Polarization in the Finnish parliament (eduskunta)

According to the old Constitution, foreign policy was the exclusive competence of the President. The power was concentrated in the hands of the president also because of the importance of personalizing the regime to maintain friendly relations with the Soviet Union. After the end of the Cold War, the need for this decreased.

Now the President of Finland directs foreign policy, but does it together with the government and through the ministerial committee of the government. However, the Constitution does not contain information on what will happen if there is no consensus between the president and the government on the issue. In 2012, a new constitutional amendment was adopted, which introduced a mechanism for resolving the conflict through the position of the Eduskunta on this issue. Such a mechanism can be applied to a narrow range of foreign policy issues (for example, the ratification of international agreements). Finland's accession to the EU contributed to the narrowing of the president's powers in favor of the government. Sometimes it can be difficult to differentiate competences on EU issues (the competence of the Cabinet) and other foreign policy issues, since the EU influences national foreign policy (the competence of the President). In accordance with the amendment to the constitution of 2012 in the EU and in all activities associated with it, Finland is represented by the Prime Minister, not the President. The President is the Commander-in-Chief and decides on participation in conflict resolution and crisis management operations.

Also before the adoption of the new constitution, the president had many powers in the field of appointment, whereas now these powers are only narrowing down. The president has the right of a suspensive veto, but it can be overcome by parliament. The president can also grant full or partial pardon after receiving the Supreme Court's application.

Earlier the President was elected by a collegium of voters out of 300 members elected by the same proportional system as the preliminaries until 1982. Since 1988, the president has been elected for a six-year term and can be in power for no more than two consecutive terms. In 1994, the new electoral system was used for the first time: if the candidate receives more than half of the votes, he is elected to the presidency; if not, then two candidates who get more votes go to the second round, which is held on the third Sunday after the first round; the candidate who received a majority of votes is considered as the winner. There is a scatter of opinions regarding the powers of the president: citizens believe that it needs to be maintained and even increased, and the political elite advocates further reduction.

7.Consideration of the Finnish Political System's Parliamentarization within the Framework of Shugart and Carey's Model

In modern political science, it is common to divide the forms of government of democratic states into presidential, parliamentary and semi-presidential ones (R. Elgie, 1999). Now let us look at Finland in the framework of the model of Shugart and Carey (1992). In their work, they propose a typology of regimes based on two dimensions: “(1) the degree of separation of powers, and (2) the nature of the Cabinet”. Authors distinguish five types of regimes and describe in detail two of them, calling them the “ideal types”. The first one is presidential system, and the authors define it using the following provisions:

1. “the popular election of the chief executive; ?

2. the terms of the chief executive and assembly are fixed, and are not contingent ?on mutual confidence; and ?

3. the elected executive names and directs the composition of the Government.” ?

This definition indicates the separate sources of formation and maintenance of the executive and legislative authorities. They will also highlight the fourth criterion, which logically follows from the previous ones:

4. “the President has some constitutionally granted lawmaking authority.”

The last criterion does not affect the methods of formation and separate functioning of the authorities. But it is important because granting legislative powers to the President guarantees the implementation of the program, which was approved by the people in the voting for this candidate. Among other things, this definition clearly highlights the fact that responsibility for the executive branch of power in the presidential system rests with the President.

Further, the authors consider the premier-presidentialism. The first description of this kind of regime was carried out by Duverger (1980). He identified them as semi-presidents and looked at them as an intermediate form, as an alternation of presidential and parliamentary phases. Shugart and Carey say that this regime has its own characteristics and call the premier-presidentialism what Duverger called the semi-presidential regime. Despite the difference in approaches to this regime, Shugart and Carey recognize Duverger's definition of a clear and suitable one. Thus, the premier-presidential system has the following characteristics:

1. “the President is elected by popular vote; ?

2. the President possesses considerable powers; and ?

3. there also exist a premier and cabinet, subject to assembly confidence, who ?perform executive functions” (Duverger, 1980). ?

Unlike the presidential system, here the President is not a "supreme" carrier of executive power, he must coexist with the prime minister. The scope of authority of both can vary depending on the country and on the specific circumstances occurring in one country. This system also differs from the presidential system in that, under the conditions of the premier-presidential regime, the powers vested in the president do not necessarily have to be legislative. The powers associated with the formation of the Government are more typical. The President may have the right to nominate ministers and make appointments to non-governmental positions. Also, in premier-presidential systems, Presidents usually can dissolve the Parliament.

Shugart and Carey identified two ideal types of system, where the president is elected publicly. There are different variations of these types, but it's impossible to fix everything. Therefore, they consider the most frequent deviations. Thus, the authority to form the Government can be assigned to both the President and the Parliament. Such a regime in the typology of Shugart and Carey is called the president-parliamentary. And they define this system in the following way:

1. “the popular election of the President;

2. the President appoints and dismisses cabinet ministers;

3. cabinet ministers are subject to parliamentary confidence;

4. the President has the power to dissolve Parliament or legislative powers, or both.”

The main difference here is that under this regime equal powers are given to shift the members of the Government to the President and the Parliament. The Parliament's authority to form the Government means that the executive branch is deprived of independent sources of support. Also in many such systems, the President has the right to dissolve the Parliament. From the foregoing, it follows that the separate survival of the authorities under this system is absent.

So, Shugart and Carey suggest the following two-dimensional scheme, where the first axis shows the scope of the President's powers to form the Cabinet, and the second shows the degree to which the survival of legislative and executive branches are separated (Fig. 1).

Fig. 1 A typology of democratic regimes in two dimensions. Source: Shugart & Carey (1992, page 18)

The ideal type of presidential system is located in the upper right corner, since the President has the authority to form the Cabinet, and the two authorities have separate sources of survival. In the lower left corner is the second ideal type of premier-presidential system, where the Cabinet is usually formed by the Parliament, and the legislative and executive branches of government do not have separate survival, as often here the President has the right to dissolve the Parliament. The upper left corner is occupied by the president-parliamentary regime, since here the President has maximum power in the formation of the Government, and the joint survival of the branches of power is absent. There can often be a conflict because of the lack of a clear delineation of the powers of the Parliament and the President. Shugart and Carey say that under this regime the President is stronger than in the presidential system.

The assembly-independent regime occupies the lower right-hand corner of the scheme. Here the President does not participate in the formation of the Government, it is formed by the Parliament, and also there is a separate survival of the branches of power.

Also, the authors find in their scheme a place for the parliamentary regime, which is located between the premier-presidential and assembly-independent types. The president is purely nominal in nature, and on the scale of separate sources for maintaining branches of power the regime is in the middle, since in most parliamentary systems the Cabinet has the right to dissolve the parliament, but at the same time the Cabinet is the trustee of the parliament. Also, no institution, except for the parliament itself, can shorten the term of its office.

In the work of Shugart and Carey, Finland acts as a premier-presidential regime, but it directly on the border between the premier-presidential and the president-parliamentary (Fig. 2). This is not surprising, since Duverger attributed Finland to the most powerful semi-presidential systems in Western Europe and noted the great importance of the authority of the head of state. The work of Shugart and Carey was written in 1992, when the President also had more power over the Cabinet and during its formation, and until 1990 he had the right to dissolve the Parliament and to appoint new elections, after the amendment introduced in 1991, the Parliament could be dissolved at the initiative of the Prime Minister. So, the location of Finland in the scheme of Shugart and Carey is fully justified: a dignifiedly strong President and the absence of separate sources for the formation and maintenance of authorities.



Fig. 2 Separate survival and presidential cabinet power: a comprehensive typology of democratic regime types. Source: Shugart and Carey (1992, page 160)

Thus, the form of government of Finland, established by the Constitution of 1919, met the criteria of the semi-presidential system: the President was elected by the population, there was a dual character of the executive power (the President and the Cabinet), the President possessed relatively extensive powers and formed the Government that was responsible to the legislature. According to the J. Nousiainen, Finland was an example of a "balanced combination of parliamentarism and presidential power" (Nousiainen, 1994). Shugart and Carey note that the Finnish Presidents played the role of arbiter rather than led the Cabinet through the Prime Minister, regardless of the party composition of the Government. The "strength" of the President largely depended on the personality of the leader and the political situation. In the history of Finland, some Presidents were moderate in using their prerogatives, leaving the Parliament a large space for action. Nausiainen calls their “the presidents of the parliamentary type”, who conduct a "soft line". But most of the Presidents enjoyed the powers vested in them, and some of them, using the terminology of M. Duverger (1980), were omnipotent. The latter belonged to U. Kekkonen (the years of the presidency of 1956-1981), which in the literature was called "glutton." Moreover, as we have said in this paper, Kekkonen realized presidential prerogatives not only in foreign but also in domestic policy: he selected Prime Ministers, "pressed" on the parties for the purpose of creating coalitions, forced the Governments to resign, appointed nonpartisan presidential offices and dissolving Parliaments.

In this paper, we want to see if the location of Finland has changed in this scheme, and if so, how. After a detailed description of Finland's political system and its changes over the past 20 years, it can be concluded that with the coming to power of President M. Koivisto in 1982, the period of parliamentarization of Finland began. The main feature of this period was the gradual dilution of the presidential elements of the regime and the strengthening of the parliamentary components of the system. The process of parliamentarization took place in two dimensions: in the legal sphere and in practice. Constitutional and legal parliamentarization consisted in changing the legislation regulating relations between the President, the Government and the Parliament. In the 1980-1990s a series of amendments was adopted to the Constitutional Act of Finland, according to which the leading role in domestic and foreign policy was gradually transferred to the Government supported by the Parliament, while the prerogatives of the President were gradually reduced. The amendments concerned the procedures for determining the composition of the Government, the dissolution of the Parliament, the vote of confidence, the veto of the President, the rights of minorities in the Parliament, the role of the President and the Government in external relations, particularly in relations with the European Union. Two amendments were also introduced concerning the term of office and the procedure for electing the President: since 1987, the President has been unable to remain in office for more than two consecutive terms, and since 1991 he has been elected in direct elections. Thus, this also contributed to limiting the power of the President and gave more opportunity to become the President to a person not from the party elite.

8.Calculation of the polarization index for the Finnish parliament

A review of the Literature on Polarization

In recent years, economists and scientists from the social fields have begun to pay special attention to the measurement and evaluation of economic and social polarization. In the earliest studies, much attention was paid to inequality in the distribution of income, which was considered one of the main causes of conflict between groups in society. Indeed, the inequality of wealth can lead to a confrontation between the rich and the poor people. But to pay attention only to economic characteristics will not be correct, because society can also be divided into separate groups according to the non-economic characteristics of individuals, such as religion, ethnicity, nationality, etc.

Two main approaches are used in analyzing social diversity: fractionation measures (Chakravarty & Majumder, 2001) and polarization. The fractionalization is a characteristic of social decomposition in the groups on any grounds. This approach depends on the number of groups: the level of fractionalization the higher the greater the number of groups. Polarization, however, considers the degree of similarity between groups: polarization and the degree of disunity of a society are greater the smaller the similarities between groups. Polarization and fractionalization can be measured in different aspects of people's lives such as income, ethnicity, race, religion, nationality, education, political views, etc. Thus, the concept of polarization gives a deeper idea of the fragmentation of society and there are plenty studies devoted to polarization, all of them are grounded on two basic documents of Esteban and Ray (1994) and Wolfson (1994) which defined two main directions in measuring social polarization.

Social polarization can be defined in several ways, but the definition given by Esteban and Ray (1994) is now regarded as classical and will be used as the main one in this paper:

“Suppose that a population of individuals may be grouped according to some vector of characteristics into clusters, such that each cluster is very similar in terms of the attributes of its members, but different clusters have members with very dissimilar attributes. In that case, we say that the society is polarized”.

Hence, in the last two decades there were plenty of papers on social indicators which pay attention to the methods of polarization measurement. Starting with pioneering works written by Esteban and Ray (1994) and Wolfson (1994), opened two main strands of polarization's measuring, considerably amount of contributions have appeared.

Polarization often refers to the distribution of socio-economic characteristics (for example, income, consumption) and refers to the extent to which the society is grouped into several clusters. It can be said that in a polarized society incomes or other relevant characteristics are concentrated around two or more different poles. In accordance with the first scheme proposed by Esteban and Ray (1991, 1994), polarization is considered in terms of intra-group identification and alienation between groups. Esteban and Ray call their approach of identification-alienation scheme. In more detail, it is assumed that everyone in a society is more part of a particular group and identifies himself with it and at the same time feels alienated from other groups. The approach created by Esteban and Ray (1994) focuses on the growth of individual groups: the polarization increases if the groups become more homogeneous inside, more separated from each other and more equal in size. This approach is used, in particular, by Graden (2000), Zhang and Kanbur (2001), Duklo et al. (2004), Esteban et al. (2007).

Such phenomenon as polarization is connected, but, of course, conceptually differs from the concept of inequality. For example, it is easy to imagine a society that converges in income to two local poles (for example, "rich" and "poor"). When measuring such a society with standard inequality indices, it would demonstrate a reduction in the level of inequality, since the overall variance of income would decrease, but it would also become more grouped around the two poles and, therefore, more bipolarized. Hence, it is obvious that polarization of society is directly connected to social tension's appearance and with likelihood of social conflicts. Thus, necessity of quantification of tension in the society and forecast of social conflicts is solid reason for developing polarization measure methods.

Presumably, a society is divided into n groups by some kind of feature. The size of group (the share in respect of the whole society) is , where , and coordinate of i-th group is . On the basis of the identification-alienation model the following index has been received

(1)

Positive constants k and . The parameter describes the degree of ”polarization sensitivity” of the system. If and , ER is equal to Gini inequality index.

Moreover, there is Reynal-Querol index (2002) which widely used in studies of ethnic polarization also based on index of Esteban and Ray. ER index has a discrete metrics hence, .

Aleskerov and Golubenko (2003) proposed an alternative approach to measuring polarization. This approach was implemented in the elections to the Finnish Parliament in 1999 and 2003. This method is inspired by the idea of the static moment of forces - the physical characteristics of the mass distribution in the system. AG index is as follows

(2)

Here, is the center of mass.

In addition, in Lipacheva (2015) another modification of AG index was produced:

(3)

Lipacheva also noticed an important distinction between (2) and (3) is if , while . The modified index, whose limit tends to zero, is more convenient to use, since, for example, there are a large number of groups in many practical cases. Consequently, this indicator takes on a low value not only in the case of one group, but also in the case of a sufficiently large number of equal groups.

The second direction, where the Wolfson's work (1994) is taken as classical, draws the disappearance of the middle class - the phenomenon that characterized the US and Great Britain in the 1980s (Wolfson, 1994), (Jenkins, 1995) and measures how incomes focus around two opposite distributive poles center and, consequently, the middle class size decreases. This approach is often mentioned as “bi-polarization”. It is go without saying that a considerable and stable middle class is a reason for development of economic and policy (Banerjee and Duflo, 2008). Thus, growing level of bi-polarization can be a serious problem for policymakers. This approach was developed and extended by such authors like Wolfson (1997), Wang and Tsui (2000), Chakravarty and Majumder (2001), Rodriguez and Salas (2003), Chakravarty and Ambrosio (2010).

The Wolfson polarization index can be considered as a function of the inequality within groups and the inequality between groups. We want to generalize this approach and formulate a class of polarization measures in the following way:

,

where f is a monotone function with respect to all arguments, is a vector of specific characteristic (e.g. income, education, political views, etc.) for a considered society of size n divided into k given groups, is a measure of the size of the k groups, and and are measures of inequality between and within groups accordingly.

Most of the work on the subject of measuring polarization in society considered the case of a one-dimensional measure. In such studies, it was assumed that society is divided into groups according to one criterion. However, the structure of society is much more complicated, and there are often disagreements on many issues. Thus, it is necessary to develop methods to measure how society is polarized in accordance with a multitude of characteristics. An average success has been achieved in this area, but there are some studies of multidimensional polarization such as Gigliarano and Mosler (2009), Bossert, Chakravarty and D'Ambrosio (2013), Nolan and Whelan (2007).

Poole and Rosenthal in their work “The Polarization of American Politics” [1984] developed an approach to the assessment of polarization in the US Congress. This approach is based on the using of well-known NOMINATE estimates, which was developed by them earlier (1983) and represent the two-dimensional coordinates of the Congressmen in the latent political space. The first dimension reflects the "liberal-conservative" dimension, while the second dimension refers to regional differences and to the acute problems of the day (e.g. civil rights, lifestyle issues, etc.) (Poole, Rosenthal, 1997).

Poole and Rosenthal suggested to measure polarization as distance between the average values of the Republicans and the Democrats of the first dimension co- ordinate. In McCarty et al. (2006), the same approach was applied for the newly created joint scales DW-NOMINATE. The most recent results of the polarization assessment in the US Congress are published on the Voteview website.

The main drawback of the approach proposed by K. Poole and H. Rosenthal is the fact that when the coordinates of congressmen are averaged, information on how scattered or dense are the clusters representing legislators is lost. Although this can influence the polarization in Congress (Aleskerov & Oleynik, 2016). The multidimensional index that is used in this study is based on the central moment of forces and does not have such a defect.

The Model

An extension of the one-dimensional modal presented in Aleskerov and Golubenko (2003) to the multidimensional one was proposed in Aleskerov and Oleynik (2016). The Aleskerov-Oleynik model is used in this study to calculate the polarization of the Finnish parliament. For this reason, let's pay more attention to description of this index.

As far as we know, Aleskerov-Golubenko index of polarization (AG index) is the first attempt to apply to the measurement of the polarization of society a concept from physics - it was originating from the concept of the central moment of the system of forces. The framework proposed by Aleshrekov and Golubenko is new and alternative to the two main directions for measuring polarization. However, the measures developed within this model meet the reasonable requirements originally established by Esteban and Ray (1994).

Imagine a society divided into n groups according to specific criteria. Bearing in mind the above definition of social polarization, we assume that there are certain features of society that determine the similarities and differences between individuals and, hence, groups of individuals. The vector of characteristics, in accordance to which individuals are grouped into clusters, is multidimensional, since we consider the multidimensional case.

We have multidimensional space , where m is a quantity of characteristics of group in the given society. Each group is characterized by a number , and a vector , . A number responds to the share of group's members in the considered society, . Hereinafter, the authors assume that each of the m characteristics can be represented as the value of some scalar variable taking the value from the interval [0; 1]. Thus, let's consider “opposite - government” scale where 0 coincides to extreme opposite and 1 - to extreme government, appropriately. Therefore, , , is a point in multidimensional representing the normalized positions of the group in the corresponding dimensions. All groups together form a system of weighted points in a multidimensional unit cube.

A center of mass of the points' system where weights are focused respectively is determined as

Whereupon, the multidimensional index of the given society's polarization can be presented as follows

where k is normalizing coefficient which is picked out in order to that the maximum of P index being equal to 1 and is distance function. Aleskerov and Oleynik (2016) propose three variants of the index given above depending on the used metric d: Euclidean, Manhattan and Chebychev distances.

If с is given as Euclidean metric, then polarization index takes a following type

.

If с is given as Manhattan metric, then polarization index takes a following type

.

If с is given as Chebychev metric, then polarization index takes a following type

.

If we use the improved Aleskerov-Golubenko index, which was given in Lipacheva (2015), then the following index variations will be obtained depending on different metrics:

Here the index d ' becomes zero in the case of one group and decreases to very low values if the groups' number in the society is very large. in Aleskerov and Oleynik (2016), the assumption that society can still be under tension even if the number of groups is large; then polarization still doesn't decrease to zero. Also, the authors say that large values of the index of polarization simplify the comparison of results and general analysis. Hence, in the empirical part of this study, it was decided to use the original form of the index, following the same arguments as Aleskerov and Oleynik.

Description of the Data

We express our deep gratitude to Antti Pajala Antti Pajala, University of Turku - Department of Political Science and

Contemporary History, Turku, Finland, e-mail: anpaja@utu.fi for providing the data for this study.

To begin with, the Finnish data needed for the scaling analyses is publicly available and includes 1983-2014 parliamentary years. http://www.fsd.uta.fi/aineistot/luettelo/FSD2117/meF2117.html The data was formed by using Poole and Rosenthal scaling techniques. Since the unity of party electoral groups is very high, it produces a “sagging problem” as Poole and Rosenthal call it. Thus, NOMINATE cannot be used in the Nordic countries because this approach needs lower group cohesion, like in the US Congress. As an alternative, Poole's Optimal Classification algorithm was implemented.

In Pajala (2012) Pajala, A. 2012. Eduskunnan ristiriitaulottuvuudet tдysistuntoддnestysten valossa vuosien 1991-2010 valtiopдivillд. Politiikka 54(2), 103-118. the Parliament's political dimensionality in the light of plenary votes during 1991-2010 period was investigated on the same data. The article presents eigenvalues for the 20 most significant dimensions. But it was established that only the first dimension is meaningful while the rest are marginal. The first dimension is the government - opposition divider and for the second dimension a reasonable interpretation was not found. Antti Pajala expected the second dimension reflects the left-right setting of the party groups, but it did not. Pajala was interested in seeing when the left-right division disappeared. He expected the left-right voting had ended in the 1980s, but in the Pajala et al. (2015) El Hannari, N., Pajala, A., Raiskila, M., Suojaranta, T. ja Vainio, J. 2015. Ryhmдyhtenдisyyden vakaus: Hallitus-oppositio -ддnestдminen eduskunnassa 1980-luvun parlamentaarisessa murroksessa. Politiikka57:2, 126-136. it was found that in the 1980s this dimension was already absent. Thus, this division disappeared in the 1970s, since in the late 1960s this division still took place.

Thus, we were provided with data computed using Poole's Optimal Classification algorithm implemented in the statistical package R. The values are within a unit circle rather than a unit square. Poole's Optimal Classification algorithm is carried out as follows:

Hence, we have data for the Finnish parliamentarians for the 1983-2014 period in Excel format, it includes 8 parliamentary convocations. There are MPs' names, their party affiliation and for some terms election districts. Here are two most explanatory dimensions: the first dimension is the abscissa (Dim1 = x-coordinate), expressing government - opposition dimension and the second one is the ordinate (Dim2 = y-coordinate), but it is not installed what it expresses.

Empirical Applications. Case of the Finnish Parliament (Eduskunta)

Eduskunta of the 1983-1986 convocation (Sorsa's third Cabinet, SDP)

The elections in the Eduskunta of the 1983-1986 convocation took place according to the proportional system. The following parties passed in the Parliament:

- The Social Democratic Party of Finland (SDP), 58

- The National Coalition Party (Kok) 45

- The Finnish Center (Kesk) 39

- Finnish People's Democratic League (SKDL) 32

- The Finnish Rural Party (SMP) 18

- The Swedish People's Party (RKP) 12

- Democratic Alternative (Deva) 10

- The Finnish Christian League (SKL) 3

- The Green Union party (Vihr) 2

One of the features of the Eduscunt of the 1983-1986 convocation was a strict party discipline. Fig. 3 clearly demonstrates this feature: clusters representing deputy parties are rather dense. Some deputy unions can be described by a more coherent consensus of opinions, their clusters are denser on the political map: SDP, Kesk, RKP and Deva. At the same time, the deputy groups of Kok, SKDL and SMP represent more scattered clusters, in particular, this applies to the last two parties: here there is the widest scatter of opinions within the party. About the consistency of opinions in other parties (Vihr, SMP) it is difficult to say something, since only few of their representatives passed into parliament. It can be said that a group of 4 parties (SDP, Kesk, RKP and SMP) shows a rather high degree of consensus in opinions.

From 1979 to 1982, a coalition of four parties (center and left) was headed by Mauno Koivisto. In 1982, President Urho Kekkonen resigned, and in his place was elected social-democrat Mauno Koivisto. Social-democrat Kalevi Sorsa again rose up at the head of the government and he reformed the Cabinet with the same four-party coalition. Soon representatives of the SKDL came out of the Cabinet because in October 1982, two devaluations were carried out, which amounted to a 10% reduction in the value of the markka (Jonung et al, 2009), this measure caused complaints from SKDL that the main victims of this measure were low-income groups.

Fig. 3. Spatial model of the Eduskunta 1983-1986s

According to David Arter (1983) The 1983 elections were widely viewed as "protest elections", because the main parties, with the exception of the Social Democratic Party (SDP), lost seats in the parliament, for example, the Liberal People's Party (LKP) lost all its seats in the Eduskunta, and the Finnish Rural Party (SMP), vice versa, more than doubled the number of seats; the Green Union party (Vihr) got seats for the first time. Three parties (SDP, Kesk, RKP), having won a majority, re-formed the government in 1983; also SMP joined the coalition for the first time. It can explain why positions of these four parties are so close and why the SKDL has the opposition place on political map for 1983-1986 period (Fig. 4). On the axis we are considering (the x-axis that is significant for us, which stands for government-opposition dimension), only the MPs from the SKDL, the Deva and the Vihr occupy strictly oppositional positions, but this group is less than one quarter of parliamentarians, so we can say that the parliament looks fairly homogeneous.

Fig. 4. Positions of the deputy unions in the Eduskinta in 1983-1986s

The coordinates of the center of mass show that on average the deputies held rather moderate positions along both axes. The polarization in the parliament of this convocation can be called quite high, which can be explained by the presence of obvious opposition to the government in the person of SKDL, Deva and Vihr; also reinforces the value of the polarization index of the demarcation of parties along the second axis, the significance of which we do not know. Presumably, this disengagement in these years could be explained by economic problems and disagreement in the parliament over economic legislation. Despite some inconsistency, it can be said that the consistency of government parties allowed the Сabinet to serve a full term (the first Сabinet that served a full term since the Second World War), its survival before the elections in March 1987 became a sign of the recently gained stability in Finnish politics.

Eduskunta of the 1987-1990 convocation (Holkeri's Cabinet, Kok)

Although the previous government was rather coherent, it was not free from tension. In 1984, disagreements arose among parties, the three socialist parties disagreed with the SDP on a number of internal issues, such as the feasibility of building nuclear power plants, financial measures to help farmers and small entrepreneurs. In 1986, the SDP party was attacked by Kesk, which was due to the fact that Kesk changed direction to the right and wanted to organize a center-right government after the 1987 elections (Berglund, 1987).

In the Eduskunta of this convocation all the same parties were presented as in the previous period. We will consider the 9 following deputy unions represented in the Eduskunta of 1987-1990s:

- The Social Democratic Party of Finland (SDP) 56

- The National Coalition Party (Kok) 53

- The Finnish Center (Kesk) 40

- Finnish People's Democratic League (SKDL) 20

- The Swedish People's Party (RKP) 13

- The Finnish Rural Party (SMP) 9

- The Finnish Christian League (SKL) 5

- Democratic Alternative (Deva) 4

- The Green Union party (Vihr) 4

Fig. 5-6 demonstrates that positions of the parties have significantly changed compared to the previous parliament. Firstly, Kesk along the government-opposition axis has moved from the government party to open opposition. As Fig. 5 shows there is also a fairly strong range of opinions within the parties as compared to the previous convocation. This is because the elections to the Finnish Eduskunta, which took place in March 1987, were a turning point in post-war Finnish politics. Since these elections contributed to the end of the practice of excluding the Conservative Party from the list of potential government parties.

Fig. 5. Spatial model of the Eduskunta 1983-1986s

Even during the campaign period, the leader of the Social Democratic Party Kalevi Sorsa remarked in his speeches that he was more supportive towards the Conservative Party than the Finnish Center party, and during his speech in Helsinki in late 1986, he suggested that the Center and the Conservative parties exchanged seats in the Finnish left-right space. Actually, the demonstrated state of affairs was already predetermined until the elections: there was no hope for the continuation of the red-green (SDP-Kesk) coalition government (Berglund, 1988). Therefore, this parliamentary convocation was interesting because an atypical right-left government coalition was formed for Finland, including such large parties as Kok and SDP, as well as RKP, and earlier right-wing but now moderate SMP. Although the SDP lost seats in the 1987 elections, it still held a leading position in government. There were dissatisfactions and opinions that President Koivisto, as a former member of the SDP, could abuse his powers during mediation in forming a government. Nevertheless, the positions of government parties are very close to the political map and there are several explanations for this. First, the views of the Socialist and Conservative Party of Finland themselves were agreed upon in many issues, for example, the need to reduce agricultural subsidies (which always was defended by centrists). Secondly, the parties were striving for the same goals: increasing Finland's competitiveness in the international market, maintaining the social security system (Karvonen and Rappe, 2006). Finally, the SDP and Kok were agreed in a hostile mood to the leader of the Center Party Paavo Vдyrynen and did not want to create a coalition with centrists.

Fig. 6 Positions of the deputy unions in the Eduskinta in 1987-1990s

The center of mass again close to zero, practically did not move from the previous position. This means that on average, there are no sharply pro-government or opposition sentiments in the parliament. As for the polarization index, it grew even more than the polarization index of the parliament of 1983-1986s. This can be seen from Fig. 5-6: mostly prosocialist forces are in opposition to government parties, also the Green Party has disagreed with the bloc of government parties on issues of nuclear energy and environmental protection. However, it is noticeable from the Fig. 3 that the parties of the opposition bloc are less coherent. Another explanation for the polarization of the factions may be the fact that in June 1990 Finland became a participant in the negotiations in Brussels on accession to the European Economic Area. There was a split in the parliament during this period because of the issue of Finland's membership in the EEC, the results of the parliamentary vote indicate the split: out of 200 deputies only 108 voted to file an application (44 against) (Aunesluoma, 2011). The country was divided into two camps. The camp of opponents of European integration included scattered secondary political forces without a strong leader: the Finnish Center, farmers and the rural population, the Union of Left Forces, environmentalists, women's organizations. Supporters of the accession represented by the Social Democrats, the National Coalition Party, the Swedish People's Party were better organized and relied on the support of industrial and commercial circles, the urban population of southern Finland (Tiilikainen, 1996).

Eduskunta of the 1991-1994 convocation (Esko Aho's Cabinet, Kesk)

In accordance with the results of the national elections held in March 1991, the eight following parties won parliamentary seats:

- The Finnish Center (Kesk) 55

- The Social Democratic Party of Finland (SDP) 48

- The National Coalition Party (Kok) 40

- The Left Union Party (Left Alliance, Vas) 19

- The Swedish People's Party (RKP) 12

- The Green Union party (Vihr) 10

- The Finnish Christian League (SKL) 8

- The Finnish Rural Party (SMP) 7

In the political arena of Finland, a new party, the Left Alliance, formed by the union of SKDL, Communist Party of Finland (SKP), the Finnish Women's Democratic League (SNDL), and also communists from the disintegrated Democratic Alternative (Deva) joined new party. This period was characterized by economic depression and high unemployment in Finland. The government took austerity measures, such as cutting public spending (Jonung, Kiander, & Vartia, 2008).

Figures 7 and 8 show that in the new convocation a new configuration of parties was formed. After the parliamentary elections, the government was formed by the next coalition of parties: Kesk, Kok, SKL and RKP. It was the first cabinet in a very long time, in which the socialist parties did not enter. Thus, this government can be called center-right.



Fig. 7 Spatial model of the Eduskunta 1991-1994s

The center of mass is again almost at zero, which means that on average the positions of the factions are moderate, there is no distortion in either side (this is achieved by a too radical but comparatively small number of votes of opposition representatives and a more moderate but larger number of votes of government parties).

Fig. 8 Positions of the deputy unions in the Eduskinta in 1991-1994s

As for polarization index, in this case it has increased again. Index calculated with help of Euclidian metric shows that polarization is about 0,8, that is very high. Thus, calculations confirm the location of parties on the political map.

Eduskunta of the 1995-1998 (Paavo Lipponen's first Cabinet, SDP) and 1999-2002 convocations (Paavo Lipponen's second Cabinet, SDP)

The previous center-right, non-socialist cabinet could not continue to exist because of the adopted unpopular program of a tough economy and a severe recession of the 1990s. And in the 1995 elections, the Social Democratic Party (SDP) won 63 of the 200 seats in Eduskunta, achieving the best result of any party since the WWII. After that, the SDP leader Paavo Lipponen formed a coalition government that went down in history under the name of the "rainbow government" because of the many parties that were included there: the SDP, the National Coalition Party, the Left Alliance, the Swedish People's Party and the Green League (which became the first green party in Europe taken the seats in the government). After the 1999 elections, the government continued its existence in the same composition. The new Finnish government was also interesting because there had never before been a cabinet of cooperation between the main right-wing party and the radical left (Arter, 1995).

The following parties have seats in the Eduskunta during these two convocations:

- The Social Democratic Party of Finland (SDP)

- The Finnish Center (Kesk)

- The National Coalition Party (Kok)

- The Left Union Party (Left Alliance, Vas)

- The Swedish People's Party (RKP)

- The Green Union party (Vihr)

- The Finnish Christian League (SKL)

- The Young Finnish Party (Nusu, only in 1995-1998 convection)

Fig 9-10 demonstrate high cohesion between parties in parliament. This fact is explained by the presence of a five-party coalition government that was formed under the balance of left and right forces, and the SDP, which had almost one third of seats in Eduskunta, agreed that they will not receive the majority of seats in the government.



Fig. 9 Positions of the deputy unions in the Eduskinta in 1995-1998s

The government needed to address pressing economic problems, do not waste time on the disputes and quickly find solutions to fight high unemployment and pay a large public debt. However, the map (Fig. 10.) shows some scatter of opinions of government parties, which shows the presence of different opinions on the solution of problems. For example, not all in the government supported the proposal to reduce child allowances, unemployment benefits and a reduction in housing subsidies. Moreover, within the party coalition there was a disagreement over the income tax: before the elections, the Left Alliance proposed the possibility of raising taxes as a means of protecting the social security system, while conservatives favored a reduction in income tax by 1% per year.

The topics raised during the 1999 election campaign differed little from those discussed in the 1995 elections: unemployment, cuts in public sector spending, income taxes and labor market reforms. Throughout the campaign, the positions of the parties were virtually indistinguishable. There were no significant differences on foreign policy issues either.

Fig. 10. Spatial model of the Eduskunta 1995-1998s

In accordance with the results of 1999, the support of the government left parties (VAS, SDP) declined compared with the results of 1995, and support for the bourgeois parties (Kesk, Kok, SKL) increased, so did the support of the Green Party. The winners in 1999, which received more seats than in the previous parliament, were Kok, Kesk, Vihr and SKL, parties could form the majority government. However, President Martti Ahtisaari (a former member of the SPD) intervened to allow the PSD to enter the government and lead it, the president invited Paavo Lipponen, who led the past government, to negotiate to form a majority government. Thus, after the election, the “rainbow government” led by Lopponen remained in power (Nurmi, H., & Nurmi, L., 2001). By 1999, the President was still an influential figure, but with much less authority than 10 years ago, it so happened that now the parliament formally became the place of implementation of political power, but in practice it is the Cabinet that sets the pace of life between the parliamentary elections. Government parties show high cohesion (Fig. 11), also the distance between government and opposition parties have decreased in comparison with previous elections which were before 1995 election (Fig. 12). It should be noted that the Green Party withdrew from the party coalition after the government approved the construction of the fifth Finnish nuclear power plant in May 2002 and took the position between the government parties and the Center Party - the core of the opposition to the left-right government.

Fig. 11. Spatial model of the Eduskunta 1999-2002s

The center of mass, explaining the general character of the parliament during the time of the parliaments of 1995-1998 and 1999-2002, remained at about the same point. As for the value of the polarization indices, they significantly decreased. The formation of a multi-party government has indeed led to a reduction in discord among the factions. Thus, the government was supported by the majority of parliament.

Fig. 12. Positions of the deputy unions in the Eduskinta in 1999-2002s

Eduskunta of the 2003-2006 convocation (Anneli Jддtteenmдki's Cabinet April-June of 2003, Kesk), (Matti Vanhanen's first Cabinet, Kesk)

In 2001, the the Finnish Christian League changed its name to the Christian Democrats, also in Eduskunta the Finns Party (the True Finns) appeared in 1995, which is follower of the Finnish Rural Party, dissolved because of financial problems.

On March 16, 2003, elections were held, which were the first to be held in accordance with the new constitution, which came into force in 2000, and "brought Finland closer to a standard form of parliamentary democracy" (Raunio, 2003). The old constitution, with a Finnish fragmented multi-party system in which there is no clearly dominant party, strengthened the president's hand in managing the negotiations (although, the president has not had a strong influence on the formation of the government since 1987). In accordance with the new constitution, the president is now in the background, and the largest party plays a leading role in shaping the government. It was expected that the elections would now be more focused on candidates for prime ministers.

After the 2003 elections, the Finnish parliament passed:

- The Finnish Center (Kesk) 55

- The Social Democratic Party of Finland (SDP) 53

- The National Coalition Party (Kok) 44

- The Left Union Party (Left Alliance, Vas) 19

- The Green Union party (Vihr) 16

- The Swedish People's Party (RKP) 11

- The Christian Democrats (KD) 6

- The Finns Party (PS) 3

The largest number of seats in parliament after these elections belonged to the Finnish Center, having intensified its support in the south of Finland. The National Coalition Party lost in its support in large part due to the fact that attention was paid to the leaders of the Finnish Center and the Social Democratic Party. In general, the election campaign was characterized by moderate party positions, as parties were cautious in their statements and speeches in order to be able to get into the coalition government. Thus, electoral debates were focused on traditional Finnish questions, such as the state's ability to provide public services and the overall effectiveness of the welfare state (Raunio, 2003). After the election, the leader of the Center's party formed a coalition government, which included Kesk, SDP and RKP, it declares the location of these parties on the political map (Fig. 13). However, this office lasted only 69 days and was dissolved after the dismissal of the prime minister Anneli Jддtteenmдki, who was accused of lying to parliament and the public about receiving secret information from the Foreign Ministry (the Iraqi scandal). The new cabinet was in the same composition was formed by Matti Vanhanen from the Party Center.



Fig. 13 Positions of the deputy unions in the Eduskinta in 2003-2006s

The National Coalition Party, not included in the government, occupies an average position on the government-opposition axis. There is also a decrease in the cohesion of opinions within the party, this can be explained by the lack of a strong leader (Fig. 14). Although it is impossible to say what exactly shows the y-axis in this case, it can be assumed that the party is lower than the rest, since it was the only party advocating a tax cut. The directly opposite view of the legalization of drugs by the Green Party (for) and the party (against) can to some extent explain their distance from each other on the map. The same position of Vihr and KD along the x-axis can be justified by the same vote on issues pertaining to the protection of environment. As well as Vas and PS have similar positions along the same axis because of their Euroscepticism (Fig. 13).

...