Vote Shifting Patterns: Comparative Cases from the Constituent Assembly Elections in Nepal

As the literature on electoral politics reveals, vote shifting is one of the most politically sensitive issues. This article focuses on the vote shifting pattern of electoral politics of Nepal. It is based on the secondary data from the first and the next Constituent Assembly (CA) elections held in 2008 and 2013 respectively, published by the Election Commission of Nepal. It aligns with the mixed of quantitative and qualitative approaches. It is limited to the data that belong to the first-past-the-post (FPTP) electoral system covering 240 constituencies and partywise vote shifting pattern. The results show that a higher vote shifting pattern was marked in the 2013 CA election in comparison to the 2008 CA election. The results also show that the Nepalese voters show a cyclic tendency towards electing a new political power at the cost of the traditional one(s) as their political aspirations are not achieved.


Introduction
Election is one of the major components of the democratic system. It is a peaceful means to get political power and authority. During an election, people exercise a sovereign power, which is vested in people according to the constitutional provision. They would choose their representatives for a certain term, mostly for a five-year time. So, democratic politicians of the world always feel 'emotions of fear' of the next election (Butler, 1955, p. 93). This suggests that the result of every election is never certain. When a large number of voters of the same constituency change their voting decision to vote another political party rather than their previous choice and it seems in result of the election is considered as a vote shifting. In the Nepalese context, as a tendency, the electors change their voting decision in each election. Consequently, the prediction of the outcome of an election is not usually accurate in the context of Nepal.
The election of the Constituent Assembly (CA) is a rare phenomenon, which falls only once an era. However, Nepal made a record of two elections of the CA within six years. The first CA election was held in 2008; however, the CA could not meet its main objective to draft a new constitution for the country. So, the next CA election was held in 2013.
Nepal has a short history of electoral politics. The first declaration was made to elect CA in Nepal by the then Rana Prime minister Mohan Shamsher on 8 January 1951. The declaration of King Tribhuvan, which was announced on 18 February 1951, also mentioned the election of the CA to constitute a new constitution for the country. In the same way, the Interim Constitution of 1951 also repeated the same version. Unfortunately, the commitment did not meet during the regime of King Tribhuvan. His successor King Mahendra dilly-dallied for the CA election. Finally, he promulgated a constitution, The Constitution of Kingdom of Nepal, on 12 February 1959 drafted by Constitution Draft Commission, which was formed by the king (Devkota, 2058 BS, p. 526). He announced an election of the House of Representatives instead of the election of the Constituent Assembly. The voting started on 18 February (1959) and lasted for two months due to the difficulty of transporting the ballot boxes to and from the remote areas (Thapa, 2013, p. 118). In spite of dissatisfaction, the political parties took part in the election. The subject of the election of the CA had disappeared from the political scenario of Nepal for a long time thereafter.
Communist Party of Nepal (Maoist) launched an armed revolt in 1996. During the 10 years of its 'People's War,' the Maoist strongly entered the election of the Constituent Assembly with a political agenda. This became a public discourse and the Nepalese society divided itself into two-parts, for and against the CA. Some supported and some doubted it. The CA election was a bottom-line agenda of the Maoist during the negotiation with the government. The uprising in April 2006 made the election of the CA as a national agenda. Ultimately, the election of the CA was held on 10 April 2008. In May 2012, after nearly four years of the political negotiation, the Constituent Assembly was dissolved without finalizing the long-awaited constitution (Pokharel & Rana, 2013, p. 4). Immediately, the new CA election had been declared. After two postponements, finally, the CA election was held on 19 November 2013. The CA election of 2013 has been called as a 'second Constituent Assembly elections' by media and people, however, the Election Commission (2070 BS, p. 1) has officially called it 'the other' (arko) Constituent Assembly election.
It is very difficult to find out the factors that determine the voting decision of the people. Do electors vote for the same party for a long time or change their choice at every election? Casting vote is a confidential act both by law and ethics. Electors would not like to disclose their voting decision. Similarly, a large number of electors is the other main cause that would not let researchers gather the accurate information about their voting behavior. However, many psephologists have tried to analyze election trends.
The election is a matter of interest and concern for most of the citizens, if not for all. In Nepal, during the campaign, the candidates become key figures, and election-related matters become wide public and private discourses from big media houses to teashops of the villages (Gaige & Scholz, 1991). The results become exciting for the mass to the extent that most of the people leave their daily routine or duty and take part in the discourse on the results of the election. Usually, it takes a few days for the common folk to return back to their usual life. Then only dim images of the election will remain in their minds. Researchers analyze the data of election at the micro-level by the various aspects. The matter of vote shifting, if any, is relatively a longer-lasting task among them, though. This article aims to explore the vote shifting patterns in two CA elections held in Nepal.
The results of these two CA elections were quite different. In the first CA election (2008), the Communist Party of Nepal (Maoist) and the Terai-centered parties won the most of the constituencies. In the next CA election (2013), the Nepali Congress and the Communist Party of Nepal (CPN-UML) secured a greater number of seats. This article sets out to analyze the vote shifting in these two elections. More particularly, it is targeted to answer the following research questions. i) What were the political party-wise vote-shifting patterns in the two CA elections of 2008 and 2013? ii) What was the geography-wise vote shifting patterns in these two elections?

Literature review
A considerable number of research studies have been conducted regarding the issue under consideration of this study. These studies mainly focus on the electoral politics of European and American countries. Some research articles are related to the Asian countries, too. Though to a small extent, local scholars have also contributed to the Nepalese electoral system and elections. The text that follows is an attempt to review the related literature. Butler (1955) believes that our knowledge about voting behavior is based essentially on insights rather than scientific proofs. He indicates that there would be many "floating votes," which are uncertain, and they influence election outcomes significantly. There may be different variables that affect the "floating votes," but he claims that the great bulk of British citizens' vote from the habit for a particular party.
In a study, Jōji (1980) finds that there are different variables to shape the individual's selection for a particular political party or candidate in Japanese politics. According to him, the elector's occupations, age, level of education, sex, industry, firm size, and lifestyle stand out as the salient factors. Similarly, Archer & Johnson (1988) suggest that economic issues are more important for election outcomes in advanced industrial democracies.
According to Bartle (1998), in the context of the British electoral system, social class is indeed an important cause of the voting decision, but it does not 'dominate.' Similarly, he concludes that religion has no effect on the voting Shastri (2003) examines the shifting paradigms in India's electoral politics. He has divided the Indian election into four phases: i) 1952 (the First General Election after Independence) -1967; ii) 1967 -1977; iii) 1977-89; and iv) 1989 to 2003. According to him, the first phase was known as the Nehruvian phase, when one party (the Congress Party) had dominated at the national level and in most of the states of India. In the second phase (1967 -1977), new social alliances across caste groups emerged at the state level and non-Congress governments came into being in several states. In the national election 1977, according to him, the Janata Party rose as a new political power and the monopoly of the Congress Party had ended. As Shastri claims, the third phase of India's electoral politics was a great paradigm shift. The fourth phase of the election created the competitive party system and completely eradicated a one-party monopoly. According to Shastri, new social and political alliances have been formed that altered the landscape of India's electoral politics. He argues that the states in India have emerged as the epicenter of the Indian politics.
A study by Blais (2004) in the five countries of North America and Europe finds that the amount of vote shift is much smaller in the United States and much larger in New Zealand (Table-1). The data show that even on the Election Day somewhere between 4% and 16% of the voters typically change their minds.  Blais, 2004, p. 802 According to Johnston, Propper, Burgess, Sarker, Bolster & Jones (2005), there is a neighborhood effect on people's voting decisions. They claim that the immediate neighborhood, comprising the area within 250 meters, influence each other.
According to Toros (2014), political values are the major social factors that have a considerable impact on voting behavior in Turkey. These factors work differently in different situations. Similarly, he claims that there are many other different social and political factors that are salient in voting behavior. A study by Zamfira (2015) in Bulgaria, Romania, and Slovakia suggests that ethnic issues, the territorial system of political administration, and socio-economic situations have important effects on voting behavior.
The reviews concerned electoral politics which especially focuses on vote shifting and voting behaviors. These trends are concerned with the context of American, European, and some Asian countries. Scholars have investigated Nepali electoral politics too. Gurung (1982) carries out a comparative study on two elections of Nepal that were held in 1959 and 1981. Nepal's first general election was held in 1959. Similarly, the first direct election of Rashtriya Panchayat was held in 1981. Both of the elections were conducted to elect the members of the national parliament, however, these two elections were held in different political environments; one was conducted in the democratic environment of the multiparty system, and the other was held in the partyless regime led by an active monarch. Gurung analyzes the sociological aspects of these two elections. He uncovers facts regarding the participation of the candidates from various castes and ethnic groups of Nepali society in these two elections. He found that most of the candidates were from the Chhetri caste followed by the Hill Brahmins set in the second position and the Newars situated in the third position in both of the elections. Gurung compares these two elections from various aspects; however, he is almost silent about the voting behavior of Nepali people. Gaige & Scholz (1991) analyze the first election of the Parliament after the restoration of multiparty democracy in 1990. The election took place after a year of the restoration, i.e., in 1991. According to them, the party workers used to meet farming voters in the early evenings after they returned from the fields. Similarly, party workers used tape recorders to replay speeches from party leaders to villagers. They claim that vote-buying was common in the Panchayat elections. It got continued in the democratic elections too. It was a negative factor that influenced the voting behavior of the Nepali election system. Baral (1995) analyzes the emerging trends of the Nepal election of 1994 and compares it with the election of 1991. The Nepali Congress (NC) became the largest party of Nepal after the result of the elections of House of Representatives (HoR) in 1991. Nonetheless, the NC had been relegated to the second position and the Communist Party of Nepal -Unified Marxist-Leninist (CPN-UML) reached the first position in the 1994 election. It means the majority of the Nepali people preferred the communist candidates in 1994. In Kathmandu valley, for example, where the 'left' political parties won all of the 13 seats (11 by the CPN-UML and 2 by Nepal Workers and Peasants Party). According to Baral, why the voters in the capital showed a pro-left swing was unclear, and such a shift in the voting behavior did not indicate any definite trend. However, he indicates that the major reason for this was antarghat (sabotage) within the party (NC).
Likewise, Manchanda (2008) analyzes the changing political environment of Nepal after the 'People's War' of the Maoists, Peoples' Movement of 2006, and the election of Constituent Assembly (CA). Citing an example of Surya Bahadur Thapa, a royalist and former Prime Minister, who cast his vote for the "Republic" in the CA. She claims that Nepali politics and society has radically been restructured. Ironically, even from Bhadrakali, the headquarters of the Nepal Army, some 200 votes were cast in the name of the Maoists. She identifies that more than 60 percent of the newly elected members were associated with left-oriented politics in the CA and it was truly inclusive because the elected members were from different social dimensions, and it broke out of the mold of Nepal's socio-political culture of "institutionalized exclusion." In a nutshell, several studies have been conducted in the field of electoral politics, voting behavior, and vote shifting. However, sufficient investigations have not been carried out in this specific area, especially in the Nepalese context. This is a very potential area of study. This study aims to search the facts about vote shifting patterns of electoral politics of Nepal.

Research methodology
This article mainly builds on secondary data. The CA election handbooks published by the Election Commission of Nepal in 2065 and 2070 BS were thoroughly utilized as the source of the data. Similarly, concerned the electronic databases were also consulted. Descriptive and analytical research designs were used for the analysis. The CA had 601 members. Among them, 575 members were elected through the voting system and 26 was nominated by the council of ministers as per the provision laid down by the Interim Constitution of Nepal, 2007. Of those elected members, 335 were elected through Proportional Representation (PR) electoral system and 240 were elected through the First-Past-the-Post (FPTP) electoral system. Since this article's key concern is to examine the general voters' vote shifting tendency, it delimited only to the examination of statistics of the FPTP electoral system.

Results and discussion
The 'first' CA election was held on 28 Chaitra 2064 BS (10 April 2008)   Nepal is naturally divided into three ecological regions. The Mountain region lies in the northern part of the country from east to west. The Hill region is in the middle part and the Terai region is in the southern part of Nepal; like the Mountain regions, these two regions are also spread from east to west. Now, Nepal has been adopted the federal system and there are seven provinces in accordance of constitution of Nepal, 2015. When the elections of CA were performing, Nepal had administratively divided into five development regions; eastern, central, western, mid-western, and far-western development regions. Analyze have made according to these ecological and development regions hereafter. According to figure-2, the highest percentage of vote shifting took place in the Mountain region. In comparing the two elections, only 3 out of 19 constituencies (below 16%) remained unchanged in this region, which means that more than 84 percent of the total constituencies had different results in the 2013 election as compared to the 2008 election. In the case of no change result, the Terai region had a little more (28 out of 115 or above 24%) than in the Mountain region. Overall, more than 75 percent of the constituencies had changed the results in the 2013 election in the Terai region. It is generally believed that the Hill region has the lowest percent of inconsistent results. Yet, in this region more than two-third of the constituencies had contrasting results. It is not too a small number in the case of vote shifting patterns.

Region-wise position of individual parties in the two elections
There were dozens of political parties that participated in the both CA elections. For the convenience of study, parties are divided into different categories, as mentioned above. The status on vote shifting of these political parties and their position in ecological and administrative regions are analyzed in the following sub-headings.

The Nepali Congress (NC)
In the 2013 election, the Terai redion proved the most faveouable site for the NC because, in this election, it secured 44 more seats than it had achieved in the 2008 CA election (Table2). It lost 9 places and sustained the same position in 7 constituencies. Similarly, it gained 34 more seats, lost 6 seats, and obtained the same result in 13 places in the Hill region. In the Mountain region, the NC secured 7 more seats and was lost 2 seats. As per the development region, the NC secured a higher number in the three increase, decrease, and no changecategories in the central region. It obtained 28 more, lost 7, and 9 constituencies are no change result. Similarly, the NC won outnumbered 7 more constituencies, lost 1, and 2 had no change in the 2013 CA election as compared to the in the 2008 election.

The Maoist
The Maoist party underwent a great loss in the CA election 2013 in comparison to the 2008 election. It was unable to increase even a single seat (Table-3), in the Mountain and Hill regions, rather it lost 12 seats in the Mountain and 52 in Hill region. Yet, in the Terai region, it regained 7 more seats but lost 37 seats, and 5 places remained as they were in the 2008 CA election. In terms of the development regions, the Western, the Mid-western, and the Far-western regions remained unfriendly for the Maoist in the CA election 2013. In these regions, the Maoist could not increase its position at all. However, it increased 4 seats in the Eastern and 3 seats in the Central regions. Table-3 also shows that the Maoist lost a larger number of seats in all five development regions of the country. Only in a small number of constituencies, the Maoist could sustain its victory of the 2008 election in the 2013 election.

The Communist Party of Nepal-United Marxist Leninist (CPN-UML)
The first CA election in 2008 did not favor the CPN-UML but this party made improvements in the next CA election in 2013. As shown in Table-  As per the development regions, the CPM-UML made improvements in the CA election 2013 in comparison to that in 2008. In the Central region, it achieved a greater number of seats in the 2013 election than in the 2008 election. Though in a small number, the CPM-UML secured better results in the 2013 election than in the 2008.

The Terai Centered parties
After the Madhes uprising of 2007, some Terai Centered parties and their leaders became popular figures in the Terai region. They had won more seats in the CA election 2008. However, they could not maintain their position in the CA election 2013. As shown in Table-5, in the 2013 CA election, they lost many constituencies, where they had won in the CA election 2008. These parties lost more than they gained in the 2013 election. The Nepalese history of electoral politics proves that the Nepalese people vote more for the party which had led the immediately past revolution. The Nepali Congress had played a significant role to abolish the century-long aristocracy Rana regime in 1951. Honoring the leading role of the Nepali Congress, the Nepalese people gave it a two-third majority (74 out of 109 seats) in the House of Representatives election held in 1959 (Gupta, 1964, p. 146). Similarly, the Nepali Congress and the Unified Leftist jointly led the revolution of 1990 that reestablished multiparty democracy in the country by displacing the partyless Panchayat system in 1990. The first election thereafter was held in 1991. The election was conducted to elect 205 members of House of Representatives, the lower house of the parliament. The result of the election shows that the Nepalese people chose the Nepali Congress as the first party with a simple majority. Similarly, the voters placed the CPN-UML in the second position, and after that there remained the other leftist parties (Election Commission, 1992). The decision of the Nepalese people seems justifiable in the sense that they cast vote according to the ratio of the role of the party in the preceding revolution.
The Communist Party of Nepal (Maoist) launched 'People's War,' the armed revolt, in 1996. It spread all over the country in a relatively short time span. Among the demands set by the Maoist on the peaceful negotiations with the government, the election of the constituent assembly was the bottom line. After the 12-point multilateral agreement between the Maoist and the other parties in 2006, they fought together against the king's despotic regime. Consequently, the Nepalese people won and achieved democracy again. The story did not finish yet. After the promulgation of the Interim Constitution of Nepal, the Madhesi people did not satisfy themselves with the provision of the constitution. Consequently, the Madhes revolt arose and spread from eastern Terai and entirely over the Terai region. It ended with the first amendment of the constitution that admitted the federal system as a form of government in the restructured 'new Nepal.' As the new revolutionary parties, the Nepali people cast more votes to the Maoist and the Terai centered parties in the first CA election 2008. So, it seems to be a political culture of the Nepalese people to flow electorally from one party to another.
The Nepalese people demonstrated a U-turn in the CA election 2013. They changed their voting decision by 180 degrees. As we discussed above, the Maoist party won in 120 constituencies, exactly 50 percent of the total FPTP seats of CA, in 2008's election, whereas it dropped down to 26 seats in the election 2013 (Election Commission, 2070 BS). The results of both of these elections were rather unexpected. As the zigzag line of figure-6 shows, it is an example of very high vote shifting patterns. It shows how drastically people change their voting decision within 5 years. Similarly, the vote shifting pattern practiced by the people over the Terai Centered parties in those two consecutive elections is also startling. It is also another example of a huge vote shifting in the Terai region. We can analyze the result of the CA election 2013 by applying the zero-sum game in the FPTP electoral system. The result shows that the loss of the Maoist and the Terai Centered parties equaled the gain for the NC and the CPN-UML. The Nepalese people, who favored the Maoist and the Terai Centered parties in 2008, decided to return electorally to the traditional political power NC and CPN-UML in 2013.

Conclusion
In the political history of Nepal, two CA elections were held: CA election 2008 (known as 'the first' CA election) and CA election 2013 (known as 'the next' CA election). Both of them were held in the same geographical constituencies, for the same propose, and applying the same electoral system. The internal structure and ideology of two major parties of Nepal, the NC and the CPN-UML, were alike in both 2008 and 2013 CA elections. Some structural changes had appeared with the CPN (Maoist) and the Terai Centered parties, although their ideology and leaders were the same in both cases.
Evidence shows that after the period of 5 years and 9 months, the Nepalese people changed their decisions to float from this party to that electorally. As the data support, the Nepalese people cast substantial votes to, thus, elected greater numbers of candidates of, the Maoist party and the Terai Centered parties in the CA election 2008. However, the voters considerably shifted to the others in the CA election 2013. This study indicates towards another fact that the Nepalese people, when dissatisfied with new political power(s), return to traditionally established political parties. As the authentic secondary data suggest, the previously held elections of the House of Representatives (1991, 1994 and 1999)  Others & Independents major political powers respectively. However, the NC could secure 37 (15%) and the CPN-UML 33 (14%) out of the total of 240 FPTP CA seats in 2008. In the 2013 CA election, the NC won in 105 (44%) constituencies and the CPN-UML in 91 (38%). The Nepalese voters re-decided to cast votes to the traditional political powers in 2013. The comparative study of the results of these two CA elections has provided an overt example of vote shifting.