In the shadow of plagues : 2020 presidential elections in Poland

Presidential elections in Poland have always drawn more voters to the ballot box than parliamentary ones. Despite the COVID-19 pandemic, the turnout in the 2020 presidential elections was the second-highest since 1989. The glib answer might be that voters were simply availing themselves of the opportunity to leave the house, however briefly, since COVID-19 measures had largely confined people to their homes for most of the year. More likely, albeit paradoxically, heightened voter interest was triggered by political autocratisation, the other plague that Poles have been struggling with, in this case since 2015. The election was won by the incumbent, Andrzej Duda, whose advantage was credited to his appeal among less-educated and older people living in rural areas. In the end, the governing Law and Justice (PiS) party, which had backed Duda’s campaign, managed to avoid cohabitation-related inconvenience for the second time.


Introduction
P residential elections in semi-presidential systems seem to be as important as parliamentary ones. This claim is grounded on the observations that all political parties invest their effort and resources to win these elections and that citizens tend to vote for a candidate for head of state more often than for prospective deputies. In this context, presidential elections in Portugal might be considered an exception since the parliamentary race mobilises voters more efficiently (IFES, 2021). It is worth emphasising that presidential elections offer a unique opportunity for scholars to research voter transitions between the first and second rounds, as citizens are expected to redefine their initial preferences due to the reduction in candidate numbers in the second round. Data from the 2017 French presidential elections bear this out -in those elections, 20 million people change their vote between first and second rounds (Romero, et al., 2020(Romero, et al., , p. 2726. The above characteristics also apply in the Polish case. However, all problems arising from presidential elections were concentrated in 2020, given the presidential race was the fourth set of major elections that Poles had participated in within a little over eighteen months. Poles had voted in autumn 2018 for regional and local authorities, in spring 2019 for the European Parliament and again in autumn 2019 to express their preferences in national parliamentary elections. As a result, Law and Justice (PiS) retained both a parliamentary majority (in place since the 2015 parliamentary elections) and the presidency. Andrzej Duda was re-elected after a gruelling race featuring multiple competitors.
After the 2015 elections, PiS began dismantling the checks and balances in Polish civic life, including taking control over the public media, the Constitutional Tribunal, the prosecutor's office, and the civil service. This was done legally but controversially since three nominees selected for the Constitutional Tribunal by the incumbent Civic Platform (PO)-led government before the 2015 parliamentary elections were not given a chance to take the oath of office by the new PiS-led government (Fomina, Kucharczyk, 2016, p. 58;Levitsky, Ziblatt, 2018, p. 67). It is important to note that the PO-led government had in a sense acted in bad faith, expanding the number of appointments to five to secure extra places for non-PiS nominees in an attempt to "stack" the court. PiS cited the illegality of this move (recognised as such by the Constitutional Tribunal) to justify rescinding the entire slate of PO nominees and making fresh appointments.
PiS' efforts to weaken Poland's democratic institutions were sanctioned by the speaker of the Sejm (the lower house of Poland's bicameral parliament) who -during the very first session of the parliament -asserted: "The good of the nation comes before the law. If the law conflicts with that good, then we cannot treat it [the law] as something sacrosanct" (Seym of the Republic of Poland, 2015, p. 78). The speaker, Kornel Morawiecki, hails from the populist Kukiz movement and is the father of Mateusz Morawiecki (who was later appointed prime minister). The elder Morawiecki is a hero of the radical "Fighting Solidarity" movement, formed in 1982 when Lech Wałęsa's Solidarity was outlawed by the Polish martial law regime and was forced to go underground.
Considering the fact that the Law and Justice government -with the support of President Duda -had been dismantling democratic institutions since 2015 (Levitsky, Ziblatt, 2018, p. 147;Markowski, 2020, p. 111;Sadurski, 2018, pp. 17-19) it could be argued that the 2020 presidential elections were a plebiscite on the future of Polish democracy. After all, The Economist Intelligence Unit (2020) assessed Poland as a "flawed democracy." We treat it as an electoral (i.e., a non-liberal) democracy (Lührmann, et al., 2020, p. 16). We do not contend that the high electoral turnout was entirely down to a perceived threat to democracy; nevertheless, we argue that the intensity of the pre-electoral conflict could have contributed significantly to the social mobilisation that drove a record number of voters to the polls.
Analyses of presidential elections in Poland after 1989 show a few common features. First, the winning candidate has always had the support of a strong political party (or, as in the case of Lech Wałęsa in the 1990 presidential elections, a broad-based political camp). Indeed, Wałęsa's victory drew on the support of the entire anti-communist opposition movement (Glajcar, 2016, pp. 22-26). Second, electoral turnout during presidential elections is consistently higher than in parliamentary, regional or European Parliament elections (Table 1). Third, since 2010, the two biggest parties have not chosen the party leader to run as the favoured presidential candidate (Glajcar, 2016, pp. 24-25). All those features were present in the 2020 race.  1 53  68  61  51  55  55  68  2  43 52  48  46  41  54  49  51  62  3  45  44  46  47  47  55  4  25  24  46 Note: Presidential elections, second round 1 Parliamentary elections 2 Regional elections 3 EU Parliament elections 4.
Source: Based on National Electoral Commission data.

Poland's Changing Electoral Code: A Brief Historical Survey
The competencies of the Polish president are central to explaining why political parties are so determined to mobilise their respective electoral bases to win control of this office. The competencies, in turn, stem from the role of the president in the Polish political system. This is thoroughly highlighted in the literature (Antoszewski, 1998;Sanford, 2002;Słomka, 2005;Alberski, 2010;Piotrowski, 2016). Nevertheless, it is important to highlight three specific elements that influence government performance, which are of paramount importance during cohabitation (i.e., when different parties control the legislature and the presidency). First, the president can initiate legislation, proposing laws to be presented to the parliament. Second, he or she holds veto power -this can be overridden by a threefifths majority vote with a quorum of half of the MPs present. Third, the president has the power to submit legislation passed by the parliament to the Constitutional Tribunal for abstract review; if the law is recognised as complying with the Constitution, the head of state cannot veto it and is obliged to sign it into law (Sula, Szumigalska, 2013, pp. 111-113;Glajcar, 2015, pp. 422, 480). This relatively strong position of a president in the Polish political system is also associated with the high level of legitimacy, as presidents have been elected by universal suffrage in Poland since 1990, and the Constitution of 1997 (Decision of the Speaker, 2020, 184) confirmed this. Both the governing United Right and the opposition parties assumed that a president opposing the government's policies could meaningfully dampen its efforts at autocratisation. Since the 2019 elections, the opposition has held a slight majority in the Senate, the upper chamber of the parliament (51 seats out of 100), which allowed them to slow down the legislative process and introduce some deliberation and consultation. Nevertheless, the United Right has been able to use its majority in the Sejm to (eventually) overcome the Senate's veto.
The Polish Constitution requires that voting in elections be universal, direct, equal, and anonymous and provides for two rounds (if needed) of voting in presidential races. Consequently, to be elected president, a candidate must gain an absolute majority of votes cast. Only once under the current constitution have presidential elections in Poland not proceeded to a second round -namely, in 2000, when Aleksander Kwaśniewski took more than 50% of the vote the first round. The Constitution provides for the postponement or delay of presidential elections under exceptional circumstances. States of excep-PP 2 '21 tion fall into three categories -martial law, a state of emergency and natural disasters. The Constitution provides that no election or referendum may be held during a state of exception or within 90 days of its expiry.
Detailed electoral regulations are gathered in the Electoral Code (Constitution of the Republic of Poland, 1997), which establishes an explicit schedule for elections: Elections shall be announced by the Speaker of the Sejm no earlier than seven months and no later than six months before the expiry of the term of the incumbent President of the Republic. The Speaker shall set the day of the elections on a bank holiday no earlier than 100 days and no later than 75 days before the expiry of the term of the incumbent President of the Republic (Electoral Code Act, 2020).
Art. 38 specifies that voters must vote in person, although this includes voting by post. Another important provision of the Electoral Code (Ibidem, 327) limits presidential campaign expenditures, calculated as a fixed amount per eligible voter (0.64 PLN in 2020). In 2020 the limit for a single candidate's campaign was calculated at 19.2 million PLN (4.3 million EUR).
According to the described provisions, the first round of the elections was set for 10 May 2020 (Decision of the Speaker, 2020). This decision triggered a series of procedures, including registration of candidates, collection of the required signatures to support the candidates, fundraising, and the launch of the electoral campaigns. COVID-19 and the corresponding restrictions imposed from March 2020 due to the pandemic interrupted the entire process. It should be stressed that the Polish law does not provide for any specific provisions for epidemiological threat (Act of 31 March 2020 on amendment, 2020, 568). While it was argued that the situation in spring 2020 was consistent with the terms defining a state of natural disaster (extraordinary conditions causing grave threat to public order or safety, which cannot be prevented or removed by the normal functioning of the state), the government resolved to introduce special provisions for COVID-19 instead (Chmaj, et al., 2020, p. 7).
The Law and Justice party was determined to hold the elections as scheduled. Polls showed high and growing support for their candidate and incumbent president, Andrzej Duda, in the first phase of the epidemic, with citizens broadly supporting the government's pandemic response. However, Law and Justice had reasons to fear that this support would quickly dissipate as mortality rates soared and the economic effects began to bite. Therefore, the government decided to adapt the binding provisions and pending electoral process to the specific conditions of the pandemic.
The first major amendments to the electoral provisions broadened access to postal voting for all citizens over 60, as well as for those in obligatory quarantine. This decision raised a lot of criticism given the increasing number of people infected with COVID-19 in Poland. Many people and organisations-including Małgorzata Kidawa-Błońska, former speaker of the Sejm and an opposition presidential candidate-called for a boycott of elections held during the epidemic, claiming that the organisation of the election would threaten the health of voters (Michalak, 2020, pp. 2-5).
In response to this criticism, the ruling party submitted a proposal to the Sejm to conduct the elections entirely by postal vote. On 6 April, the Sejm approved the measure. It was then taken up by the Senate, which decided to reject the proposal after nearly four weeks of deliberation (Act of 6 April 2020 on special terms, 2020). The major concerns were the secrecy and freedom of the vote (Kalisz, Szulecka, Wolhy, Bakierski, 2021 2020, pp. 6-7), protection of the ballots from theft, and protection of personal data (Bodnar, 2020). The Senate communicated its decision on 5 May, only five days before the planned polling date.
Nevertheless, while the Senate deliberated through April, the governing party began preparing for the elections -albeit without legal grounds -printing ballots and ordering the Polish postal service to prepare for their distribution and collection (Mikołajewska, 2020). However, the government's efforts hit an obstacle when local governments refused to submit the voter files to the postal service before the authorising legislation passed the Senate (Balcer, Banasik, 2020). This situation led to general uncertainty about whether the elections would be held and on what terms. Even members of the ruling "United Right" governing coalition raised questions. Intra-coalition negotiations led to a compromise -the Senate veto would be overturned in the Sejm, but the ballots would not be distributed, and the elections would be declared invalid and rescheduled to a later date. While the coalition parties agreed, this solution had absolutely no basis in Polish law. Thus, the elections were never cancelled but were not held either. The newly approved statute required further amendment to allow another vote to take place.
The governing parties agreed on a set of provisions to reschedule elections, which were put to the parliament in yet another authorising statute (Act of 2 June 2020 on special terms, 2020). The new provisions established a hybrid process, in which the default option would be traditional in-person voting, but with the possibility to vote by post available to any voter wishing to take up the opportunity. This was a compromise between the need to ensure that elections were held before the expiry of Duda's first term (6 August 2020) and the need to prevent the spread of COVID-19. Allowing voters to choose whether to vote by post or in person was also designed to improve their sense of security during the elections.
However, the statute introduced changes that went well beyond running an election under pandemic conditions. For example, while candidates who had registered already were permitted to run, the legislation allowed for new candidates to register, opening the door for parties to replace their candidates but also leaving doubt about whether the existing election was being rescheduled or an entirely new one being organised. It was an important distinction since it affected campaign financing. The new provisions decreed that newly registered candidates could spend only half the statutory limit. In contrast, those formerly registered had the right to spend the full amount, albeit including amounts spent before 10 May. On 3 June 2020, the speaker of the Sejm announced new elections according to these provisions, to be held on 28 June (with a second round scheduled for 12 July).
The ruling party's decision not to refer to provisions on natural disasters -combined with the challenges of the pandemic -led to chaos and uncertainty, further worsened by the opposition's inability to take a consistent stand on the nature and timing of elections. Consequently, large sums were wasted on elections that kept being delayed, and voters were left with doubts concerning both the course of the campaign and the safeguards in place to protect the public's health and wellbeing. The broad acceptance of the finally applied solution was associated with all the parties' willingness to avoid further delays and even more complex problems.

The 2020 Presidential Campaign: Party Competition and Voter Turnout
Throughout the campaign, the primary conflict was between the ruling Law and Justice party and the opposition. The incumbent President Duda was perceived as the frontrunner, with about 50% of respondents declaring support for him in February and early March. His standing in the polls grew in the first few weeks of the pandemic, exceeding 60% at the highest point and giving the ruling right-wing the hope of winning the elections in the first round (Trząsalska, 2020). The opposition, on the other hand, was unable to unite and submit a single candidate to face Duda. This brings us to the second major conflict-that within the pro-democratic opposition. The opposition was fractured among several party candidates-each working to advance his or her party's position and consolidate his or her position as the leading anti-PiS force. The largest parliamentary opposition party, the Civic Platform, initially appointed the deputy speaker of the Sejm, Małgorzata Kidawa-Błońska, as its candidate. However, her lucklustre campaign, lack of charisma and, finally, her decision to call for a boycott of the postal vote led to a dramatic fall in her popularity and opened a path for other candidates.
The primary beneficiary of the collapse in support for Kidawa-Błońska was the nonpartisan candidate and former TV celebrity and political commentator Szymon Hołownia, who presented himself as a "political outsider" who could bridge the partisan divide (Hołownia, 2020). Although trying to strike a balance between left and right, he was outspoken about the ruling party's breaching of the constitution and the rule of law (Bartkiewicz, Malinowski, 2020). At the same time, his strong commitment to Roman Catholicism and consistently conservative views were designed to win over right-wing voters.
A similar "middle way" strategy was adopted by the other candidate who gained at the expense of Kidawa-Błońska -namely, Władysław Kosiniak-Kamysz, the leader of the Polish Peasant Party (PSL). PSL had been part of a governing coalition with PO (2007PO ( -2014 and in electoral alliance with them in the European Parliamentary elections. However, since the parliamentary elections of 2019, PSL has strived to distance itself from the PO-led "Civic Coalition" (KO) -which formed in 2018 as an electoral grouping opposing PiS -and has adopted a moderate stance to attract traditionalist, conservative voters. Kosiniak-Kamysz (Kosiniak-Kamysz, 2020), a mild-mannered, educated man whose family had been involved in politics for generations, seemed a perfect candidate for this purpose and, in fact, his support grew significantly over the campaign, exceeding 10% at certain points (Trząsalska, 2020).
Out of the remaining candidates, the only significant ones were two representatives of parliamentary opposition parties. The rightist, nationalist and conservative Confederation was represented by Krzysztof Bosak. His campaign stressed traditional values and opposed both the liberal lifestyle agenda and the high-cost social transfers championed by PiS (Bosak, 2020). The Left (Lewica) -a coalition of different leftist organisations -took a while to select a candidate. In the end, the coalition settled on Robert Biedroń, Poland's first openly gay politician. With differences among the many sides supporting him, Biedroń's agenda was unclear and his campaign weak, meaning he could not mobilise the same share of voters who had supported the left in the 2019 parliamentary elections.
The real gamechanger came after the first election date passed without a vote. The KO, the main opposition grouping, seized the opportunity to replace their first pick, Kidawa-Błońska, with the mayor of Warsaw, Rafał Trzaskowski. This change was like a breath of fresh air for the KO after a lacklustre start to its campaign. A relatively young, energetic leader -who came in for ferocious attacks from the ruling party and Poland's public broadcaster -Trzaskowski managed to mobilise vast groups of voters, as evidenced by his success in collecting 1.6 million signatures in support of his candidacy in four days (Szczęśniak, 2020).
Trzaskowski's campaign focused on mobilising the pro-democratic electorate, which was eager for change after five years of the "United Right" in power. But he also addressed more moderate voters, presenting himself as a politician willing to speak with his adversaries, too -for example, by inviting strictly pro-government, right-wing media to ask him questions (Wiśniewski, 2020). Trzaskowski, who has served as Warsaw's mayor since 2018, offered an agenda of diversity and decentralisation, stressing the potential of local and regional governments. He also promised to improve public services and the state's commitment to environmental policies. Without forgetting about the limitations of the presidency in the Polish political system, he promised a presidency of openness and cooperation but based on clearly defined principles of democracy and the rule of law (Trzaskowski, 2020). His campaign did not emphasise his high level of education, knowledge of foreign languages and prior achievements in Polish politics but nor did his staff strive to downplay them, instead pairing his reputation as a worldly, "can-do" politician with an image of a committed activist and family man. Trzaskowski's wife, Małgorzata, was active in the campaign, promising to represent the interests of women (Czuma, 2020). The Civic Platform party organisation backed Trzaskowski's campaign, but he did not highlight his party affiliation.
While Trzaskowski presented himself as an open-minded modern politician, the right described him as a leftist extremist, promoting liberal ideals and supporting gay marriage. The issues of LGBTQ+ rights moved to centre-stage in the campaign, as the government and the public media presented the issue as an ideological threat to Polish traditions and Christianity (Kalisz, et al. 2021, pp. 51-53). This charge was leveraged to enhance the ruling party's image and its presidential candidate as defenders of Polishness against the liberal West and enforced tolerance. The other point of focus of Duda's campaign involved social transfers for families and the elderly. In his campaign, Duda acted on his image as the supporter of traditional families and small business, vowing to protect them from the excesses of modernity and globalisation (Duda, 2020).
The image of the right and Duda himself as defenders of the Polish tradition and Roman Catholicism was also enhanced by the Church. Despite the bishops' declarations of impartiality and no political involvement (Miziński, 2020), many priests spoke out to support the incumbent president or criticise his opponents, raising criticism of politicising the Church, especially among the Catholic intelligentsia in Poland (Boniecki, 2020). Church premises were used to display posters of Duda as well (Boserup, Andrusz, 2020, p. 14).
Another specific feature of Andrzej Duda's campaign was how state institutions put the thumb on the scale for him. The most important role was played by Prime Minister Mateusz Morawiecki, whose schedule was almost entirely dedicated to campaigning for Duda. Further, the Minister of Health Łukasz Szumowski (responsible for organising health care during the pandemic) and Minister of Development Jadwiga Emilewicz (in charge of anti-crisis measures during the lockdown and afterwards) planned initiatives and announcements that appeared designed to make the government and president look good. All governments, to some extent, take advantage of incumbency to maximise their electoral chances, but according to neutral observers, the situation in Poland breached all democratic norms (Boserup, Andrusz, 2020, p. 15). The conduct of the ostensibly neutral public broadcaster was no exception -even the supervising National Broadcasting Council, dominated by representatives of Law and Justice, criticised it for bias (Kublik, 2020).
The 2020 campaign was exceptional not only due to the organisational issues caused by COVID-19 and the government's policies. Another variation was the massive mobilisation of supporters of both major candidates, reflected not only in the record-high voter turnout on the election days themselves but also during the campaign. This was evidenced by high attendance at meetings with major candidates, as well as the blanketing of posters and banners seemingly on every available surface -fences, balconies, windows and even the facades of houses and flats. While the coronavirus pandemic reduced (or virtually eliminated) the distribution of flyers and electoral gadgets, citizen activity made the campaign very prominent not only in urban areas but also in rural areas and outside localities.
The overall image of the electoral campaign has to be viewed in a broader context of the democratic backslide under "United Right" rule. The verbal attacks against the LGBTQ+ community just one reflection of the vision of illiberal democracy implemented by the governing coalition since 2015. By presenting themselves as the protectors of traditional order against Western norms and standards of openness and tolerance, Duda's staff strived to enhance the ideological discrepancy between the pro-liberal democracy opposition groups and the government's supporters in an effort to mobilise the conservative part of the Polish society.
The course of the elections caused controversy. First, confusion in the organisation of the polls during the pandemic raised a lot of doubts surrounding the legality of the new election schedule and the late amendment of electoral provisions. Second, fairness and candidate equality were compromised by the untoward involvement of the government -including with state budget funds -in the campaign. Third, the bias and unfairness of the campaign information provided by the ostensibly neutral public broadcaster compromised citizens' rights to full and correct information. Fourth, many voters residing abroad did not receive the voting packages or were unable to return them to Poland and were thus deprived of their right to vote. These doubts were raised by opposition candidates and parties, as well as NGOs and observers of the Organisation for Security and Cooperation in Europe (OSCE), and also reflected in multiple objections against the result submitted to the Supreme Court. However, the court did not consider the objections (Michałowski, 2020) and declared the elections results valid and binding.

Results
In comparing the 2015 and 2020 presidential elections, a few salient changes can be identified. Voter turnout in the first round in 2015 was 48.96% and 55.34% in the second round; the corresponding figures in 2020 were 64.51% and 68.18%, respectively. Therefore, in five years, there was a significant rise in voters participating in the presidential election. A comparison of the second round of presidential elections in 2015 and 2020 (Table 2) revealed a slight drop (-0.5%) in support for Duda favouring his competitor. To compare if there were significant changes from 2015 to 2020 on the municipal level (N=2,418) in mean percentage support for second-round candidates, a paired samples t-Test was conducted (Table 3). The created model for Duda revealed (t=-32.7, p<.001) a significant rise from 2015 (M=57.90, SD=14.90) to 2020 (M=61.60, SD=14.60). In contrast, the created model for Komorowski/Trzaskowski (t=32.7, p<.001) revealed a significant drop from 2015 (M=42.10, SD=14.90) to 2020 (M=38.40, SD=14.60). It is also notable that the effect size of the difference in support between candidates changed significantly (d=+0.257), showing that growing support for Duda had also become more intense. The obtained results are interesting because, in 2020, Trzaskowski attracted a larger number of supporters (N=10,018,263) than Komorowski (N=8,112,311) in 2015. Still, on the municipal level, the average percentage support score for his candidacy was lower than Komorowski's. At first, it may seem that the obtained results are contradictory. However, when the percentage of municipal support was converted to binary coding, it was revealed that the number of municipalities with higher support for Duda (N 2015 =1,550, N 2020 =1,815) outnumbered those supporting Komorowski andTrzaskowski (N 2015 =866, N 2020 =600) by ratios of 2:1 and 3:1, respectively. The McNemar's test revealed significant (X 2 =193, p<.001) change from 2015 to 2020, indicating the notable difference in the number of municipalities supporting Duda.
Also, as shown in Table 8 and Table 9, most people supporting Trzaskowski were residents of large towns and cities, which are less numerous than medium-to-small towns PP 2 '21 and villages. Therefore, it may be concluded that the difference in the number of supporters and average mean municipal support for candidates showed that Duda did very well in the vast majority of villages and smaller municipalities. In comparison, Trzaskowski performed well in larger but significantly less numerous municipalities.
The first round of voting (Table 4) revealed that the two leading candidates were Duda (43.50%) and Trzaskowski (30.46%). The difference in the percentage number of supporters for both main candidates in the first round was significant (13.04%). Yet, it changed considerably in the second round, especially for Trzaskowski, who netted a substantial boost in support (+18.51%). The final difference in support between the candidates was 2.33%. Table 4 Percentage support for candidates in 1st and 2nd round of voting -presidential elections in Poland 2020 To explain and understand such a significant shift in support, we correlated the municipal percentage party support in 2019 and the municipal percentage presidential support using linear regression (Table 5). The constructed model for Duda in the 1st round was significant (F(6, 1300)=9,067, R 2 =.98, p<.001), and standardised estimates revealed significant positive correspondence with PiS (β=1.16, p<.001) voters. Further analysis revealed significant F(6, 1300)=3,192, R 2 =.94, p<.001) strong positive correspondence for Trzaskowski in regard to KO (β=0.74,p<.001) and SLD (β=0.38, p<.001) voters in the 1st round. The analogical constructed model regarding the second round of voting for Duda confirmed significant (F(6, 1300)=7,635, R 2 =97, p<.001) strong positive correspondence with PiS voters (β=0.77, p<.001) and for Trzaskowski (F(6, 2404)=12,409, R 2 =97, p<.001) revealed a very weak positive correspondence only with SLD voters (β=0.09, p<.01).
From the obtained data, it was possible to notice significant distinctions when comparing both candidates' potential party electoral support. There was a very strong positive correlation between PiS voters and support for Duda, clearly showing that supporters of the party were strongly mobilised to support Duda in both rounds. Trzaskowski, on the other hand, had strong positive support of the chosen parties in the first round. Still, in the second round, his supporters became so politically diverse that there was no clear positive connection between him and any specific party formation. Those results were partially confirmed by data gathered by Ipsos, which were focused on the diagnosis of second-round voters who voted for neither Duda nor Trzaskowski in the first round. A study by Ipsos showed that the majority of this electorate voted for Trzaskowski (Table 6). Table 5 Municipal party support in 2019 and presidential candidates support correspondence -presidential elections in Poland 2020  Based on the presented data, we concluded that the majority of voters who did not choose either Duda or Trzaskowski in the first round swung to Trzaskowski in the second round. Such a radical shift of diverse voters in favour of one candidate not only balanced the chances of election for both candidates but offers analysts the opportunity to explore the demographic traits of the two leading candidates' voters.

Party
Based on data collected by Ipsos (Table 7) in the first round of voting, we can sketch a profile of the typical voter for both Duda and Trzaskowski, who were the leading candidates. In the first round, most Duda voters were either farmers, workers, unemployed or pensioners. A vast majority of them held only primary or vocational education and were over 50 years old, living in villages and middle-sized towns. In contrast, most Trzaskowski voters were either employers or managers with higher education, representing a very diverse age group ranging from 30 to 50 years old and mainly living in the largest PP 2 '21 cities (Warsaw, Kraków, Łódź, Wrocław and Poznań). The typical voter profile for each candidate was thus starkly different. Similar comparisons were conducted for the second round of voting, and, once more, the analysis was based on data collected by Ipsos (Table 8). The data show that the typical Duda supporter in the second round had the same traits as in the first round. In other words, he or she was older, less-educated, in menial work, and from a relatively small population centre compared to Trzaskowski's voters.
The comparison of both voting rounds reveals that the general social profile of the voters supporting each candidate did not change significantly, and it clearly shows that Duda can be classified as the candidate of the old-age, slightly less-educated "working class" and Trzaskowski the candidate of young or middle-aged, slightly better-educated "employer class." Although there is no convincing empirical data, it may be stated -based on indirect premises -that those groups of voters may also present some ideological traits as well. It is not a new fact that older people tend to be more conservative, and the young are more liberal and progressive in their political attitudes. Moreover, people raised in large towns and metropolises have a more liberal and progressive attitude than people from small towns and villages. Also, people with higher education tend to be more liberal, open-minded, and progressive than people with lower levels of education. Based on those premises, it may be stated that voters supporting Duda may have a more conservative world view and those voting for Trzaskowski are more liberal/ progressive. Although it is not necessarily that simple, especially taking into account that in the second round, voters of Trzaskowski were composed of radically different groups of interests ranging from conservatives (Confederation, PSL), liberals (KO), radical leftists and socialists (SLD). From the raw data obtained during the analysis, the significant boost in Trzaskowski's support in the second round may have been provoked by a strong protest against electing Duda. Indeed, a considerable number of voters may have voted for Trzaskowski because they did not want to vote for Duda and had no other choice. With this perspective in mind, we assumed most Duda voters had a genuine desire to vote for him because they did not have the pressure to compromise their attitude, and their vote was fully intentional. When it comes to Trzaskowski's voters, matters are much more complex. It may be assumed that he represented the best possible "central" option. In other words, all those who supported him genuinely saw him as a better option than Duda and were strongly against Duda but had no other options.

PP 2 '21
These claims can be supported by post-election findings published by CBOS in their report regarding the motivation of voting on candidates in the 2020 presidential elections in Poland (Rogulska, 2020, pp. 2-4). From the report's analysis, there is clear evidence that Duda's voters were primarily motivated by a positive evaluation of his current actions and fulfilment of expectations (33%) and a high level of credibility (18%). Thus, they were genuinely motivated to vote for him, based chiefly on individual satisfaction with his performance in office. When it comes to Trzaskowski (Ibidem, 2020, pp. 6-7), the primary substantive (positive) motivations were the apparent need for change (12%), his appealing programme (11%), and the desire for a political counterbalance (10%). However, the most cited reason was open negative voting (17%) -in other words, voting for Trzaskowski to prevent Duda from winning the election. This perspective is confirmed by empirical findings suggesting clearly different roles and expectations of electoral groups toward their leading candidates.

Discussion
Voting could be summarized as "sincere" in the first round and as "strategic" in the second. However, in investigating the second round, the approach of Merrill and Grofman (1999, p. 2) offers insights. The authors analyse the case of voters who once the governing party but now view it as having stayed too long in power and strayed too far from its original policy platform. Thus, some voters might long for the status quo ante and prefer the opposition party, especially if it is perceived that the ruling party may drift toward even more radical policies in the near future. This approach considers not only the ideological proximity of voters and parties but also the direction of parties' future policy decisions, otherwise known as the directional approach. This directional approach can be used to explain the voting behaviour of the voters who in 2015 supported Duda and then in 2020 switched to his challenger Trzaskowski. According to the Ipsos exit poll conducted during the second round, Trzaskowski was supported by 8.6% of Duda voters from 2015, while the former was backed only by 2.8% of Bronisław Komorowski voters. Komorowski was the PO candidate in 2015.
The results of the elections and of our research illustrate the difference between sincere and strategic voting. Voters of candidates defeated in the first round mostly went for Trzaskowski in the second, regardless of the significant programmatic differences between candidates (for instance, Trzaskowski and Krzysztof Bosak). Thus, it can be argued some votes might be recognised as protest ones; instead, they were against Duda rather than being for his rival. This indicates that the presidential elections remained, as predicted early in 2020, a clash between the ruling right and its pro-democracy opposition. Developments related to the pandemic, above all multiple legal changes leading to rescheduling and reorganising of elections, did not transform the dynamics of the elections. It seems instead that the pandemic enhanced existing lines of division, forcing further polarisation between the two camps.
Analysing why Duda won and identifying the visible demographic differences between Duda and Trzaskowski's supporters requires a broader insight into the context of the electoral campaign, the conflict over the systemic changes introduced by PiS after 2015, and, finally, the unique conditions of the pandemic in 2020.

Conclusions
The results of the elections presented in the present article cast the distinction between sincere and strategic voting into sharp relief. The constituents of candidates who were defeated in the first round mostly swung their support behind Rafał Trzaskowski in the second one, even though there were often significant ideological differences between them. Thus, much of the second-round electoral behaviour was "protest voting," with many voters more opposed to Duda than being adamant in support for Trzaskowski. Duda was perceived as the representative of the United Right, responsible for the democratic backslide in Poland, and voting against him reflected citizens' attitude to the nonliberal democracy policies. While the election results show that the right is still able to mobilise a sizeable sincere electorate in support of its traditionalist anti-liberal agenda, it can marshall very few strategic voters, as evidenced by Duda's failure to significantly improve his result between the two rounds of voting.
The electoral results clearly show not only the ideological division but also the divide between voters living in big cities and rural areas. Such interpretation would refer to the research previously conducted by Kitschelt (1992, pp. 21-27), who hypothesised that party systems in Central and Eastern Europe are structured along the promarket/libertarian versus antimarket/authoritarian values. Similarly, Markowski (1997, p. 242) also proposed a two-dimensional mapping of this space, distinguishing the economic aspect (populism vs. liberalism) and the socio-cultural divide (secular libertarian cosmopolitan vs religious authoritarian nationalism).
Even if the studies of Kitschelt or Markowski (1999) reflect the situation in Poland in the twentieth century, it can be said that there is evidence that there are still overlapping cleavages. What we call the rural side of the division was represented in the last presidential election by voters from villages, farmers, more often with lower education, the elderly and also more frequent Catholic church-goers, since many Church representatives expressed their support for Duda. The other camp of this division, described as "urban," consists of inhabitants of big cities, with higher education, more often younger, employers, students and managers. Thus, considering the electoral campaign and its major points, two distinct voter profiles in Poland differ in social characteristics and political preferences.
The voters of President Duda are also supporters of the United Right government. Being more religious and conservative, they respond to the slogans involving the protection of traditional values of Catholicism and patriotism, ready to accept the democratic backslide if they perceive it as a guarantee of individual and collective security and prosperity. On the other hand, the opposition side supports liberal democracy, speaks up for minority rights, tolerance, and acceptance of more variable lifestyle choices. The importance of this value is evidenced by the flow of support from other non-United Right to Rafał Trzaskowski, as opposition against unsettling of the liberal democratic institutions was the shared platform of those candidates. Despite significant differences highlighted by the opposition candidates during the electoral campaign, voters of Szymon Hołownia, Władysław Kosiniak-Kamysz, and Robert Biedroń voted for Rafał Trzaskowski in the second round.
The percentage of Duda supporters remained the same in 2020 compared to 2015, but in the four regions which traditionally voted for Civic Platform and its candidates, the number of Trzaskowski voters decreased. Thus, we may conclude that the division between western and eastern Poland during presidential elections became less important. At the same time, we have witnessed the tendency observed earlier by Zarycki and Nowak (2000), who pointed out that urban-rural cleavage can be linked with ideological divides, i.e., city inhabitants are more liberally oriented. In 2020, the differences between eastern areas of Poland (historically occupied and affected by Russia), and western parts (historically associated with Germany), are of lesser importance than the multifaceted cleavage between the pro-liberal residents of cities and traditional conservative and more religious rural population. Whether this cleavage widens or is overcome, it will continue to play a critical role in Polish political competition.
The data that support the findings of this study are available from the Centre for Qualitative Research at the Institute of Political Science, University of Wrocław: https:// www.cequar.uni.wroc.pl/Datasets.
On behalf of all authors, the corresponding author states that there is no conflict of interest.