumbent had no political affiliation, the answer coded was "N/A."

NELDA25: Were there reliable polls that indicated popularity of ruling political party or of the candidates for office before elections? Details...

If there were reports about the election which suggested that reliable polling data existed and indicated the popularity of the ruling party or of candidates, then a "yes" was coded.This may mean that either domestic or international polling agencies were able to carry out and publish survey results.

Examples of Yes:
Mexico - 70-1994-0821-P1; Polling showed the PRI candidate, Ernesto Zedillo, to be the favorite.

Examples of No:
Turkey - 640-1995-1224-L1; In Turkey there was a ban on publishing polling results. It is worth noting that observers did make predictions that were fairly reliable about the popularity of the candidates, but we cannot say that there were any reliable polls.

NELDA26: If yes (NELDA25): were they favorable for the incumbent? Details...

If NELDA25 is coded as "yes", then NELDA 26 was coded "yes" if the polls were favorable for the incumbent. If the polls were not favorable for the incumbent, then a "no" was coded. If no reliable polling data was found (if NELDA25 is coded "no"), "N/A" was coded. Incumbent here means the leader if the leader's office was contested and the ruling party/parties if the legislature was contested. Favorable was interpreted within the context of the election in the sense of doing well, which usually means being in a position to win power.

Examples of Yes:
Mexico - 70-1994-0821-P1; Polling showed the candidate of the incumbent's PRI, Ernesto Zedillo, to be the favorite.

Examples of No:
Myanmar (Burma) - 775-1990-0527-A1; pre-election polls were favorable to the opposition party, the National League for Democracy.
Mexico - 070-2006-0702-L1; Polls right before the election showed the two main candidates, Calderon and Obrador, neck-and-neck. Since the polls were not favorable for the incumbent party's candidate, "No" was coded. Generally, in cases where the polls showed candidates in dead heat, "No" should be coded.

Special Cases:
Guatemala - 90-1985-1103-L1: NELDA25=yes, NELDA 26=N/A and Guatemala - 90-1985-1103-P1: NELDA 25=yes, NELDA 26=N/A;? These two elections denoted a transition from military to civilian rule. Since the military leaders were not involved, despite accurate polling of the candidates, it is not possible to determine whether the polls were favorable to the incumbent or not.

NELDA27: Was the vote count a gain for the opposition? Details...

If the election vote count displayed a gain from previous elections for the opposition, even if they did not win, a yes is coded. If there was no gain in the vote by the opposition, a no is coded. If these were the first contested elections, then the answer is yes if the opposition gained some seats in parliament or representation in the office under contestation. In the case of a presidential election, vote count gains are coded in terms of victory, not in percentage of votes. So, if a losing candidate runs in two consecutive elections, but loses the second election by a smaller margin, it cannot be said that the vote count was a "gain for the opposition." This is different for legislative or parliamentary elections: an opposition party may not become the majority party, but can still win more seats than in a previous election, thereby making the vote count "a gain for the opposition." Also important is the perception of gain for the opposition, regardless of the precise numbers in the vote count. If opposition is not allowed or absent, the answer is N/A. When coding for a multiple round election, the last round should refer to the election overall. Thus, we count a gain when there is a gain from the previous election (as opposed to a higher percentage of votes compared to the last round). For the 1st and interim rounds, if it is not absolutely obvious that the opposition will gain seats or lose seats as opposed to the last election, then a N/A is coded.

Examples of Yes:
Guatemala - 90-1974-0303-L1; The opposition Christian Democrats scored a gain in seats in the election.
Greece - 350-1977-1120-L1; The opposition picked up more seats than in previous elections. However, the ruling New Democracy party still won a majority of seats in the legislature.
Mexico - 070-2006-0702-L1; In multiparty democracies, there may be an overall opposition loss of seats in a legislature (the ruling party increases its lead) that still leads to a "Yes" coding. This is when one opposition party increases its lead vis-a-vis others or when the media protrays the election as a success for an opposition party. In the Mexico case, the ruling PAN gained seats. While the opposition PRI lost many seats, the opposition PRD gained seats. "Yes" was coded with respect to the PRD.

Examples of No:
Cameroon - 471-1997-0517-L1; the ruling RPDC party picked up even more seats in this election than in the previous election.

NELDA28: Is there evidence that reports critical of the government's handling of the election reached large numbers of people? Details...

If there is evidence that reports critical of government handling of the election reached a large domestic audience, then a "yes" was coded. If there were some critical reports, but they did not reach a large domestic audience, then a "no" was coded. If there were no reports critical of the handling of the election, for example if the election was considered free and fair, a "N/A" was coded.

Examples of Yes:
Philippines - 840-2004-0510-P1; The opposition alleged massive vote fraud in this election. Protests ensued.
Mexico - 70-1985-0707-L1; The election was widely regarded as fraudulent by the population.

Examples of No:
Cambodia - 811-1972-0604-P1; The government closed down many newspapers to prevent reports critical of the government's handling of the election from reaching the population.

Special Cases:
Iran's ruling authorities strictly control freedom of expression, so any criticism of the unfairness of an election cannot reach a wide audience domestically. For this and similar cases, the answer is "no."

NELDA29: Were there riots and protests after the election? Details...

If so, a "yes" is coded. The riots and protests should at least somewhat be related to the handling or outcome of the election.

Examples of Yes:
Philippines - 840-2004-0510-P1; Demonstrators protested the handling of the election.
India - 750-1998-0216-L2; The previous elections in 1996 resulted in a deadlocked parliament. New voting had to be called three years ahead of schedule after politicians spent the previous two years trying and failing to cobble together an enduring minority or coalition government. On February 22, 1998, the popular BJP candidate Vajpayee went on a hunger strike to protest the dismissal of his party's government in the Uttar Pradesh state. Many Hindu nationalists rioted as well.

Examples of No:
Argentina - 160-2005-1023-L1; there were not protests or riots after the election.

NELDA30: If yes (NELDA29): did they involve allegations of vote fraud? Details...

Question 30 was coded as "yes" if the riots or protests are backed with allegations of vote fraud. If there are no allegations of vote fraud fueling the riots or protests, a "no" was coded. If question 29 is coded "no", "N/A" was coded here.

Examples of Yes:
Guatemala - 90-1982-0307-P1; The three losing opposition candidates led protests against what they viewed as fraudulent elections.
Philippines - 840-2004-0510-P1; Opposition figures alleged massive vote fraud. Protests ensued.
Algeria - 615-1999-0415-P1; In Algeria there are many riots and protests after elections. During the presidential election of 1999 the civil war was still going on and unrest was significant. In this election six candidates pulled out of the race because they thought the elections would not be free. The government harassed the Islamist parties and many people protested and rioted as a result. and riot.

Examples of No:
Cambodia - 811-1972-0604-P1; Although there were protests, they were food riots and not related to vote fraud.

Special Cases:

NELDA31: If yes (NELDA29): did the government use violence against demonstrators? Details...

Question 31 was coded "yes" if the government responded with violence against the protests or riots. If no violence was used, "no" was coded. If there was violence, but it was not associated with government actors, the answer is "no". If question 29 is coded "no," "N/A" was coded here.

Examples of Yes:
Guatemala - 90-1982-0307-P1; Protests were marked by violent clashes with police.

Examples of No:
Armenia - 371-2003-0219-P1; thousands of people marched peacefully in the capital, but no violence was used against the demonstrators.

NELDA32: Were results that did not favor the incumbent cancelled? Details...

In Presidential elections where the incumbent or a figure favored by the incumbent ran, "favorable results" were decided on the basis of how that candidate did. In legislative elections where the leader's office is not contested, "favorable results" were decided on the basis of whether the incumbent found the results favorable. Once this is established, results that favor the incumbent led to a "N/A" coding here (whether cancelled or not). Results that did not favor the incumbent led to a "yes" if cancelled, and "no" if left intact. If it is unclear whether the results favored the incumbent or not, "N/A" is coded regardless of whether they were cancelled or not.

Examples of Yes:

Iran - 630-1984-0415-L2; The Council of Guardians (formed in July 1980- 30629), annulled results in 20 constituencies where results were not favorable to the incumbent, citing the 'unhealthy atmospheres' in the constituencies concerned during the election campaign.
Examples of No:
Czech Republic - 316-2002-1025-L2; the incumbent won the election and the results were not cancelled.

Special Cases:
Myanmar (Burma) - 775-1990-0527-A1; The election results were never officially cancelled. However, the military regime delayed transfer of power to the elected party, eventually imprisoned or killed almost all of the elected opposition, and declared that new elections would need to be held, until the issue was largely forgotten. Because the government never officially issued a declaration canceling the results of the election, the answer coded was "no."

NELDA33: Was there significant violence involving civilian deaths immediately before, during, or after the election? Details...

If there was any significant violence relating to the elections that resulted in civilian deaths, a “yes” is coded.

Examples of Yes:
Cambodia - 811-1993-0523-L1; The Khmer Rouge murdered members of rival political parties and marred the election campaign.
Pakistan - 770-1993-1006-L1; The Montreal Gazette states, "Pakistan has a history of volatile politics, although this campaign, coming after nearly a year of political chaos, has been more peaceful than most, with only about 50 people killed." Compared to previous elections, the number of deaths is quite minimal; however, because this question entails comparing the number of deaths to other countries, fifty deaths are considered a significant amount.

Examples of No:
Costa-Rica - 094-2006-0205-P1; there was no evidence of significant violence involving civilian death.
Syria - 652-1953-0710-P1; One bomb went off on election day. This was counted as being relatively peaceful in Syria.

Special Cases:

NELDA34: Were results that were favorable to the incumbent cancelled? Details...

In Presidential elections where the incumbent or a figure favored by the incumbent ran, "favorable results" were decided on the basis of how that candidate did. In legislative elections where the leader's office is not contested, "favorable results" were decided on the basis of whether the incumbent found the results favorable. Once this is established, results that did not favor the incumbent led to a "N/A" coding, whether cancelled or not. Results that did favor the incumbent led to a "yes" if cancelled and "no" if left intact. If it is unclear whether the results favored the incumbent or not, "N/A" is coded, regardless of whether the election was cancelled. For partial overturns of votes, "yes" was coded only where results favorable to the incumbent were cancelled and the overturn was deemed significant or substantial, in other words, if it tipped the power balance away from the government in power before the election. If the cancelled results were minimal and seen as something of low local concern, then "no" was coded.

Examples of Yes:
Guatemala - 90-1982-0307-P1; The results were cancelled by coup leaders who subsequently took over the government.
Philippines - 840-1986-0207-P1; The Marcos government won the election through electoral fraud. International and domestic pressure forced Marcos to flee the country.

Examples of No:
Albania - 339-2001-0624-L4; the results favored the incumbent, and were not cancelled.

NELDA35: If yes (NELDA34): was this in part a result of wide-spread protests? Details...

The case for the "yes" answer is clear. If question 34 was coded "no" or "N/A," question 35 was coded "N/A." (If 34 was "unclear", then 35 is "unclear".)

Examples of Yes:
Philippines - 840-1986-0207-P1; The Marcos government won the election through electoral fraud. The United States and the local Catholic Church both rejected the official results of the election, spurring wide-spread local protests.
India - 750-1989-1122-L1; The original results in Prime Minister Rajiv Gandhi's district were overturned and voting had to be rerun due to fraud and intimidation. Gandhi subsequently won the election, but by a smaller margin.

Examples of No:
Benin - 434-1968-0505-L1; All candidates were chosen by the incumbent military government. However, 74% of the population boycotted the election, thus causing the election results to be annulled. Though this was a boycott, it was not due to wide-spread protests after the election and was coded as "no."

NELDA36: If yes (NELDA34): was this in part a result of outside pressure? Details...

The case for the "yes" answer is clear. If question 34 was coded "no" or "N/A", question 36 was coded "N/A." (If 34 was "unclear", then 36 is "unclear".)

Examples of Yes:
Philippines - 840-1986-0207-P1; The Marcos government won the election through electoral fraud. Under heavy pressure from the United States, Marcos went into voluntary exile in Hawaii.

Examples of No:
Gabon- 481-1990-0916-L1 - Results of the election had been annulled by the supreme court in 32 out of 110 constituencies due to serious, government-recognized, irregularities and the elections in theses constituencies were scheduled to be re-run, however this was not due to outside pressure.

Special Cases:
"

NELDA37: If yes (NELDA34), was a new election held? Details...

The case for the "yes" answer is clear. If question 34 was coded "no" or "N/A", question 36 was coded "N/A."? (If 34 was "unclear", then 36 is "unclear".)

Examples of Yes:
India- 750-1998-0216-L1; The Indian Election Commission ordered a rerun of polling in 599 stations, which contained hundreds of thousands of citizens, due to the deaths of over two dozen people at some of these polling stations. However, since this was only a partial rerun, the answer coded was "no.

Examples of No:
Pakistan - 770-1977-0307-L1 - Results were annulled due to widespread election manipulation, however new elections were not held as a military coup took place after the election

Special Cases:

NELDA38: If yes (NELDA37), did victory go to a different party or candidate than at the "initial" stage? Details...

If question 37 was coded "yes", and a different candidate or party won than in the previous stage, then "yes" was coded. If the same candidate or party won, then "no" was coded. If Q37 was coded "no" or "N/A", Q36 was coded "N/A."

Examples of Yes:

Bangladesh - 771-1996-0215-L1 - Due to irregularities in the voting process and protests by the main opposition parties, who boycotted these elections, the results were cancelled, an interim government installed, and new elections held. The new election was held on June 12, 1996, which were won by the Awami League, one of the opposition parties that had boycotted these elections.

Examples of No:
Burkina Faso - 439-1997-0511-L1 - The Supreme Court overturned the ruling CDP party's victories in 4 constituencies, but in the re-run, the CDP again won the seats.

NELDA39: Was the incumbent replaced? Details...

If the incumbent leader of the country changed after the election, "yes" was coded. A "yes" was coded even if the election did not involve the leadership position but led to events that dislodged the leader from power. Otherwise, if the election did not involve the leadership position, "N/A" was coded. Data for incumbents again come from the Archigos dataset. If the incumbent changes "soon" after the elections but the change is not a consequence of the election or of election-related protests or other similar events, the coding is "no".

Examples of Yes:
Mexico- 70-1988-0706-P1; The incumbent, de la Madrid, was replaced by Salinas.
Indonesia - 850-2004-0705-P2; The incumbent lost the election and was replaced.
Turkey - 640-1995-1224-L1; In Turkey's 1995 legislative election the incumbent party was replaced by another one in dominating parliament, and the prime minister was changed r a few months later. The answer here was "yes."

Examples of No:
Albania - 339-1996-0526-L3; The incumbent leader, Sali Berisha, was not replaced in the election.

Special Cases:
Iran - 630-1989-0728-P1; The leader of Iran is considered the Supreme Leader, not the president. After to the death of Ayatollah Khomeini, Khameini resigned the presidency to become the Supreme Leader. Though this election is a presidential election, it was not the cause of the change of the leader in the country (the Supreme Leader is not elected).
Lebanon--1960-2006. The Archigos dataset codes the office of the president as the incumbent leader, which is not directly elected. However, in this case, because the president is selected by the parliament after elections, we consider the office of the incumbent to be contested in all Lebanese parliamentary elections.

NELDA40: If yes (NELDA39), did the leader step down because the vote count gave victory to some other political actor? Details...

If 39 was "yes" and if the leader was replaced due to victory by another political actor, question 40 was coded as "yes." If the incumbent was replaced, but not by the vote count, then question 40 was coded "no." If question 39 was coded "no" or "N/A," the coding here is "N/A."

Examples of Yes:
Mexico - 70-1988-0706-P1; The incumbent, de la Madrid, was replaced by Salinas, who won the election.
Indonesia - 850-2004-0705-p2; The incumbent lost the election and was replaced.
Peru - 135-2006-0409-P2; The incumbent president, Alejandro Toledo, did not contest this race. Alan Garcia of the opposition APRA party won the presidential election and Toledo stepped down.

Examples of No:
Guatemala - 090-1982-0307-P1: The vote count did give victory to another political actor, but the leader actually stepped down as the result of coup before the elected leader could take office.

NELDA41: If yes (NELDA39), was the leader replaced as a result of widespread protests? Details...

If 39 was "yes" and if the leader was replaced due to or partly due to protests, question 41 was coded "yes."? If the incumbent was replaced but not by protests, question 41 was coded "no." If question 39 was coded "no" or "N/A," the coding here is "N/A."

Examples of Yes:
Kyrgyzstan - 703-2005-0227-L2; The incumbent was replaced due to protests over a fraudulent legislative election.

Examples of No:
Algeria: 615-1999-0415-P1; The new leader, Bouteflika, took his position without incident.

Special Cases:

NELDA42: If yes (NELDA39), was there a coup that prevented the elected leader from taking office? Details...

If there was a coup that overthrew the elected leader, question 42 was coded "yes." If not, it was coded "no." If question 39 was coded "no" or "N/A," the coding here is "N/A." A coup takes place if sources have called the change a coup or if the change in leadership is a sudden, forcible, and illegal removal of a government, usually by the military or some part thereof.

Examples of Yes:
Guatemala - 90-1982-0307-P1; A military coup two weeks after the election prevented the winner of the presidential election from taking office.

Examples of No:
Kyrgyzstan - 703-2005-0227-L2; The incumbent was replaced due to protests over a fraudulent legislative election, rather than a coup.

Special Cases:

NELDA43: What was the name of the leader who was in office before the election? Details...

A free-text answer field for the name of the leader prior to the election is coded here per Archigos. This question was answered regardless of whether the office of the incumbent was contested in the election or the incumbent was replaced.

Examples:

Bahrain - 692-2006-1125-L2; The leader in office before the election was Hamad Isa ibn al-Khalifah

Special Cases:
Lebanon--1960-2006. The Archigos dataset codes the office of the president as the incumbent leader, which is not directly elected. However, in this case, because the president is selected by the parliament after elections, we consider the office of the incumbent to be contested in all Lebanese parliamentary elections.

NELDA44: What was the name leader who was in office after the election? Details...

A free-text answer field for the name of the leader after the election is coded here per Archigos. This question was answered regardless of whether the office of the incumbent was contested in the election. The two names will be the same in 43 and 44 if there is no change. "After the election" is intended to capture any changes in leadership caused by the election. If this was a multi-round election, this question should reflect who is in power after each round, so, for example, there may be no change in leader until the last round.

Examples:

Argentina: 160-1973-0311-L1; the leader after the election was Hector Campora
br /> Special Cases:
Lebanon--1960-2006. The Archigos dataset codes the office of the president as the incumbent leader, which is not directly elected. However, in this case, because the president is selected by the parliament after elections, we consider the office of the incumbent to be contested in all Lebanese parliamentary elections.

NELDA45: Were international monitors present? Details...

If evidence of one or more official delegations of foreign election monitors was found, "yes" was coded. International observers are typically invited by the host government and are organized or sponsored by international organizations or international non-governmental organizations. Journalists or foreigners giving personal accounts of the election are not be considered international monitors in this context. Because international election monitoring is a relatively recent phenomenon, source material before the 1990s may not mention whether international observers were present or not. Thus, before the 1990s, if no reference was made to international monitors, the questions should most likely be coded as "no." If the source material is too sparse to make a judgment, unclear should be coded.

Examples of Yes:
Turkey - 640-2002-1103-L1; Turkey invited election observer missions from the OSCE and the Council of Europe Parliamentary Assembly.

Examples of No:
Benin - 434-1964-0119-L1; no international monitors were present at the election.

Special Cases:
Myanmar (Burma) - 775-1990-0527-A1; The government banned all foreigners from entering the country before the election and denied formal requests for international monitoring. However, three days before the election (May 24, 1990), the government granted 60 visas to journalists for the election; the government also loosened the restrictions on Western diplomats to oversee parts of the elections and serve as international observers. Because these Western diplomats witnessed some polling, as well as the vote counting, and issued formal statements about their observations, the answer coded was "yes."

NELDA46: If yes (NELDA45), were Western monitors present? If NELDA45 is coded as "no" this is coded as "N/A." Details...

If international monitors in NELDA 45 were primarily from Western countries (as defined by OECD membership) or Western international organizations, "yes" was coded. If monitors were present, but not from Western countries, "no" was coded. If monitors were not present, NELDA45 is coded "no", and "N/A" was coded here.

Examples of Yes:

Georgia - 372-1992-1011-L1; Monitors came from Britain, US, Canada, and European Parliament.
Examples of No:
Sudan - 625-2000-1213-P1; The Arab League and the OAU sent monitors, but the West refused to due to expected irregularities in the vote.

NELDA47: If yes (NELDA46), were there allegations by Western monitors of significant vote-fraud? If NELDA46 is coded as "no," this is coded "N/A." Details...

If there were no Western monitors, or no international monitors, "N/A" was coded. If there were Western monitors present and there were allegations of significant vote-fraud by any Western monitors, then "yes" was coded. If there were no allegations of fraud, "no" was coded.

Examples of Yes:
Sri Lanka - 780-1999-1221-P1; Several human rights groups and election monitoring groups noted vote-rigging and violence.

Examples of No:
Indonesia - 850-2004-0705-P2; Monitors did not declare any vote fraud, and praised the quality of the election.

NELDA48: Were some monitors denied the opportunity to be present by the government holding elections? Details...

If monitors were not allowed to be present by the government, "yes" was coded. If no obstacle to monitors was present, "no" was coded, even if monitors did not attend the elections.

Examples of Yes:
Moldova - 359-2005-0306-L1; The Moldovan government expelled Russian monitors, claiming they were conducting espionage. This was part of a larger trend of worsening relations with Russia. OSCE monitors were not expelled.

Examples of No:
Senegal - 433-1978-0226-P1; there is no evidence that any international monitors were present.

Special Cases:

NELDA49: Did any monitors refuse to go to an election because they believed that it would not be free and fair? Details...

If monitors specifically did not attend an election because they did not think that it could be democratic, especially if they were invited to observe by the host government, "yes" was coded. If there is no evidence that monitors refused to monitor the election because of pre-election concerns about the quality of the election, "no" was coded. If there is no evidence that any international monitors considered going to the election, "no" was coded.

Examples of Yes:
Sudan - 625-1996-0306-P1; The West refused to send monitors due to concerns over the fairness of the elections

Examples of No:
Senegal - 433-1978-0226-P1 - The government did not request monitors observe the election.

NELDA50: Is country said to be in good relations with US before the elections? Details...

If the country holding elections was said to have good relations with the US prior to the elections, "yes" was coded. If the country did not have good relations with the US, "no" was coded. If there is no evidence of any substantial relationship, "no" was coded, unless sources were very poor, in which case "unclear" was coded.

Examples of Yes:
Guatemala - 90-1982-0307-L1; Guatemala was strongly supported by the US due to its staunch anti-Communist credentials and strategic location.
Pakistan - 770-2002-1010-L1 Before the election, Pakistan allied with the US to support the war on terrorism, and the US increased aid to Pakistan.

Examples of No:
Iran - 630-1997-0523-P1; Iran was openly hostile to the US and bilateral relations were widely known to be poor, so a "no" is coded.

Bulgaria - 355-1976-0530-L1; Bulgaria was loyal to the USSR, and did not have relations with the US.

NELDA51: If yes (NELDA50), is there a negative change in relations with the US after the election? Details...

If the country had relations with the US (either good or bad) before the election and there was a negative change after the election, "yes" was coded. If no negative change was observed, "no" was coded.

Examples of Yes:
Bosnia-Herzegovina - 346-2002-1005-P1; Nationalist parties that the US opposed gained power in the election.

Examples of No:
Bulgaria - 355-1976-0530-L1; Bulgaria was loyal to the USSR, and did not have relations with the US. This election did not lead to a change in relations.

Special Cases:

NELDA52: If yes (NELDA50), is there a positive change in relations with the US after the election? Details...

If the country had relations with the US (either good or bad) before the election and there was a positive change after the election, "yes" was coded. If no positive change was observed, "no" was coded. If any observed change was not connected the election or events surrounding the election, a "no" was coded.

Examples of Yes:
Pakistan - 770-1993-1006-L1; After the newly elected Prime Minister Bhutto assumed power, Pakistan asserted that negotiations with the U.S. about Pakistan's nuclear policy became more positive.
Guatemala - 90-1984-0701-A1; Due to a new focus on human rights in US foreign policy, Guatemala's transition to democratic elections was hailed by the US government and led to better relations.
Bulgaria - 355-1990-0610-L1, L2; The former Soviet satellite state held free elections in which the former communists won. Because the "G-24" (including the USA) promised economic aid to Eastern European states that held elections, relations with the US warmed.
Cambodia - 811-1993-0523-L1; The US lifted its embargo against Cambodia at the beginning of 1992, normalizing economic relations with the country.

Examples of No:
Bulgaria - 355-1976-0530-L1; Bulgaria was loyal to the USSR, and did not have relations with the US. This election did not lead to a change in relations.

NELDA53: Is the country said to have a substantial economic, military or political relationship with a Western country or IGO? Details...

If the country had a substantial relationship with another Western country (as defined by membership in the OECD) or IGO, such as the World Bank or the IMF, "yes" was coded. Only organizations that can provide substantial economic or military benefits count. The UN as a whole, for example, does not count, although a specialized and powerful UN body such as the International Atomic Energy Agency may count if that agency played a large part in mediating the country's relations with the outside world.

Examples of Yes:
Honduras - 91-1981-1129-P1; Honduras had close economic and political ties to the US and received significant military funding as well.
Indonesia - 850-2004-0705-P2; The United States had a strong security and strategic relationship with Indonesia.
Turkey - 640-1991-1020-L1; Turkey had substantial relationships with the EU since it was under consideration for membership and had strong economic ties. It also had strong military relations with the US.
South Africa - 560-1966-0330-L1; South Africa had a substantial relationship with the United Kingdom, the country's former colonizer, who secretly dealt arms to South Africa despite UN voluntary arms embargo.

Examples of No:
Syria - 652-1973-0525-L1; Syria was aligned with the USSR and did not have any significant relations with Western countries or IGOs.

Special Cases:

NELDA54: If yes (NELDA53), which one? Details...

If question 53 was "yes," question 54 is a free-text answer field for the name of the Western countries (as defined by membership in the OECD) or IGOs that the country had a relationship with. If Q53 was not "yes", the coding here was "N/A".

Example: Bahrain - 692-1972-1201-A1 --Bahrain was a former protectorate of the UK, which had a large presence in the area, and was one of Bahrain's main trading partners.

NELDA55: Is there a negative change in the country's economic, military or political relationship with a Western country or IGO after the election? Details...

If a negative change was observed and it was causally connected to the election, a "yes" was coded. Note that this question is coded without attention to the existence of a significant relationship per NELDA53: what matters is whether or not there is a negative change in an existing relationship with Western actors or IGOs, not whether it is significant. For example, although Cuba and the US do not have a significant economic, military or political relationship, the relationship can, in theory, deteriorate even further as the result of an election.

Examples of Yes:
Bosnia-Herzegovina - 346-2002-1005-L1; The US expressed its displeasure with gains by ultra-nationalist parties in the election.
Examples of No:
Cameroon - 471-1965-0320-P1 - Cameroon had close ties with France, and the election did not affect the relations.

NELDA56: Is there a positive change in the country's economic, military or political relationship with a Western country or IGO after the election? Details...

Analogous to NELDA55, if a positive change was observed and it was causally connected to the election in some way, a "yes" was coded. Again, this question is coded even if the relationship is not significant, per NELDA53.

Examples of Yes:
Guatemala - 90-1984-0701-A1; Due to a new focus on human rights in US foreign policy, the transition to democratic elections was hailed by the American government and led to better relations.
Cambodia - 811-1993-0523-L1; After Cambodia called its first multiparty elections following a suspension of regular elections, the US normalized economic and political relations with the country.
Mongolia - 712-1990-0729-L1; After successful elections, the International Community applauded electoral reforms, and the IMF considered Mongolia's application for membership.
Congo (Brazzaville) - 484-1961-0326-P1; In these elections, the first presidential elections after independence, France was pleased with the election of the incumbent leader, and contributed substantial and support towards building a new dam.

Examples of No:
Bosnia-Herzegovina - 346-2002-1005-L1; There was a negative change in relations with the US and OSCE due to the parties that gained power.

NELDA57: Is aid cut-off, or threatened to be cut-off, by an outside actor at any point before or after the election? Details...

If aid was cut off or there was a threat to cut off aid by an outside entity (Western or non-Western) before or after the election, "yes" was coded. If no threat or cut off was found, "no" was coded. If a threat was unrelated to the election, a "no" was coded.

Examples of Yes:
Bosnia-Herzegovina - 346-2000-1111-L1; Western officials implied that aid could be cut off if nationalist parties took power.
Haiti - 041-1995-0625-L1; A US foreign aid bill passed by the House of Representatives on July 11, 1995, included the stipulation that Haiti would only receive aid if its electoral process met certain standards.
Liberia - 450-1985-1015-L1; The U.S. government, which is by far Liberia's largest foreign aid donor, paid careful attention to the election. Congress demanded that U.S. assistance, $86 million this year, be withheld if the election was not found free and fair.
Albania - 339-1978-1112-L1; Though China had cut off aid to Albania in July of the same year, this was a reaction to Hoxha's improving relations with the West, and not connected to the election.

Examples of No:

Special Cases:

NELDA58: Did an outside actor attempt to influence the outcome of the election by making threats to withhold, or by withholding, something of value to the country? Details...

If an external actor (Western or non-Western) attempted to influence the quality of the election, the way it was held, or its outcome by making threats to withhold something of value to the country, "yes" was coded. Potentially valued items include foreign aid, membership in or benefits associated with international organizations, trading relationships, investment, diplomatic relations, or other forms of external support. Attempting to influence the election may include efforts to improve the quality of the election by threatening to, for example, withdraw aid if election fraud takes place, but may also include threats conditioned on the loss or victory of a particular party or candidate.

Examples of Yes:
Fiji - 950-1992-0523-L1; In the 1992 Fiji election, Australia withheld aid until democratic elections were held.
Myanmar (Burma) - 775-1990-0527-A1; In 1988, when the government used violence against pro-democracy demonstrators, all aid to the country was cut off by foreign states. The 1990 elections were held in part because of international pressure from foreign countries, and economic aid was withheld until Myanmar held elections.
Nigeria - 475-1998-0425-L1; The Commonwealth warned Nigeria that unless it continued its transition to democracy, further international sanctions and expulsion from the Commonwealth were possible. Nigeria's suspension from the Commonwealth continued, however.
Taiwan - 713-1995-1202-L1; China denounced the poll as a fraud, and mounted a pre-election campaign, including provocative military exercises for a possible invasion in an attempt to influence the outcome.

Examples of No:
Azerbaijan - 373-2003-1015-P1; No outside actor attempted the influence the election. However, the US invested $2 million dollars in the election in order to ensure that it was free and fair; they were critical of previous elections.
Burundi - 516-1965-0510-L1; No evidence of outside actors attempting to influence the election. In addition, the election was not considered important as power was centered in the monarchy.

Countries Covered

The list of countries covered follows Gleditsch and Ward's List of Independent States. All states in existence for any period between 1960 and 2006 are listed below. If they have not been included in the NELDA database, the reason for exclusion is listed in the parenthetical following the country name. Micro-states are defined as those countries with a population < 500,000 citizens at the time of the election. Coverage dates follow each included country, and if the beginning year is not 1960, it is equal to the country's year of independence.

Afghanistan, 1960-2006
Albania, 1960-2006
Algeria, 1962-2006
Andorra (micro-state)
Angola, 1975-2006
Antigua & Barbuda (micro-state)
Argentina, 1960-2006
Armenia, 1991-2006
Australia (developed democracy)
Austria (developed democracy)
Azerbaijan, 1991-2006
Bahamas (micro-state)
Bahrain, 1971-2006
Bangladesh, 1971-2006
Barbados (micro-state)
Belarus, 1991-2006
Belgium (developed democracy)
Belize (micro-state)
Benin, 1960-2006
Bhutan, 1960-2006
Bolivia, 1960-2006
Bosnia-Herzegovina, 1992-2006
Botswana, 1966-2006
Brazil, 1960-2006
Brunei (micro-state)
Bulgaria, 1960-2006
Burkina Faso, 1960-2006
Burundi, 1962-2006
Cambodia, 1960-2006
Cameroon, 1960-2006
Canada (developed democracy)
Cape Verde (micro-state)
Central African Republic, 1960-2006
Chad, 1960-2006
Chile, 1960-2006
China (indirect national elections)
Colombia, 1960-2006
Comoros, 1975-2006
Congo, 1960-2006
Congo, Democratic Republic of (Zaire), 1960-2006
Costa Rica, 1960-2006
Cote d'Ivoire, 1960-2006
Croatia, 1991-2006
Cuba,1960-2006
Cyprus, 1960-2006
Czech Republic, 1993-2006
Czechoslovakia, 1960-1992
Denmark (developed democracy)
Djibouti, 1977-2006
Dominica (micro-state)
Dominican Republic, 1960-2006
East Timor, 2002-2006
Ecuador, 1960-2006
Egypt, 1960-2006
El Salvador, 1960-2006
Equatorial Guinea, 1968-2006
Eritrea, 1993-2006
Estonia, 1991-2006
Ethiopia, 1960-2006
Federated States of Micronesia (micro-state)
Fiji, 1970-2006
Finland (developed democracy)
France (developed democracy)
Gabon, 1960-2006
Gambia, 1965-2006
Georgia, 1991-2006
German Democratic Republic, 1960-1990
German Federal Republic (developed democracy)
Ghana, 1960-2006
Greece (developed democracy)
Grenada (micro-state)
Guatemala, 1960-2006
Guinea, 1960-2006
Guinea-Bissau, 1974-2006
Guyana, 1966-2006
Haiti, 1960-2006
Honduras, 1960-2006
Hungary, 1960-2006
Iceland (developed democracy)
India, 1960-2006
Indonesia, 1960-2006
Iran, 1960-2006
Iraq, 1960-2006
Ireland (developed democracy)
Israel, 1960-2006
Italy (developed democracy)
Jamaica, 1962-2006
Japan (developed democracy)
Jordan, 1960-2006
Kazakhstan, 1991-2006
Kenya, 1963-2006
Kiribati (micro-state)
Korea, People's Republic of, 1960-2006
Korea, Republic of, 1960-2006
Kuwait, 1961-2006
Kyrgyz Republic, 1991-2006
Laos, 1960-2006
Latvia, 1991-2006
Lebanon, 1960-2006
Lesotho, 1966-2006
Liberia, 1960-2006
Libya, 1960-2006
Liechtenstein (micro-state)
Lithuania, 1991-2006
Luzembourg (micro-state)
Macedonia (FYROM), 1991-2006
Madagascar, 1960-2006
Malawi, 1964-2006
Malaysia, 1960-2006
Maldives (micro-state)
Mali, 1960-2006
Malta (micro-state)
Marshall Islands (micro-state)
Mauritania, 1960-2006
Mauritius, 1968-2006
Mexico, 1960-2006
Moldova, 1991-2006
Monaco (micro-state)
Mongolia, 1960-2006
Montenegro (micro-state)
Morocco, 1960-2006
Mozambique, 1975-2006
Myanmar (Burma), 1960-2006
Namibia, 1990-2006
Nauru (micro-state)
Nepal, 1960-2006
Netherlands (developed democracy)
New Zealand (developed democracy)
Nicaragua, 1960-2006
Niger, 1960-2006
Nigeria, 1960-2006
Norway (developed democracy)
Oman, 1960-2006
Pakistan, 1960-2006
Palau (micro-state)
Panama, 1960-2006
Papua New Guinea, 1975-2006
Paraguay, 1960-2006
Peru, 1960-2006
Philippines, 1960-2006
Poland, 1960-2006
Portugal (developed democracy)
Qatar (no elections)
Rumania, 1960-2006
Russia (Soviet Union), 1960-2006
Rwanda, 1962-2006
Saint Kitts and Nevis (micro-state)
Saint Lucia (micro-state)
Saint Vincent and the Grenadines (micro-state)
Samoa/Western Samoa (micro-state)
San Marino (micro-state)
Sao Tome and Principe (micro-state)
Saudi Arabia (no elections)
Senegal, 1960-2006
Seychelles (micro-state)
Sierra Leone, 1961-2006
Singapore, 1965-2006
Slovakia, 1993-2006
Slovenia, 1991-2006
Solomon Islands (micro-state)
Somalia, 1960-1990
South Africa, 1960-2006
Spain (developed democracy)
Sri Lanka, 1960-2006
Sudan, 1960-2006
Surinam (micro-state)
Swaziland, 1968-2006
Sweden (developed democracy)
Switzerland (developed democracy)
Syria, 1960-2006
Taiwan, 1960-2006
Tajikistan, 1991-2006
Tanzania, 1961-2006
Thailand, 1960-2006
Togo, 1960-2006
Tonga (micro-state)
Trinidad and Tobago, 1962-2006
Tunisia, 1960-2006
Turkey, 1960-2006
Turkmenistan, 1991-2006
Tuvalu (micro-state)
Uganda, 1962-2006
Ukraine, 1991-2006
United Arab Emirates (no elections)
United Kingdom (developed democracy)
United States of America (developed democracy)
Uruguay, 1960-2006
Uzbekistan, 1991-2006
Vanuatu (micro-state)
Venezuela, 1960-2006
Vietnam, Democratic Republic of, 1960-2006
Vietnam, Republic of, 1960-1975
Yemen, 1990-2006
Yemen (Arab Republic of Yemen), 1960-1990
Yemen, People's Republic of, 1967-1990
Yugoslavia (Serbia), 1960-2006
Zambia, 1964-2006
Zanzibar, 1963-1964
Zimbabwe, 1965-2006

Sources

Primary sources used to code NELDA variables include those listed below, although other sources were used when required, such as detailed case histories of specific countries. In some cases our data conflict with existing data on elections, and we have made extensive efforts to uncover sources of disagreement and ensure consistent coding.

Elections in Africa: a data handbook , edited by Dieter Nohlen, Michael Krennerich, and Bernhard Thibaut (Available electronically through University library).

Elections in Asia and the Pacific: a data handbook. Volume I, Middle East, Central Asia, and South Asia edited by Dieter Nohlen, Florian Grotz, and Christof Hartmann (Available electronically through University library).

Elections in Asia and the Pacific: a data handbook. Volume I, Middle East, Central Asia, and South Asia, edited by Dieter Nohlen, Florian Grotz, and Christof Hartmann (Available electronically through University library).

Elections in Asia and the Pacific: a data handbook. Volume II, South East Asia, East Asia, and the South Pacific , edited by Dieter Nohlen, Florian Grotz, and Christof Hartmann (Available electronically through University library).

Elections in the Americas: a data handbook edited by Dieter Nohlen (two volumes, hard copy only).

Keesing's Record of World Events

Lexis-Nexis Academic

Pro-Quest Historical Newspaper Databases

CIA World Factbook

Theodora, Archiva of CIA Factbooks from 1989-present

US State Department

Library of Congress Country Reports

IFES Election Guide

BBC Country Profiles

Center on Democratic Performance, Election Results Archive, Binghamton University

Lijphart Elections Archive

Latin American Election Statistics

IDEA

African Elections Database

Economist Intelligence Unit Country Profiles and Country Updates

Inter-Parliamentary Union. Chronicle of Parliamentary Elections. Covers legislative elections and electoral developments. Printed edition 1966-present, limited availability online.

Intercoder-Reliability Summary

NELDA1: Were regular elections suspended before this election?

Overall rate of agreement: 85% (N=384).

Rate of agreement given that both coders answered yes or no: 88% (N=363).

NELDA2: Were these the first multi-party elections?

Overall rate of agreement: 90% (N=384).

Rate of agreement given that both coders answered yes or no: 92% (N=379).

NELDA3: Was opposition allowed?

Overall rate of agreement: 91% (N=384).

Rate of agreement given that both coders answered yes or no: 93% (N=376).

NELDA4: Was more than one party legal?

Overall rate of agreement: 96% (N=384).

Rate of agreement given that both coders answered yes or no: 97% (N=381).

NELDA5: Was there a choice of candidates on the ballot?

Overall rate of agreement: 96% (N=384).

Rate of agreement given that both coders answered yes or no: 96% (N=382).

NELDA6: If regular, were these elections early or late relative to the date they were supposed to be held per established procedure?

Overall rate of agreement: 70% (N=384).

Rate of agreement given that both coders answered yes or no: 86% (N=269).

NELDA7: Before elections, are there clear indications that incumbent made a prior decision to give up power?

Overall rate of agreement: 78% (N=384).

Rate of agreement given that both coders answered yes or no: 88% (N=156).

NELDA8: Did the incumbent reach their term limit?

Overall rate of agreement: 70% (N=384).

Rate of agreement given that both coders answered yes or no: 83% (N=148).

NELDA9: Had the incumbent extended their term in office or eligibility to run in elections at any point in the past?

Overall rate of agreement: 75% (N=384).

Rate of agreement given that both coders answered yes or no: 93% (N=152).

NELDA10: Was country ruled by a "transitional leadership" tasked with "holding elections"?

Overall rate of agreement: 92% (N=384).

Rate of agreement given that both coders answered yes or no: 93% (N=382).

NELDA11: Before elections, were there significant concerns that the elections would not be free and fair?

Overall rate of agreement: 72% (N=384).

Rate of agreement given that both coders answered yes or no: 75% (N=366).

NELDA12: Was the incumbent or ruling party confident of victory before elections?

Overall rate of agreement: 60% (N=384).

Rate of agreement given that both coders answered yes or no: 74% (N=304).

NELDA13: Were opposition leaders prevented from running?

Overall rate of agreement: 83% (N=384).

Rate of agreement given that both coders answered yes or no: 86% (N=369).

NELDA14: Did some opposition leaders boycott the election?

Overall rate of agreement: 77% (N=384).

Rate of agreement given that both coders answered yes or no: 86% (N=337).

NELDA15: Is there evidence that the government harassed the opposition?

Overall rate of agreement: 70% (N=384).

Rate of agreement given that both coders answered yes or no: 80% (N=334).

NELDA16: Were these the first multi-party elections?

Overall rate of agreement: 65% (N=384).

Rate of agreement given that both coders answered yes or no: 76% (N=327).

NELDA17: Was the economic situation (growth and stability) in the country said to be good?

Overall rate of agreement: 70% (N=384).

Rate of agreement given that both coders answered yes or no: 73% (N=371).

NELDA18: Was the country said to be in an economic crisis?

Overall rate of agreement: 75% (N=384).

Rate of agreement given that both coders answered yes or no: 78% (N=370).

NELDA19: Was the country said to be a large recipient of outside economic aid?

Overall rate of agreement: 73% (N=384).

Rate of agreement given that both coders answered yes or no: 79% (N=355).

NELDA20: Was the office of the incumbent leader contested in this election?

Overall rate of agreement: 77% (N=384).

Rate of agreement given that both coders answered yes or no: 87% (N=341).

NELDA21: Did the incumbent run?

Overall rate of agreement: 85% (N=384).

Rate of agreement given that both coders answered yes or no: 95% (N=154).

NELDA22: If no: was there a chosen successor?

Overall rate of agreement: 88% (N=384).

Rate of agreement given that both coders answered yes or no: 81% (N=53).

NELDA23: If yes: did a successor assume power as a result of the election?

Overall rate of agreement: 88% (N=384).

Rate of agreement given that both coders answered yes or no: 95% (N=21).

NELDA24: Did the incumbent's party lose?

Overall rate of agreement: 64% (N=384).

Rate of agreement given that both coders answered yes or no: 89% (N=243).

NELDA25: Were there reliable polls that indicated popularity of ruling party candidates for office before elections?

Overall rate of agreement: 56% (N=384).

Rate of agreement given that both coders answered yes or no: 66% (N=324).

NELDA26: If yes: were they favorable for the incumbent?

Overall rate of agreement: 60% (N=384).

Rate of agreement given that both coders answered yes or no: 74% (N=70).

NELDA27: Was the vote count a gain for the opposition?

Overall rate of agreement: 68% (N=384).

Rate of agreement given that both coders answered yes or no: 81% (N=316).

NELDA28: Is there evidence that reports of the government's handling of the election reached large numbers of people?

Overall rate of agreement: 53% (N=384).

Rate of agreement given that both coders answered yes or no: 79% (N=248).

NELDA29: Were there riots and protests after the election?

Overall rate of agreement: 83% (N=384).

Rate of agreement given that both coders answered yes or no: 84% (N=378).

NELDA30: If yes: did they involve allegations of vote fraud?

Overall rate of agreement: 81% (N=384).

Rate of agreement given that both coders answered yes or no: 91% (N=34).

NELDA31: If yes: did the governments use violence against demonstrators?

Overall rate of agreement: 79% (N=384).

Rate of agreement given that both coders answered yes or no: 85% (N=27).

NELDA32: Were results that did not favor the incumbent cancelled?

Overall rate of agreement: 49% (N=384).

Rate of agreement given that both coders answered yes or no: 94% (N=126).

NELDA33: Was there significant violence involving civilian deaths immediately before, during, or after the election?

Overall rate of agreement: 78% (N=384).

Rate of agreement given that both coders answered yes or no: 81% (N=367).

NELDA34: Were results that were favorable to the incumbent canceled?

Overall rate of agreement: 65% (N=384).

Rate of agreement given that both coders answered yes or no: 96% (N=240).

NELDA35: If yes: was this in part a result of wide-spread protests?

Overall rate of agreement: 96% (N=384).

Rate of agreement given that both coders answered yes or no: 100% (N=2).

NELDA36: If yes: was this in part a result of outside pressure?

Overall rate of agreement: 95% (N=384).

Rate of agreement given that both coders answered yes or no: 0% (N=2).

NELDA37: If yes: was a new election held?

Overall rate of agreement: 94% (N=384).

Rate of agreement given that both coders answered yes or no: 100% (N=2).

NELDA38: If yes: did victory go to a different party or candidate than at the "initial" stage?

Overall rate of agreement: 96% (N=384).

Rate of agreement given that both coders answered yes or no: 100% (N=2).

NELDA39: Was the incumbent replaced?

Overall rate of agreement: 60% (N=384).

Rate of agreement given that both coders answered yes or no: 87% (N=196).

NELDA40: If yes: did leader step down because the vote count gave victory to some other political actor?

Overall rate of agreement: 84% (N=384).

Rate of agreement given that both coders answered yes or no: 79% (N=70).

NELDA41: If yes: was leader replaced as a result of wide-spread protests?

Overall rate of agreement: 87% (N=384).

Rate of agreement given that both coders answered yes or no: 97% (N=69).

NELDA42: If yes: was there a coup that prevented the elected leader from taking office?

Overall rate of agreement: 86% (N=384).

Rate of agreement given that both coders answered yes or no: 99% (N=70).

NELDA45: Were international monitors present?

Overall rate of agreement: 69% (N=384).

Rate of agreement given that both coders answered yes or no: 77% (N=345).

NELDA46: If yes: were Western monitors present?

Overall rate of agreement: 64% (N=384).

Rate of agreement given that both coders answered yes or no: 90% (N=126).

NELDA47: If yes: were there allegations by Western monitors of significant vote-fraud?

Overall rate of agreement: 67% (N=384).

Rate of agreement given that both coders answered yes or no: 90% (N=121).

NELDA48: Were some monitors denied the opportunity to be present by the government holding elections?

Overall rate of agreement: 84% (N=384).

Rate of agreement given that both coders answered yes or no: 99% (N=324).

NELDA49: Did any monitors refuse to go to an election because they believed that it will not be free and fair?

Overall rate of agreement: 85% (N=384).

Rate of agreement given that both coders answered yes or no: 98% (N=329).

NELDA50: Was the country said to be in good relations with the US before elections?

Overall rate of agreement: 77% (N=384).

Rate of agreement given that both coders answered yes or no: 81% (N=362).

NELDA51: If yes: Was there a negative change in relations with the US after the elections?

Overall rate of agreement: 79% (N=384).

Rate of agreement given that both coders answered yes or no: 93% (N=323).

NELDA52: If yes: Was there a positive change in relations with the US after the elections?

Overall rate of agreement: 69% (N=384).

Rate of agreement given that both coders answered yes or no: 82% (N=319).

NELDA53: Was country said to have a substantial economic, military or political relationship with a Western country or IGO?

Overall rate of agreement: 76% (N=384).

Rate of agreement given that both coders answered yes or no: 86% (N=341).

NELDA55: Was there a negative change in relations after the elections?

Overall rate of agreement: 70% (N=384).

Rate of agreement given that both coders answered yes or no: 92% (N=279).

NELDA56: Was there a positive change in relations after the election?

Overall rate of agreement: 59% (N=384).

Rate of agreement given that both coders answered yes or no: 79% (N=273).

NELDA57: Was aid cut-off, or threatened to be cut-off by an outside actor at any point before or after the election?

Overall rate of agreement: 85% (N=384).

Rate of agreement given that both coders answered yes or no: 87% (N=377).

NELDA58: Did an outside actor attempt to influence the outcome of the election by making threats to withhold or by withholding something of value to the country?

Overall rate of agreement: 85% (N=384).

Rate of agreement given that both coders answered yes or no: 89% (N=367).

Research funded by generous support from Yale University, the MacMillan Center for International and Area Studies, and the Institute for Social and Policy Studies.

Overview
Contact
Variables
Countries Covered
Sources
Intercoder-Reliability Summary

Overview

Contact and Data Access

Variables

Countries Covered

Sources

Intercoder-Reliability Summary

NELDA

National Elections Across Democracy and Autocracy

Principal Investigators:

Susan Hyde & Nikolay Marinov, Department of Political Science, Yale University

Contents