Scholars and journalists argue that the EU is transitioning towards a value union. If true, this value convergence is happening along the EU enlargements, warranting the research questions: To what extent have the political values of EU citizens converged across member states since the 1970s and has this trend been influenced by the enlargement rounds of the EU? Relying on longitudinal surveys and political values derived from key EU documents, a quantitative intertemporal convergence analysis on eight political values leads to four conclusions: First, no overarching convergence trend is observable as four values diverge across EU countries. However, divergences are more profound than convergences. Second, the original members of the EU constitute the most homogeneous subgroup analyzed. Third, the EU enlargement rounds do not necessarily increase value differences between EU countries. Fourth, new EU members sometimes adopt the predominant attitude of the other EU countries.
Introduction
During its various integration steps, the European Union evolved to be a unique cooperation project between 27 nations. Several scholars1 argue that this process was accompanied by a shift from an economic union, promoting trade between the six founding members, towards a political union of values as an increasing amount of policy areas were incorporated in the responsibilities of the EU. However, even while advocating for the first predecessor organization of the EU, Robert Schuman2 declared peace an inherent reason for the founding of the European community. Hence, certain political values accompanied the EU to this day. Furthermore, the increasing importance of these values is exemplified by official statements,3 media discussions,4 and academic research5. Especially, scholarly contributions stress the role of common values in overcoming the internal difficulties of the EU6. In contrast, current conflicts regarding the linkage of EU budget funds to rule of law conditionalities exemplify possible value rifts inside the EU7. The fact that the two main counterparties, Poland and Hungary, entered the EU as part of the EU’s enlargement towards the east in the early 2000s exemplifies long-lasting concerns8 over a deterioration of EU values resulting from the EU enlargement rounds.
Past empiric academic contributions have tried to identify these value convergence patterns in the EU and sought to determine the influence EU enlargement rounds have on these trends. Whereas landmark studies by Inglehart and Baker,9 Haller et al.,10 or more recently Akaliyski11, confirmed different value clusters within the EU, several studies analyzed the convergence of values over time. However, these studies are limited in the number of countries incorporated, the period covered, or the values analyzed. For instance, whereas Houwelingen et al.12 only look at 15 countries from 2002 to 2016, Akaliyski13 focuses on general cultural and behavioral values instead of political ones inherent to the EU idea and structure. Additionally, various studies solely focus on the cross-country development of select values, like the acceptance of homosexuality14 or the free movement of labor.15 Furthermore, Oshri et al.16 already focus on one key political value within the EU, democratic identity, and conclude that democratic values tend to increase over time in EU member states. In contrast, Gerhards17 maintains that each EU enlargement round increases the differences in values across the EU. However, similar to the other mentioned contributions, this study mainly focuses on cultural instead of political values. Therefore, the current literature lacks a longitudinal cross-country convergence analysis of codified political values across all EU countries. This gap in the academic discussion warrants further research into the convergence of key political values in the EU and leads to the following two research questions: To what extent have the political values of EU citizens converged across member states since the 1970s and has this trend been influenced by the enlargement rounds of the EU?
The methodology to determine, if political values converged across EU countries along the enlargement rounds, is described in detail in the second chapter. Additionally, this chapter introduces the main sources for the used value variables, namely the European Values Study (EVS), the Eurobarometer (EB), and the International Social Survey Programme (ISSP). Moreover, the chapter recapitulates the EU enlargement rounds to establish the timespans and groups of EU members that require an extensive analysis focus. Furthermore, the analyzed political values are derived in the third chapter. This process focuses on significant EU documents establishing EU political values and enabling the enlargement rounds. In the fourth chapter, the analysis of possible value convergences relies on the matching of these derived values with the mentioned value variables from the different surveys. Subsequently, the value indicators are operationalized referring to these matched value variables. The resulting analysis is structured into the general development of the political values and their development along the enlargement rounds.
Clarification of Terms and Methodology
Political Values
Research in social sciences has not produced a uniform definition of values, mainly due to difficulties of directly measuring or observing values. However, it is commonly accepted that values describe something more elementary or basic than, for example, beliefs or opinions. Hence, values underlie these parts of human behavior. Although this definition is more functional than purely definitional, it allows the indirect analysis of values by referring to the attitudes, opinions, and behaviors of citizens as explained in chapter 2.3 and carried out in chapter 4.
As values, in general, lack a common definition, the same applies for the subset of political values. However, adhering to the functional definition introduced in the prior paragraph, political values are values that are concerned with the political sphere. Hence, this category of values can be derived from citizen’s political behavior and their perception of certain political events, orders, and situations. For instance, how a citizen might rank different government systems could reveal their value of democracy. Consequently, the indicators derived in chapter 3 seek to match these definitions to observe the political values of EU citizens.
EU Enlargement Rounds
The EU enlargements describe the widening of the EU integration process to countries beyond the six founding members. Furthermore, during the EU enlargement rounds, states with shared traits have gained accession to the EU at the same time. These shared characteristics further warrant a selective analysis as they imply shared political values. Whereas different scholars have proposed a variety of nuanced definitions and taxonomies, this research paper follows Nugent’s naming and structure proposal, reflected in Table 1.
Table 1 : EU Enlargement Rounds25
| Enlargement Round | Year | Member States Acceding |
| First | 1973 | Denmark, Ireland, United Kingdom |
| Mediterranean (1st phase) | 1981 | Greece |
| Mediterranean (2nd phase) | 1986 | Spain, Portugal |
| EFTA | 1995 | Austria, Finalan, Sweden |
| 10+2 (1st phase) | 2004 | Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, Slovenia |
| 10+2 (2nd phase) | 2007 | Bulgaria, Romania |
| Balkan | 2013 | Croatia |
Quantitative Longitudinal Convergence Analysis
The quantitative analysis of convergence has long been established in growth economics, where two types of convergence measurements have been developed. Whereas β-convergence refers to the catching up of poor regions to richer ones, σ-convergence refers to the reduction of disparities between regions.26 However, these methods can be applied beyond the analysis of income, equality, and other economic indicators, as proven by Hancke and Axisa.27 Researched in cooperation with the EU, their paper analyzes the convergence of social values in 13 EU countries between 2003 and 2015 using σ-convergence. Partly referring to this approach, this research paper employs σ-convergence to measure the convergence of political values across EU countries. Furthermore, σ-convergence allows the measurement of convergence using fewer observations,28 which makes it suitable for this research paper as the number of observations is inherently limited at 28 EU countries.
In the first step, common political values are inferred from key EU documents. Afterward, these values are matched with available variables in the EVS, ISSP, and the EB. In case these variables were surveyed on a nominal or ordinal level of measurement, they are transformed onto an interval scale by referring to the share of respondents choosing certain answer options. This step is necessary as the computation of σ-convergence is not possible for variables on nominal or ordinal scales as these scales are not suited for a mean calculation. Subsequently, value indicators are derived from the identified variables. In case indicators are matched to more than one variable, these variables must be weighted and averaged for the versions of the public opinion surveys, in which they are available. Moreover, values are omitted from the following analysis if no proxy is derivable for them from the public opinion surveys. Consequently, indicators for the remaining political values are available on a positive interval scale and can be computed for each EU country, which participated in the respective versions of the public opinion surveys. Finally, the σ-convergence of these operationalized indicators is determined.
Usually, σ-convergence is measured by using either the standard deviation or the coefficient of variation,29 exemplified by the use of standard deviation in the mentioned research paper by Hancke and Axisa.30 The coefficient of variation is determined by dividing the standard deviation with the mean and, hence, reflects the variance of observations proportional to the mean of these observations. Due to the various designs of the public opinion surveys large differences between the possible means of answers are possible. Hence, the relative measurement of the coefficient of variation offers results that are comparable across values and time free from scale effects.31 Additionally, the coefficient of variation allows for simpler assumptions regarding the significance of the observed convergences. However, due to the proportional characteristic of the coefficient of variation, this method of measuring convergence is more sensitive to values near zero as small changes in the variables are reflected in larger proportional changes. Nonetheless, the benefits of the coefficient of variation and the specific application in this research paper prevail this caveat. Hence, the coefficient of variation is used to measure convergence. Consequently, a decrease in the coefficient of variation of the operationalized value indicators over time signifies convergence while a rise in the coefficient of variation indicates divergence.
Data sources: European Values Study, Eurobarometer, and International Social Survey Programme
Due to the chosen scope of this research paper, public opinion surveys consulted need to maintain consistent question and variable formats over a long period, while also surveying the respective EU countries for each version in the respective years. The Eurobarometer series, the European Values Study, and the International Social Survey Programme meet these criteria.
First, the Standard Eurobarometer was established in 1974 and has been carried out 93 times since then with relatively constant question and variable formats.32 However, adapting to the respective political circumstances, questions have been amended, exchanged, or added. In general, the Standard Eurobarometer asks questions regarding the personal life of respondents, their perception of the EU, wishes, values, and concerns. Each survey is carried out in the respective EU countries by independent contractors, making sure to meet the required random sample size of approximately 1000 survey participants.
Second, the European Values Study has been carried out five times, starting in 1981 and being repeated every nine years. While the first surveys were carried out in 14 European countries with a minimum random sample of 1000 respondents, more countries and larger minimum random sample sizes were incorporated in later survey versions.33 Moreover, the survey focusses on basic human values and is especially concerned with "family, work, environment, perceptions of life, politics and society, religion and morality, national identity." 34
Third, the International Social Survey Programme was established in 1984 and has been carried out yearly since then.35 Each year one of the eleven reoccurring topics is covered by the survey, which is carried out by independent organizations in every participating country. Conducted in 1985, 1990, 1996, 2006, and 2016, the "Role of Government" series is the relevant topic series for this research paper. Generally, the individual organizations survey at least 1000 respondents per country and version.36
Finally, despite the described traits of the three longitudinal public opinion surveys, shortcomings still exist regarding the chosen methodology described in chapter 2.3. Because all survey series started after the first EU enlargement round in 1973, the analysis of convergence will not be possible with special regard to this enlargement round. Furthermore, the countries participating in the EVS and the ISSP are not necessarily always reflecting the respective EU member states at the time of the survey versions. For instance, Greece and Luxembourg were already part of the EU before 1981, but are not included in the first and second version of the European Value Study.37 Whereas these countries are the only inconsistent cases for the EVS, the ISSP warrants special attention due to its global focus. Consequently, it is the longitudinal survey with the lowest correspondence with the respective EU member states at the time of the respective survey. Hence, implications deriving from variables referring to this survey have to be regarded as less conclusive than these from the other two surveys. Furthermore, questions in all three surveys were not always incorporated consistently in all versions. Hence, the matching of survey variables to the derived political values is limited to the survey questions that have been repeated consistently at least once.
Derivation of Value Indicators
The EU integration processes were accompanied by several treaties, communications, directives, and other drafted documents. While most of these documents are dealing with specialized technicalities, certain documents are referring directly to the shared political values of the EU. These references allow the derivation of political values, which will be analyzed in the fourth chapter. The Treaty of Rome in 1958, the Copenhagen Criteria published in 1993, the Treaty Establishing a Constitution for Europe drafted in 2004, and the Treaty on European Union after the Lisbon Treaty in 2009 are all referring to common political values of the EU. The following paragraphs highlight these value references and cluster them into the shared political values of the EU.
Treaty of Rome 1958
The Treaty of Rome, and especially the Treaty establishing the European Economic Community, created the economic union between the six founding members of the EU. However, more importantly for the purpose of this research paper, it includes one of the first references to shared political values of the EU. The preamble of the treaty states the preservation of peace and freedom as a primary target of the newly founded community.38 Additionally, the preamble also references capitalistic market conditions, which are specified throughout the treaty.39 In contrast to later documents, like the Copenhagen Criteria, the Treaty of Rome does not list specific political accession criteria but rather defers accession criteria to a later treaty between the six founding countries.40
Copenhagen Criteria 1993
The Copenhagen Criteria codified political accession criteria for potential new EU member state candidates for the first time. After the Treaty of Rome, several enlargement rounds were decided along lengthy intra-EU discussions, that did not consistently refer to codified EU law.41 Defined by the European Council in 1993, the Copenhagen Criteria, therefore, provide a second crucial definition of common EU political values, which include democracy, rule of law, human rights as well as respect for and protection of minorities. Additionally, the political-economic value of a market economy is stated by these accession criteria,42 similar to the mention in the preamble of the Treaty of Rome.
Draft for EU Constitution 2004
After the European Council met in Laeken in December 2001, it adopted the Laeken Declaration, which called for the draft of an EU constitution, including shared values. While the drafted constitution got rejected by referenda in France and the Netherlands in 2005, the enthusiasm of EU officials stemming from the Laeken Declaration43 and successful ratifications in other EU countries signify the importance of the political values referred to in the constitutional draft. For instance, the draft for the EU constitution got ratified in Spain after a public referendum and in Germany, Italy, Austria, and Belgium after a vote in the respective parliament.44 Consequently, the political values referred to in the draft are part of the relevant codified political values of the EU.
First, the draft's preamble refers to the universal values of human rights, freedom, democracy, peace, and the rule of law, while also introducing the political values of solidarity between countries and EU citizens and transparency.45 Furthermore, these values are reaffirmed in various sections of the draft.46 Additionally, Art I-2 refers to the protection of minorities, specified as non-discrimination, tolerance, and justice, similar to the EU documents analyzed in the prior sections. Moreover, free market and economic competition are referred to again in Art. I-3, which adds sustainable development and environmental protection as shared political values. Finally, like the Copenhagen Criteria, Art. I-58 sec. 1 grants access to the EU for European countries respecting the values stated in Art I-2.
Treaty on European Union
As one of the two core EU treaties, the Treaty on European Union has been constantly amended during the EU integration process, leading to the final inclusion of political values after the Lisbon Treaty in 2009. This inclusion followed the failed ratification of the draft for an EU constitution, which resulted in similar political values being referenced in the treaty. Hence, human rights, freedom, democracy, rule of law, sustainable development, and solidarity are mentioned as guiding principles of the EU.47 These core political values were already mentioned in the Treaty of Amsterdam 1997,48 further signifying their importance. Moreover, these values are again emphasized as common political values in Art. 2, which also names the protection of minorities. While Art. 3 refers to the historical political value of peace, it restates the comparably new political value of sustainable development, which has only been previously introduced by the draft for an EU constitution. Like the Copenhagen Criteria and Art. I-58 of the constitutional draft, Art. 49 of the Treaty on European Union states the respect for the values mentioned in Art. 2 as accession criteria for potential new member states.
Synthesis of the Values
The prior sections provided an overview of the political values codified and introduced by EU documents. Accordingly, the political values of the EU can be divided into a group of historical values, one of repeating accession criteria, and one of comparably new political values.
The first group entails peace, freedom, and a market economy, which were mentioned throughout EU documents from the beginning of the EU, signifying their historical significance. The second group entails democracy, rule of law, human rights, and respect for minorities as well as their protection, which are all referred to in the Copenhagen Criteria or the various treaty articles linking to accession criteria. Finally, sustainable development and solidarity are part of the last group, entailing political values repeatedly mentioned in the most recent EU documents. As the political value transparency is only once mentioned as a common political value and otherwise referred to as a norm for the EU institutions, it is not included in the following analysis. The result of the value synthesis of the analyzed EU documents is displayed in Table 3.
Table 2: Codified common political values of EU documents, grouped
| Treaty of Rome | Copenhagen Criteria | Draft EU Constitution | Treaty on EU | Group of Political Values | |
| Peace | X | X | X | Historical Political Values | |
| Freedom | X | X | X | ||
| Market Economy | X | X | X | ||
| Democracy | X | X | X | Accession Criteria | |
| Rule of Law | X | X | X | ||
| Human Rights | X | X | X | ||
| Minority Protection / Respect for Minorities | X | X | X | ||
| Sustainable Development | X | X | New Political Values | ||
| Solidarity | X | X | |||
| Transparency | X |
Operationalization of the Derived Political Values
The Eurobarometer,49 the European Values Study,50 and the International Social Survey Programme51 offer 31 variables, that match one of the identified political values. However, seven of those variables are not applicable to the research questions as they were only surveyed once in one of the longitudinal survey series. Hence, these variables are omitted from the analysis in chapter 4 and only the remaining 24 variables are matched to one of the derived specific political values. Most variables stem from the EVS, which contributes 20 variables, supported by the ISSP and the EB, which contribute 2 variables respectively. However, none of the survey series offered viable variables referencing the political value solidarity, which results in the exclusion of this value for the analysis. The following paragraphs describe the matched variables, the operationalization of these variables, and the derivation of 13 indicators for the remaining eight political values. The summary of the operationalized political values is presented in table 4.
Two variables from the EVS are combined to characterize the political value of peace. Both variables refer to the engagement in peace movements and are part of the 1990, 1999, and 2008 versions of the EVS. Answers to the survey questions were given on a nominal scale, which allows for an interval interpretation of the share of respondents that mentioned an engagement in the peace movement in response to the survey questions. As both variables occur simultaneously in the three survey versions, they are weighted equally to compute the indicator PEACE – TAKING PART.
Second, the political value of freedom only matches one variable from the EVS, available for 1981, 1990, 1999, and 2008. Hence, the resulting indicator FREEDOM is derived from a variable weighing equality against individual freedom on a nominal scale. Therefore, the indicator is represented on an interval scale of the share of respondents choosing freedom over equality.
Similar to the indicator for peace, two variables from the EVS are operationalized for the indicator of the political value market economy. Whereas the first variable classifies respondents’ preferences regarding private and government ownership of business, the second variable does the same regarding the effects of market competition. As both variables use the same interval scale and are both available for 1990, 1999, 2008, and 2017, the variables can be equally weighted and combined into the indicator MARKET ECONOMY, while maintaining the interval scale.
Fourth, two indicators characterize the political value of democracy. The first indicator DEMOCRACYVALUE derives itself by combining two EVS variables, which were surveyed on an ordinal scale. Whereas the first variable asks for attitudes towards having a democratic system, the second one weighs democracy against other forms of government. While both variables are available for 1999 and 2008, only the first one was surveyed for the 2017 version as well. Hence, the indicator is only computed for 1999 and 2008. Due to the ordinal scale of both variables, the indicator is represented on an interval scale of the share of respondents, that valued democracy over other forms of government in their answers. The second indicator DEMOCRACY-PROTEST refers to a variable from the ISSP and is available for 1985, 1990, 1996, 2006, and 2016. Survey participants were asked to assess if public meetings to protest the government should be allowed. The ordinal scale of the variable is transferred into an interval scale for the indicator by computing the share of respondents, that tended towards allowing these public meetings.
Furthermore, the political value of rule of law is operationalized by referring to an ISSP variable available in 1985, 1990, 1996, 2006, and 2016. This nominal variable represents respondents’ sentiments regarding the need to obey the law and potential exceptions to this principle. Hence, the indicator RULE OF LAW uses an interval scale of the share of respondents that chose to "obey the law without exception".
Sixth, the same type of variables used for operationalizing peace are also used to characterize the political value of human rights. Hence, the resulting indicator HUMAN RIGHTS – TAKING PART consists of two EVS variables, which reflect the engagement of respondents in human rights movements on a nominal scale. Therefore, the indicator is derived by computing the share of respondents mentioning an engagement. Equal weights are applied to the two variables. However, whereas both variables are available for 1981, 1990, 1999, and 2008, only one variable was surveyed in 2017 as well. As a result, the convergence of the indicator is only analyzed from 1981 to 2008.
Moreover, the political value of respect for and protection of minorities is characterized by seven variables from the EVS, resulting in two indicators. The first indicator MINORITY – NEIGHBOR refers to six variables, that asked respondents to point out minority groups, which the respondents would not like to have as neighbors. Corresponding with the specific minority groups in the individual countries, the variables are only surveyed if the specific minority exists in the survey country. However, as the indicator is not referring to dissent with specific minority groups but rather with minorities in general, the available variables are equally weighted for each country individually. Additionally, as some minority groups were not a factor in the surveyed countries in every survey period, all variables are only available for 1999, 2008, and 2017. Due to the nominal scale of the variables, the interval indicator is derived by equally weighing the share of respondents that mentioned the individual minority groups for the available variables. The second indicator MINORITY – HOMOSEX refers to a variable reflecting the justifiability of homosexuality according to the respondents. As the variable was already surveyed on an interval scale in 1981, 1990, 1999, 2008, and 2017, the indicator adopts the existing scale.
Finally, the political value of sustainable development is represented by four indicators, the most out of all political values. The first indicator SUSDEV – TAKING PART follows the same logic as the indicators for peace and human rights. Additionally, while both variables are available for 1981, 1990, 1999, and 2008, only one variable was surveyed in 2017. Accordingly, 2017 is not included in the analysis. The second indicator SUSDEV–APPROVAL refers to the approval of ecological and nature protection organizations, exemplified by two variables from the Eurobarometer series. As both variables are reflected on an ordinal scale, they are operationalized by computing the share of respondents that at least somewhat approve of these organizations. Afterward, the two variables are equally weighted to determine the indicator on an interval scale. As both variables are taken from a Eurobarometer trend file aggregating Standard Eurobarometer versions from 1970 to 2002, the indicator can only be derived for 1982, 1984, 1986, 1989, and 1994. The third indicator SUSDEV–GIVE uses a variable from the EVS available in 1990, 1999, 2008, and 2017. The respondents provided their intent to giving part of their income for environmental protection on an ordinal scale, which allows for an indicator, that reflects the share of respondents agreeing to give part of their income. The final indicator SUSDEV-TAX follows a similar logic as the third one, as it refers to an EVS variable asking people about their agreement with an increase in taxes for the prevention of environmental pollution. Hence, this indicator is also reflected on an interval scale of the share of respondents agreeing to this proposal. However, the variable is only available for 1990 and 1999.
Results: Convergence of Political Values across EU Countries
The results of the data analysis are represented in table 4, which shows the coefficients of variation for the different political value indicators for the EU countries which were part of the EU at the time the different longitudinal surveys were conducted. To identify the effect of enlargement rounds, EU countries are categorized into groups reflecting the memberships of the EU after each enlargement round following the taxonomy introduced in table 1. The only enlargement round omitted from this categorization is the first phase of the Mediterranean enlargement in 1981 as it only includes one country, Greece, which was also not part of the first two versions of the EVS. Therefore, each analyzed group includes the countries that joined the EU during the respective enlargement round and all EU members up to this point. As the table cannot functionally display the coefficients of variation for these different subgroups of EU countries, inferences from these data points are described qualitatively in the following paragraphs. Otherwise, these coefficients of variation are available in the data appendix.
Trends of Individual Political Values
The political value of peace has diverged consistently across EU countries from 1990 to 2008. While this divergence is observed for all individual groups of EU countries as well, the EU enlargements since 1981/1986 have amplified this trend. In fact, in 2008, the coefficient of variation increases with every enlargement round since the Mediterranean enlargements and is larger than the one determined for the group of original member countries and for the nine EU members in 1973. However, it is crucial to note, that, for the whole EU, the mean of the indicator ranges from approximately 0,0085 in 1990 to 0,016 in 2008. As such low means indicate engagements in peace movements of around 9 to 16 respondents per 1000 survey participants, relatively small divergences in the composition of the country-specific sample could lead to large changes in the result. Consequently, the coefficients of variation reach values above 1, indicating a standard deviation larger than the mean. Nonetheless, the similar divergences across the different EU country groups still suggest an EU value divergence driven by all member states, but especially by the ones entering the EU during the first phase of the 10+2 enlargement. In fact, in the 2008 version of the EVS, engagements in peace organizations for all countries entering the EU during the 2004 enlargement are significantly lower than the mean.
The political value of freedom has converged consistently across EU countries from 1981 to 2008. While from 1981 to 1990, this convergence is mostly a result of an initial convergence within the original EU members and within the EU countries after the first EU enlargement in 1973, the inclusion of the 1995 and 2004 enlargement rounds continues the convergence trend from 1990 to 2008. The later continuation is taking place despite a simultaneous divergence between the original EU members and a simultaneous divergence between the EU countries of 1986. Furthermore, in 2008, the countries of the 2004 enlargement contributed to a decrease in the coefficient of variation of 15 % in comparison to the EU countries of 1995. Hence, new EU members were generally more in line with the average values of the prior EU member countries.
The value market economy only converges from 2008 to 2017, after the coefficient of variation of the indicator MARKET ECONOMY remained relatively constant from 1990 to 2008. While the original EU countries, the EU countries of 1973, and the EU countries of 1986 are converging to each other from 1990 to 1999, they are all diverging again after 2008. Furthermore, the latter two country groups already diverged from 1999 to 2008, while the original EU members continued their convergence trend in this period. A distinct convergence on the EU level from 1990 to 1999 was prevented by the 1995 enlargement as especially Austrians held more favorable attitudes towards privatization and market competition than other EU countries at the time. The countries of the 2004, 2007, and 2013 enlargement rounds contribute to the observed convergence from 2008 to 2017. In contrast, the countries of the first, the Mediterranean, and the EFTA enlargement rounds contribute to a divergence, which is only offset by the proportionally large enlargement round of 2004. Furthermore, the original EU countries remain the most homogeneous group regarding the political value market economy, followed by the EU countries of 2013, 2007, and 2004.
Attitudes towards democracy diverged almost constantly in the EU as shown by the indicators DEMOCRACY-PROTEST and DEMOCRACY-VALUE. As the indicator DEMOCRACY-PROTEST only relies on a sample size of three countries for the ISSP versions of 1985 and 1990, these results are omitted from the analysis. Furthermore, the group of original EU countries consisted only of Germany in 1990, further rendering these initial survey versions unsuitable for analysis. However, after 1996 both indicators display similar divergence trends with relatively stable coefficients of variation for the EU countries of 1973, 1986, and 1995 till 2006 and 2008 respectively. Additionally, both indicators signify convergence within the group of EU founding members. Despite these trends, the general attitudes towards democracy in the EU diverged significantly from 1996 to 2006, and from 1999 to 2008. This is mostly due to the 2004 enlargement, proven by the 2006 and 2008 survey versions. In 2006, 77,2 % of respondents from countries of this enlargement round thought that public protests against the government should probably or definitely be allowed. In contrast, this attitude was voiced by 90,4 % of respondents in countries, that were part of the EU before.
After 1996, the attitudes of EU citizens towards the rule of law have converged across EU countries. Furthermore, all subgroups of EU countries have become more homogeneous regarding this political value from 1996 to 2006 and from 2006 to 2016. Before 1996, the variable consulted is only available for a sample size of three, with Ireland substituting for Italy in 1990. Hence, it is not possible to derive viable conclusions for this period. In contrast to most other political values and their indicators, the group of original EU members is not consistently the most homogeneous subgroup. However, by 2016, the founding members of the EU represent the most homogeneous group again.
From 1981 to 2008, EU countries have become more heterogeneous in terms of the political value of human rights. However, from 1990 to 1999 a period of convergence occurs, mostly driven by the accession of Austria, Finland, and Sweden in 1995. This convergence is happening despite the group of original EU members and the group of EU countries in 1973 continuing their global divergence trend during this period as well. Finally, from 1999 to 2008, the enlargement rounds of 2004 and 2007 are increasing the divergence, exemplified by the 25,8 % increase in the coefficient of variation in 2008 if the countries of the 2004 enlargement round are included in the calculation. This increase is mostly due to the low engagement in human rights organizations in countries joining the EU after 1995. Whereas, in 2008, the average value of the indicator HUMAN RIGHTS-TAKING PART is 6,4 % for the EU countries up until the 1995 enlargement, this indicator is only 0,5 % for the countries of the 2004 enlargement round. Finally, the original EU country members are again one of the most homogeneous groups, this time joined by the group of EU countries of 1973.
Both indicators for the values of minority protection and respect for these groups diverged from 1981 to 2017. Additionally, both indicators are characterized by a temporary period of convergence before the coefficient of variation reaches a higher than initial level. However, whereas the indicator for the justifiability of homosexuality displays continuous convergence for all country groups up until the inclusion of the EFTA enlargement, the second indicator reflects continuous divergence for almost all subgroups of EU countries after 1990. Hence, the net divergence of the indicator MINORITY-HOMOSEX is a result of the 2004 and 2007 enlargement rounds, leading to a peak coefficient of variation in 2008. This global maximum is a result of the low acceptance of homosexuality in the countries that joined the EU in 2004. Whereas the indicator for these member states averages 3,24 on the 1-10 scale, the average indicator of the other EU countries is 5,88 in 2008. However, while being part of the EU, the countries of the 2004 enlargement round converged towards the attitude of the other EU countries as the indicator of the countries joining the EU in 2004 reached 4,46 in 2017. Furthermore, while the indicator MINORITY-NEIGHBOR is diverging across all subgroups of EU countries since 1990, the accession of new EU members in 2004 is amplifying this trend. In fact, in 29,3 % (2008) and 35,9 % (2017) of cases, citizens in these countries said they would not like to have a specific minority group as neighbors. In contrast, this indicator is only 16,1 % (2008) and 13,3 % (2017) for citizens of the EU countries prior to the 2004 enlargement round.
The indicators SUSDEV-GIVE and SUSDEV-TAX display a net divergence among the EU countries from 1990 to 2017, and from 1999 to 2008 respectively. However, the indicators are showing opposite trends from 1999 to 2008 despite the two indicators both displaying the propensity of citizens to incur a financial loss to protect the environment. This is mostly due to convergence among the original EU members and among the EU members of 1973 for the indicator SUSDEV-GIVE. In contrast, the second indicator for these subgroups is diverging during this period, exemplified by Germany deviating from the mean of the original EU members by about 1 % in 1999 and more than 41 % in 2008. Furthermore, all subgroups of EU countries are diverging in the indicator SUSDEV-TAX from 1999 to 2008. Moreover, the enlargement rounds of 2004 and 2007 are amplifying this general divergence trend. Similar to this effect of these more recent EU members, they are contributing to a divergence in the indicator SUSDEV-GIVE from 2008 to 2017. However, in this case, these EU members are negating the continued convergence among original EU members, EU members of 1973, of 1986, and EU members after the EFTA enlargement round. This continued trend among these subgroups of EU countries began in 1999 after these groups diverged initially from 1990 onwards. Finally, both indicators display a greater homogeneity among the countries with a longer EU membership than the ones that joined the EU during the enlargement rounds in the 21st century.
Furthermore, whereas the indicator SUSDEV-APPROVAL is converging between 1982 and 1994, the indicator SUSDEV-TAKING PART is diverging from 1981 to 2008. However, for the indicator SUSDEV-APPROVAL, the sample size of EU members aside from the founding countries is too small in 1984, 1989, and 1994. Hence, derivations for the whole EU are mostly influenced by the original EU members. Nonetheless, in 1982 and 1986, when more data points for other EU countries are available, these other EU members are amplifying the homogeneity among EU countries. Till 1999, all subgroups of EU countries are displaying divergence in the indicator SUSDEV-TAKING PART. Afterward, the introduction of the EU countries of the 2004 and 2007 enlargement rounds is contributing to the continued divergence among EU countries despite the convergence processes within the original EU members and the EU countries of 1973. This is mostly due to the countries of the 2004 and 2007 enlargement rounds reporting an average indicator about 71 % smaller than the average indicator of the countries with a longer EU membership. Finally, the original EU countries and the EU countries as of 1973 are the most homogeneous groups regarding the indicator SUSDEV-TAKING PART in 2008.
General Convergence Trends
First, the political values of EU citizens across EU member countries diverged in four instances, namely in the political values of peace, democracy, respect for and protection of minorities, and sustainable development. In contrast, convergences occurred in the political values of freedom, market economy, and rule of law. The political value sustainable development presents a special case as three indicators diverged, while only one converged. However, as the converging indicator SUSDEV-APPROVAL is plagued by persistent sample size issues, sustainable development is classified as diverging. Hence, the political values of the same subgroup defined in chapter 3.1 are not following uniform trends. The observed trends can be classified as modest or pronounced. As a measurement for this distinction, the percentual changes in the coefficients of variation from the first available data points to the last available ones are displayed in table 4. A more than 50 % reduction in the coefficient of variation is classified as a very pronounced convergence, whereas a more than 100 % increase in the coefficient of variation is classified as a very pronounced divergence. Furthermore, a convergence between -33 % and -50 % is classified as pronounced, while a pronounced divergence occurs between 50 % and 100 %. The remaining trends are defined as modest. These placements allow a further comparison between divergences and convergences as they equate to proportional changes. As a result, the indicators describing the engagement of citizens in human rights, peace, or sustainable development organizations display a very pronounced divergence. This outcome can partially be attributed to the low variable values but is also a testament to the already large differences between EU countries at the beginning of the surveys. Furthermore, only the political value of rule of law converged profoundly during the observed period. This observation further strengthens the conclusion of more profound divergences across EU countries as the political value of sustainable development also profoundly diverged in two indicators. In sum, three indicators show very pronounced divergence, two indicators display profound divergence, and four indicators are characterized by modest divergence. In contrast, only one indicator each shows very pronounced convergence and pronounced convergence, while two indicators display modest convergence.
Second, the founding members of the EU are the most homogeneous group in terms of the analyzed values as they are one of the two most homogeneous subgroups of EU countries at the time of the last available data for all political values analyzed. Similar conclusions for other subgroups of EU countries are not possible as no clear pattern can be identified.
Third, the effect of the enlargement rounds on the value convergence depends on the political value analyzed. Whereas the enlargement rounds of the 20th century mostly increase heterogeneity initially but follow the general convergence trend set by the original EU countries, the enlargement rounds of the 21st century have more ambiguous effects. Hence, it is possible to update the conclusion of Gerhards,52 who postulates that each enlargement round increases the value differences among EU countries. Whereas the enlargement rounds of the 21st century amplified or caused a divergence in the political values of peace, democracy, human rights, respect for and protection of minorities, and sustainable development, these enlargement rounds contributed to an increased homogeneity across EU countries in the political values of freedom, rule of law, and market economy.
Fourth, after countries become EU members, they sometimes contribute to a temporary convergence process among EU countries, mostly by adopting the political values of the more established EU members. For instance, after they joined the EU, the countries of the 2004 enlargement round converged towards the predominant attitude regarding the political values market economy, democracy, and rule of law.
| Political Value | Indicator | Range | Number of Variables | Source of Variables |
| Peace | PEACE-TAKING PART | ]0;1[ | 2 | EVS |
| Freedom | FREEDOM | ]0;1[ | 1 | EVS |
| Market Economy | MARKET ECONOMY | ]0;10[ | 2 | EVS |
| Democracy | DEMOCRACY-VALUE | ]0;1[ | 2 | EVS |
| DEMOCRACY-PROTEST | ]0;1[ | 1 | ISSP | |
| Rule of Law | RULE OF LAW | ]0;1[ | 1 | ISSP |
| Human Rights | HUMAN RIGHTS-TAKING PART | ]0;1[ | 2 | EVS |
| Minority Protection / Respect for Minorities | MINORITY-NEIGHBOR | ]0;1[ | 6 | EVS |
| MINORITY-HOMOSEX | ]0;10[ | 1 | EVS | |
| Sustainable Development | SUSDEV-APPROVAL | ]0;1[ | 2 | EB |
| SUSDEV-TAKING PART | ]0;1[ | 2 | EVS | |
| SUSDEV-GIVE | ]0;1[ | 1 | EVS | |
| SUSDEV-TAX | ]0;1[ | 1 | EVS |
Conclusion
The examination of key EU documents leads to the definition of nine political values of the EU. Of these nine values, eight are operationalizable referring to the EVS, the EB, and the ISSP. The derivation of the σ-convergence among the EU countries across time results in four main conclusions.
First, no overarching convergence trend is observable as four out of eight political values diverge across EU countries over time. However, divergences are more profound than convergences, mostly driven by the indicators measuring citizen engagement in different organizations. Second, the original members of the EU consistently constitute the most homogeneous subgroup analyzed. Third, the EU enlargement rounds do not necessarily increase value differences between EU countries. Fourth, new EU members adopt the predominant attitude of the other EU countries in some cases. These conclusions extend the existing research on convergence among EU countries by a study incorporating more countries, more political values, and a more recent dataset. Furthermore, the resulting conclusions contradict existing assumptions of increasing divergence due to EU enlargement rounds and paint a more varied picture.
However, four caveats restrict the significance of the derived conclusions. First, the EVS, the EB, and the ISSP were not always carried out in all EU countries at the time of the respective survey versions. Second, due to the decentral organization of these surveys, minor differences between the implementation in different countries occurred, rendering the report of some responses unusable for the purposes of this research paper. Consequently, the conclusions had to be derived from a patchwork of different periods and country compositions. Third, insufficient consistency among variables in the three surveys resulted in some indicators consisting of only one variable, which might have introduced an overreliance on these specific questions. Finally, the scope of the research paper did not allow for quantitative significance testing of the observed trends.
Hence, possible future research based on the conclusions derived in this research paper could be focused on verifying the results by introducing random coefficients to test the significance of the observed trends. Furthermore, this research paper purely derives the convergence trend of eight political values and does not include research on the qualitative reasons for these convergence trends. Hence, future qualitative research could focus on specific political values and try to explain the observed differences in convergences trends. Moreover, further research could increase the depth of analysis and identify convergence catalysts among the different subgroups of EU countries.
Footnotes
1 As an example see: Dullien, S. and Torreblanca, J.I. (2012). "What Is Political Union? Policy Brief 70", Council on Foreign Relations, London: European Council on Foreign Relations, Available here, (Accessed on: 3 January 2021); Grant, C. (1999). "Europe's Emerging Political Union", Centre for European Reform (blog), Available here, (Accessed on : 3 January 2021); Laffan, B. (2004). "The European Union and Its Institutions as 'Identity Builders'", In Richard K. Herrmann, Thomas Risse, and Marilynn B. Brewer (eds.), Transnational Identities: Becoming European in the EU, , 1st ed., Lanham: Rowman & Littlefield Publishers, pp. 75–96.
2 Schuman, R. (1950). "Declaration of 9 May", Declaration, Salon de l'Horloge at the Quai d'Orsay in Paris, May 9.
3 As an example see: European Union (2016). "The EU in Brief." Text. European Union, 16 June, Available here , (Accessed on: 3 January 2021).
4 As an example see: The Economist (2020). "Can the European Union Learn to Love a Common Culture?", The Economist, 17 October, Available here, (Accessed on : 3 January 2021).
5 Akaliyski, P. (2017). "Sources of Societal Value Similarities across Europe: Evidence from Dyadic Models", Comparative Sociology, 16 (4), pp. 447–70.
6 As an example see: Guiso, L., Herrera, H. and Morelli, M. (2016). "Cultural Differences and Institutional Integration", Journal of International Economics, 38th Annual NBER International Seminar on Macroeconomics, 99 (March), pp. S97–113.
7 Burchard, H.von der (2020). "Hungary and Poland Escalate Budget Fight over Rule of Law", POLITICO, 26 November, link (Accessed on: 3 January 2021).
8 Fuchs, D. and Klingemann, H.D. (2002). "Eastward Enlargement of the European Union and the Identity of Europe", West European Politics, 25 (2), pp. 19–54; Schimmelfennig, F. (2001). "The Community Trap: Liberal Norms, Rhetorical Action, and the Eastern Enlargement of the European Union", International Organization, 55 (1), pp. 47–80.
9 Inglehart, R. and Baker, W.E. (2000). "Modernization, Cultural Change, and the Persistence of Traditional Values", American Sociological Review, 65 (1), pp. 19–51.
10 Haller, M., Hadler, M., Balog, A. and Kellerman, P. (2004). "Ist Der Nationalstaat Überholt? Überlegungen Und Fakten Über Die Sinnvollste Einheit Bzw. Analyseebene in Der International Vergleichenden Sozialforschung. Kritik. Replik", Ist Der Nationalstaat Überholt? Überlegungen Und Fakten Über Die Sinnvollste Einheit Bzw. Analyseebene in Der International Vergleichenden Sozialforschung. Kritik. Replik, 23 (3–4), pp. 141–69.
11 Akaliyski, 2017.
12 Houwelingen, P. van, Iedema, J. and Dekker, P. (2019). "Convergence on Political Values? A Multi-Level Analysis of Developments in 15 EU Countries 2002-2016", Journal of European Integration,41 (5), pp. 587–604.
13 Akaliyski, P. (2019). "United in Diversity? The Convergence of Cultural Values among EU Member States and Candidates." European Journal of Political Research 58 (2), pp. 388–411.
14 Gerhards, J. (2010). "Non-Discrimination towards Homosexuality: The European Union's Policy and Citizens' Attitudes towards Homosexuality in 27 European Countries", International Sociology, January.
15 Gerhards, J. (2008). "Free to Move?: The Acceptance of Free Movement of Labour and Non Discrimination Among Citizens of Europe", European Societies, 10 (1), pp.121–40.
16 Oshri, O., Sheafer, T. and Shenhav, S.R. (2016). "A Community of Values: Democratic Identity Formation in the European Union", European Union Politics, 17 (1), pp. 114–37.
17 Gerhards, J. (2007). Cultural Overstretch?: Differences between Old and New Member States of the EU and Turkey, 1st ed, New York: Routledge.
18 Halman, L. (2007). "Political Values", In Russel J. Dalton and Hans-Dieter Klingemann (eds.), The Oxford Handbook of Political Behavior, New York: Oxford University Press, p. 309.
19 Reich, B. and Adcock, A. (1976). Values, Attitudes and Behaviour Change, 1st ed., Vol. Essential Psychology, 3 vols., North Yorkshire: Methuen; Rokeach, M. (1968). Beliefs, Attitudes, and Values: A Theory of Organization and Change, 1st ed. San Francisco: Jossey-Bass, p. 124.
20 Halman 2007, p. 309
21 Almond, G.A. and Verba, S. (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. 1st ed. Princeton, New jersey: Princeton University Press.
22 Halman 2007, p. 310.
23 Nugent, N. (2017). The Government and Politics of the European Union, 8th ed., London: Macmillan Education.
24 Ibid., p. 57.
25 Ibid.
26 Monfort, P. (2008). "Convergence of EU Regions - Measures and Evolution", Working Paper 01/2008, EU Regional Policy Working Papers, Luxembourg: EU Publications Office, p. 4-5, Available here, (Accessed on : 3 January 2021).
27 Hancke, B. and Axisa, A. (2020). "Measuring Social Convergence across the EU", Luxembourg: EU Publications Office, p. 9, Available here, (Accessed on : 3 January 2021).
28 Ibid.
29 Monfort 2008, p. 5.
30 Hancke and Axisa 2020, p. 9.
31 JRC (2020). "Coefficient of Variation - European Commission", Joint Research Centre of the European Commission, 3 December, Available here , (Accessed on : 3 January 2021).
32 European Commission (2020). "PublicOpinion", Public Opinion, Available here, (Accessed on: 3 January 2021).
33 European Values Study (2018a). "EVS Methodology", European Values Study, Available here, (Accessed on: 3 January 2021).
34 European Values Study (2020a). "European Values Study 2017: Integrated Dataset (EVS 2017)", GESIS Data Archive, Cologne: GESIS - Leibniz Institute for the Social Sciences.
35 International Social Survey Programme (2020). "International Social Survey Programme", Available here, (Accessed on : 3 January 2021).
36 European Values Study 2018, p. vii-x.
37 European Values Study (2015). "EVS 1981 - 2008 Variable Report Longitudinal Data Files", Variable Report 29, GESIS-Variable Reports, Cologne: GESIS - Leibniz Institute for the Social Sciences, p. 32.
38 See preamble: European Economic Community (1958). Treaty Establishing the European Economic Community, Document 11957E/TXT, Available here, (Accessed on: 3 January 2021).
39 Ibid., pt. III, title I, chapter I
40 Ibid., Art. 237.
41 See: Janse, R. (2018). "The Evolution of the Political Criteria for Accession to the European Community, 1957–1973", European Law Journal, 24 (1), pp. 57–76.
42 EU Publications Office. n. d. "Accesion Criteria (Copenhagen Criteria)", Glossary, EUR-LEX, Available here, (Accessed on: 3 January 2021).
43 Hurrelmann, A. (2007). "European Democracy, the 'Permissive Consensus' and the Collapse of the EU Constitution", European Law Journal, 13 (3), p. 343.
44 BBC News (2007). "EU Constitution: Where Member States Stand." BBC News, March 25, Available here, (Accessed on: 3 January 2021).
45 See Preamble: European Union (2004). "Draft Treaty Establishing a Constitution for Europe", C 310, Official Journal of the European Union, Luxembourg: European Union.
46 Ibid., Art. I-2 & Art. I-3.
47 See Preamble: European Union (2012). Consolidated Version of the Treaty on European Union, Document 12012M/TXT, Available here, (Accessed on: 3 January 2021).
48 See pt. 1, sec. 8: European Union (1997). Treaty of Amsterdam Amending the Treaty on European Union, Document 11997D/TXT, Available here, (Accessed on: 3 January 2021).
49 Schmitt, H., Scholz, E., Leim, I. And Moschner, M. (2008). "The Mannheim Eurobarometer Trend File 1970-2002 (Ed. 2.00)", Edited by European Commission, GESIS Data Archive, Cologne: GESIS - Leibniz Institute for the Social Sciences.
50 European Values Study (2020a). "European Values Study 2017: Integrated Dataset (EVS 2017)", GESIS Data Archive, Cologne: GESIS - Leibniz Institute for the Social Sciences.
51 ISSP Research Group (2008). "International Social Survey Programme: Role of Government I-IV - ISSP 1985-1990-1996-2006", GESIS Data Archive, Cologne: GESIS - Leibniz Institute for the Social Sciences.
52 Gerhards, J. (2007). Cultural Overstretch?: Differences between Old and New Member States of the EU and Turkey, 1st ed, New York: Routledge.