This is an open access article distributed under the terms of the Creative Commons
Attribution License (In times of hot ideological disputes, such as those that have recently taken place in Poland, important questions about democracy often arise. It would be difficult to find, in such debates, a political philosopher or social scientist who would deny that freedom is the key attribute of democracy. However, in both theoretical considerations on democracy and everyday beliefs about what is, and what is not democracy, the concept of freedom is understood in various ways.
One of the most important issues in contemporary normative theories of liberal democracy is the opposing two concepts of political freedom: 'unconditional' and 'conditional'. The first focuses on maximizing individual autonomy, and stipulates that one should act according to the principle of ‘laissez-faire et laissez passer, le monde va de lui meme’ (let us act and go, the world goes by itself) (Szacki, 1994, p. 40). The second concept refers to the idea of individual autonomy that must be negotiated as a compromise with other socio-political values (e.g., Wnuk-Lipiński, 2005).
The above distinction has a longstanding tradition within the broadly understood doctrine of liberal democracy. For example, in the 1970s, the key differences between the concepts of 'unconditional' and 'conditional' freedom can be seen as the basis for a fundamental dispute between the libertarian and egalitarian versions of liberalism, represented by R. Nozick (1999) and J. Rawls (1971), respectively. However, little is known as to whether these two ideas of political freedom correspond with a sense of freedom subjectively perceived in human minds.
The main purpose of this paper is to find out whether individual preferences of unconditional or conditional freedom are related to differences in understanding political community, and to differences in attitudes towards liberal democracy. To answer this question we will first have to show that both hypothetical forms of freedom are reflected in psychologically different experiences of freedom.
When Do People Feel Free? Unconditional and Conditional Freedom
Erich Fromm (1970/1941) and Isaiah Berlin (1969), well-known and respected political philosophers, initiated the tradition of distinguishing between unconditional and conditional freedom. The first one (called ‘freedom from’ and ‘negative freedom’, respectively) means maximum autonomy of the individual, being free from external orders and restrictions. Full freedom assumes minimal coercion from the authorities and no political interference in one’s private life. In this approach freedom becomes the most important value to be pursued in political action. The second one (called ‘freedom to’ and ‘positive freedom’, respectively) refers to pro-development activity, which gives people the opportunity to accomplish their own potential. This approach to freedom can be seen as ‘conditional’ in two ways. On the one hand, it emphasizes the importance of other values that must be taken into account during political actions aimed at maximizing freedom; on the other hand, it assumes that a certain combination of other values (like: respect, equality, fairness, solidarity, etc.) may be necessary to achieve freedom, i.e., to successfully strive for the accomplishment of individual possibilities and life goals.
These fundamental differences in understanding and experiencing freedom have some empirical support in the results of the studies conducted in different theoretical contexts. For example, in the 1960s and 1970s, psychologists studied individual differences in internal vs. external locus of control. Rotter (1966, 1990) showed that as a result of repeated life experiences, people form generalized beliefs about causal connections linking their behavior and its consequences. As a result, they may feel internally free (that is, capable of achieving one’s own goals and needs), or externally controlled. Rotter did not use the term ‘freedom’, but it can be assumed that people with a strong inner sense of control feel more free and have an intrinsic sense of freedom. Additionally, later studies, conducted within the cognitive theories of attribution by Kruglanski and Cohen (1973, 1974), show that people describe themselves and other people as free when they see that they can do what they want, and they can follow their values, goals and needs. Freedom so felt has an ‘exogenous character’ (Kruglanski, 1975).
In the first decade of the 21st century, Alford (2003) conducted a number of studies on experiencing and understanding freedom. His main goal was to find out to what extent Americans’ understanding of freedom is similar to the concepts of positive and negative freedom proposed by Berlin (1969). It turned out that the majority of people surveyed by Alford did not directly talk about freedom in either a ‘negative’ or ‘positive’ sense. In answers to the question of what freedom is, in the descriptions of the situations of experiencing freedom or lack of it, respondents spoke about money and power. According to Alford, his respondents, especially young Americans, devaluated freedom as an independent value and appreciated it less than money and power. They most often thought about freedom in terms of ‘negative freedom’ - a total lack of external restrictions, which seems impossible, and, as such, not worth their effort. Either freedom is full and unlimited, or not at all.
Studies conducted by Skarżyńska and Radkiewicz on a nationwide sample of adult Poles showed that one can distinguish between ‘reflective’ and 'absolute' freedom (Skarżyńska, 2018). The former refers to living in accordance with personal values and one’s worldview (it is closer to positive freedom), while the latter means an unrestrained striving for pursuing one's own goals and desires (it is closer to negative freedom). The authors noticed that, paradoxically, if someone more strongly experiences and understands freedom as any unrestrained action, the more he/she accepts such elements of authoritarian power as the use of hate speech or force against protesting citizens.
The Current Study
Both the concept of Berlin's positive and negative freedom, and Fromm's concept of 'freedom to' and 'freedom from' were originally used in the field of political philosophy to describe certain types of relationships linking an individual and socio-political environment. They are not psychological constructs that refer to empirically accessible ways of understanding and experiencing freedom by humans. Based on the previous studies, we suggest that the psychological sense of the philosophical distinction regarding unconditional and conditional freedom should be sought in their extrinsic or intrinsic location, respectively. Extrinsic experience of freedom is based on the belief that being free is possible only when one has opportunities to fully achieve one’s own goals and desires, and the factors limiting these opportunities are aversive and should be eliminated or ignored. It is more a sense of freedom resulting from the lack of external restrictions/barriers than from the pursuit of personal goals or values. An expression of extrinsic freedom is identical to unlimited expression of what, at the time, someone considers as ‘being oneself’.
An intrinsic sense of freedom is based on the belief that being free means compatibility between one's own actions and preferred values, life goals or worldview. The expression of intrinsic freedom consists of accomplishing what someone considers good, right, true, etc. However, this harmony between self and actions is only possible as long as it does not violate the well-being of someone else. Personal freedom cannot be accomplished at the expense of the rights and freedoms of other people. In contrast to an extrinsic sense of freedom, in the case of an intrinsic sense of freedom, limiting one's actions by the rights and freedoms of other people is not perceived as detrimental.
Personal Values and Moral Foundations in Intrinsic and Extrinsic Experiences of Freedom
Assuming that at the empirical level both forms of freedom reveal truly identifiable, relatively independent variables, the question arises about their psychological meaning, i.e., to what extent they reflect intra-individual differences between people. Since at the very roots, in the field of political philosophy, the concept of unconditional and conditional freedom is strongly related to the preferences of certain values and moral choices, we believe that the area of axiological and moral characteristics should be the first to be taken into account.
For this purpose, we have referred to the theory of basic human values, in which Shalom Schwartz and colleagues defined values as beliefs determining personal goals that go beyond specific situations and provide essential rules of conduct (e.g., Schwartz, 2006; Schwartz et al., 2012). Schwartz’s model comprises ten types of values situated at the individual level that can be ordered on two higher-level dimensions: Self-enhancement vs. Self-transcendence and Openness to change vs. Conservation. The first dimension opposes the set of values related to individual need for power, control and success (power and achievement), to such values that aim at human solidarity, taking care of others, and harmonious coexistence between man and nature (benevolence and universalism). In the second higher-level dimension, openness to change means the preference for personal autonomy and expression of self (self-direction, stimulation), and its opposite are the values that strengthen community and give a sense of belonging and security (conformity, tradition, security).
First of all, as the value of self-direction promotes intellectual autonomy and independence of action, it seems obvious to expect that self-direction should be positively related to both intrinsic and extrinsic freedom (H1). Furthermore, we believe that intrinsic and extrinsic freedom substantially differ in the level of concern for the well-being, rights and freedoms of other people, as well as in the level of striving for unlimited fulfillment of one's goals and desires. Therefore, we expected that one’s inclination toward an intrinsic vs. extrinsic sense of freedom should be reflected in different patterns of relationships with self-enhancement and self-transcendence values. The more people are guided by values of self-enhancement, the more they should prefer extrinsic freedom and the less they should prefer intrinsic freedom (H2). Conversely, the more people are guided by values of self-transcendence, the more they should prefer intrinsic freedom and the less they should prefer extrinsic freedom (H3).
The second source of the hypotheses was the moral foundations theory (MFT) developed by Jonathan Haidt and colleagues (Haidt 2007, 2014). MFT explains the origins of human moral reasoning on the basis of innate, modular foundations and emphasizes the central role of automatic affective evaluations (moral intuitions) in making moral judgments. According to Haidt (2014) there are six modular foundations underlying moral reasoning: care/harm, fairness/reciprocity, ingroup/loyalty, authority/respect, purity/sanctity and liberty. In the context of intrinsic and extrinsic freedom, three of them seem particularly important: care/harm, which means that it is wrong to hurt others, and it is right to prevent or alleviate harm; fairness/reciprocity, which leads to disapproval of those who violate principles of justice and reciprocity; and liberty, which gives moral priority to human freedom and disapproves of various forms of violation of individual autonomy by other people and the state.
The three moral foundations mentioned above are directly related to human well-being - they express respect for goodness, freedom and the rights of a human being, helping others, and loyalty to individuals. However, although they all have an ‘individualizing’ character, and aim to protect individuals, we think they should be related to an intrinsic and extrinsic sense of freedom in a somewhat different way. Someone who is guided by care/harm and fairness/ reciprocity seeks self-realization in a way that does not violate the well-being of other people, and works to ensure that no one is deprived of his/her rights. Therefore, we argue that moral intuitions of care/harm and fairness/reciprocity should favor experiencing intrinsic freedom (H4). In turn, as the foundation that gives freedom the highest moral value, the moral intuition of liberty should favor experiencing extrinsic freedom (H5).
Intrinsic and Extrinsic Freedom as Predictors of the Preference for Political Community
Referring to political science, one can argue that a democratic society can function on the basis of two quite different ideas of the political community: liberal and communitarian (e.g., Szacki, 1997). According to liberal beliefs, a citizen is free, morally judicious, and autonomic. Liberals stress that community is made up of free individuals who have different beliefs, religions, and worldviews. They have to co-exist within the law and be respectful of each other. Citizens have a set of liberties that do not need to be justified. The state has to ensure the freedom of an individual by means of rules that prevent lawlessness, chaos, and tyranny. Liberal civil society is a moral community, which exists to preserve and enlarge the freedom of its members, for their security and peaceful coexistence, and for the protection of their property.
On the other hand, according to communitarian beliefs, a citizen is deeply rooted in communities and formed by socialization. Without norms, values, language, history, and culture adopted from the community, an individual would only be a biological being and nothing more. A community is primal and forms individuals’ identity. Communitarians believe that freedom of the political community is the condition of individual freedom. And freedom of the political community is based on the authority of law, customs, and/or religion.
As one can see, both orientations are based on a different approach to individual freedom. In a communitarian orientation, political community is aimed towards a common good, which has priority over individual rights and freedoms. Individuals are expected to be ready to limit their own freedom for the overarching goals of the community. Considering this key characteristic, we suppose that preference for communitarian political community can only be favored by an intrinsic sense of freedom (H6).
In a liberal orientation, an individual has a superior position to the community, and the primary task of the liberal political community is to protect and maximize individuals’ rights and freedoms. This naturally aligns the liberal orientation with an extrinsic sense of freedom. On the other hand, pointing to intrinsic or extrinsic freedom as distinctive for the liberal orientation seems questionable, since - regardless of the differences - they are both anchored in striving for personal autonomy and independence. Therefore, we suppose that preference for liberal political community can be favored by both an intrinsic and extrinsic sense of freedom (H7).
Intrinsic and Extrinsic Freedom as Predictors of the Support for Liberal Democracy
Our previous research shows that support for the principles of liberal democracy and aversion to an authoritarian government are higher the more people value freedom that is understood as self-fulfillment/self-realization and limited by the well-being of other people (Skarżyńska, 2018). Therefore, we suppose that only an intrinsic sense of freedom should be a direct positive predictor of support for principles of liberal democracy (H8). This is in line with the normative understanding of liberal democracy, which aims to protect human rights and provide fair living conditions not only for the majority, but also for various minorities. It gives freedom of choice to all citizens, but it also tries to prevent disrespecting the law and violating the rights of other people (Król, 2017; Szacki, 1994).
On the other hand, we believe that the positive effect of an extrinsic sense of freedom on support for liberal democracy exists indirectly, i.e., it is mediated through the preference for liberal political community, which is expected to directly promote the principles of liberal democracy (H9). This expectation results from two premises. First, as we argued earlier, an extrinsic sense of freedom seems to be related to a liberal orientation and not to a communitarian one. Second, there is no doubt that liberal democracy, as the political system, builds inherently upon the core values of the idea of liberal political community.
Method
Participants and Procedure
The sample consisted of 800 adult Poles: 51.8% females and 48.2% males. 17.8% of respondents were aged 18 to 30, 30.1% were aged 31-44, 22.1% were aged 45-60, and 30% were above 60. Primary and lower education had been completed by 7.4% of respondents, junior high school - 2%, vocational - 37.4%, secondary and post-secondary – 33.7%, and 19.5% of the respondents had completed higher education. Place of residence: 37.8% of respondents lived in the countryside, 6% in small towns up to 20 thousand inhabitants, 27.6% in towns above 20 to 100 thousand, 18.8% in towns above 100 to 500 thousand, and 9.8% in cities above 500 thousand inhabitants.
A survey study was conducted. Respondents were selected for the random-quota sample based on a two-stage procedure: 1) random sampling of urban and rural areas, and then 2) random selection of respondents (quotas defined by the combination of three criteria: sex, age, and education). Interviews were conducted in respondents’ homes by trained interviewers using a computerized questionnaire format.
Measures
Intrinsic and Extrinsic Sense of Freedom
Two forms of experiencing freedom were measured using twelve items – six for each component. Respondents received the following instruction: ‘There is now a lot of talk and writing about freedom. And when do you feel really free? We present a dozen statements about freedom. Please read each of them carefully and on the attached scales, indicate to what extent you agree or disagree with each of these statements’ (responses on a scale from 1-disagree strongly to 5-agree strongly). Examples of statements for intrinsic freedom: ‘I feel really free when I do what is consistent with my values‘, ‘I feel really free when I can do what I want, provided that it does not violate the rights of other people‘; examples of statements for extrinsic freedom: ‘I feel really free when the state or government does not tell me how to live‘, ‘I feel really free when I do not have to wonder if I can do it or not, whether it should be done or not‘. The exact wording of the scale is shown in the Results section.
Personal Values
To measure axiological preferences we used the Portrait Values Questionnaire developed by Schwartz (2006). The applied measurement consisted of a total of 40 items including ten types of values: self-direction, stimulation, hedonism, conformity, tradition, group security, achievement, power, benevolence and universalism. In accordance with the PVQ format, the items (descriptions) referred to the views and behavior of an unknown person. The respondent’s task was to assess – on a scale from 1 (completely unlike me) to 6 (completely like me) – to what extent his/her feelings and behavior were similar to those of the persons in the presented descriptions. For descriptive statistics and intercorrelations amongst personal values, see Appendix A.
Moral Foundations
We used the Moral Foundations Questionnaire (MFQ) (Graham et al., 2011) developed to measure five moral intuitions: care/harm, fairness/reciprocity, ingroup/loyalty, authority/respect and sanctity/ purity. In addition, we used the measurement of a sixth moral intuition described by authors in later publications as liberty (Haidt, 2014; Iyer, Koleva, Graham, Ditto, & Haidt, 2012). The measurement of each moral fundament included 6 items: three on the subscale of moral relevance (response options ranged from 1 = not at all relevant to 6 = extremely relevant) and three on the subscale of moral judgments (response options ranged from 1 = strongly disagree to 6 = strongly agree). Examples of moral judgments: care/harm – ‘Compassion for those who are suffering is the most crucial virtue‘ (α = .74); fairness/reciprocity – ‘Justice is the most important requirement for a society’ (α = .74); ingroup/loyalty – ‘It is more important to be a team player than to express oneself‘ (α = .67); authority/respect: ‘Respect for authority is something all children need to learn‘ (α = .67); sanctity/purity: ‘People should not do things that are disgusting, even if no one is harmed‘ (α = .72), and liberty: ‘Everyone should be free to do as they choose, as long as they don’t infringe upon the equal freedom of others’ (α = .66). For descriptive statistics and intercorrelations amongst moral intuitions see Appendix B.
Political Community: Liberal and Communitarian Orientation
This instrument was developed by Radkiewicz to measure preference for liberal or communitarian political community (Radkiewicz & Jarmakowski-Kostrzanowski, 2018). A 32-item scale is based on the assumption that within the broadly understood democratic order one can distinguish two internally coherent beliefs about citizenship, civicness and civil society – liberal and communitarian orientation. They reflect clearly separate views on the right and preferred ways of organizing social life meant as a political community. In the course of empirical verification, among the wider group of criterial dimensions, the following four categories of beliefs were distinguished, which apparently differentiate both orientations: civic liberty, citizen's identity, relations between an individual and society, and view of the state. Each category contained 4 items. The exact wording of the scale is shown in Appendix C.
Internal reliability was α = .87 for liberal orientation (16 items) and α = .91 for communitarian orientation (16 items). At the level of measurement both orientations proved to be significantly positively correlated: r = .53, p < .001. The confirmatory factor analysis showed the best goodness of fit (amongst four competing models) for the two-factorial model with four categories of beliefs (χ2 = 1150.8, df = 436; CFI = .920; SRMR = .058; RMSEA = .045). The four categories (as described above) were defined as the latent variables separately for both orientations. Among a number of criterion variables used to assess the construct validity, the most distinct and expected relationships were obtained with Cameron's (2004) national identification scale: r = -.07, p < .001 for liberal orientation, and r = .39, p < .001 for communitarian orientation. This correlational pattern was even more expressive at the level of partial correlations, when the collinearity effects were controlled: r = -.37, p < .001 and r = .51, p < .001, respectively.
Support for Democratic Principles
This 9-item scale developed by Kaase (1971) taps attitudes on a number of consensual and contested principles common to democratic systems, including basic democratic values and support for political rights versus social order. The author identified the five following principles: 1) the right of individual political participation; 2) popular control of the government through regular elections in which a change of leadership is possible; 3) understanding the legitimacy of political conflict; 4) opposition to violence as a legitimate means of resolving political conflict; and 5) consensus on fundamental democratic attitudes. Examples: ‘Every citizen has the right to take his convictions to the street if necessary’, ‘In principle, every democratic party should have a chance to govern’, ‘Everyone should have the right to express his opinion even if he/she differs from the majority’, ‘It is not conceivable to have a viable democracy without a political opposition’ (responses on a scale from 1-disagree strongly to 6-agree strongly). Internal reliability of the scale amounted to α = .80.
Results
Factor Validity of the Measurement of Intrinsic and Extrinsic Sense of Freedom
In order to estimate factor validity of the model containing intrinsic and extrinsic sense of freedom we performed confirmatory factor analyses using IBM SPSS Amos 21.0. We contrasted two alternative models: 1) one-factorial, with 54 degrees of freedom, containing one general latent variable underlying the covariance of all 12 items; and 2) two-factorial, with 53 degrees of freedom, containing the target structure with two factors, as depicted in Figure 1.
Figure 1
Confirmatory factor analysis: Model representing intrinsic and extrinsic sense of freedom as a target factor structure.
The one-dimensional model did not show a satisfactory goodness of fit: (χ2 = 537.6; GFI = .873; SRMR = .075; RMSEA = .106), while the goodness fit statistics for the two-factorial model proved to be much better and satisfactory (χ2 = 254.2; CFI = .925; SRMR = .051; RMSEA = .065, 95% CI [0.060, 0.069]. The difference in goodness of fit between the two models, measured by the difference in χ2 statistics, was statistically significant (χ2 = 283.4; df = 1; p < .001).
The results of the confirmatory factor analysis were the basis for creating subscales of intrinsic (M = 4.19, SD = 0.54) and extrinsic (M = 3.91, SD = 0.62) sense of freedom, which proved to be significantly positively correlated: r = 0.58, p < .001. Internal reliability for both subscales amounted to α = .81 and .75, respectively.
Personal Values and Moral Intuitions as Predictors of Intrinsic and Extrinsic Sense of Freedom
In the next step, we focused on looking for evidence that personal values and moral intuitions are important factors underlying intrinsic or extrinsic sense of freedom. First, we hypothesized that the values of self-transcendence should favor experiencing intrinsic freedom, whereas the values of self-enhancement should favor experiencing extrinsic freedom. Second, we expected that moral intuitions of care/harm and fairness/reciprocity should positively predict intrinsic freedom, whereas the moral intuition of liberty should positively predict extrinsic freedom.
Personal Values
Table 1 shows two regression analyses with intrinsic and extrinsic freedom as dependent variables regressed on ten personal values, and including covariates (extrinsic and intrinsic freedom, respectively). In the case of intrinsic freedom as the dependent variable, we found five significant effects: power (β = -.31, p < .001), self-direction (β = .16, p < .001), group security (β = .18, p < .001), benevolence (β = .13, p = .002), and achievement (β = .10, p = .015). These five predictors contributed 24% variance to the total R2 over and above the effect of covariate (extrinsic freedom). For the extrinsic freedom as the dependent variable, we found three significant effects: power (β = .32, p < .001), self-direction (β = .15, p < .001), and conformity (β = .11, p = .008). These three predictors contributed 14% variance to the total R2 over and above the effect of covariate (intrinsic freedom).
Table 1
Effects of Personal Values on Intrinsic and Extrinsic Sense of Freedom – Regression Analysis
| Predictor | Intrinsic freedom
|
Extrinsic freedom
|
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | B | SE | 95% CI | η2 | β | B | SE | 95% CI | η2 | |
| Constant | 1.40** | 0.13 | [1.13, 1.63] | 0.40* | 0.16 | [0.08, 0.70] | ||||
| Intrinsic freedom | .54 | 0.59** | 0.04 | [0.50, 0.63] | ||||||
| Extrinsic freedom | .45 | 0.40** | 0.03 | [0.35, 0.45] | ||||||
| Self-direction | .16 | 0.14** | 0.03 | [0.07, 0.20] | .02 | .15 | 0.13** | 0.04 | [0.05, 0.21] | .01 |
| Hedonism | .01 | 0.02 | 0.00 | [-0.03, 0.03] | .00 | .05 | 0.03 | 0.02 | [-0.01, 0.06] | .00 |
| Stimulation | -.06 | -0.03 | 0.02 | [-0.07, 0.01] | .00 | .01 | 0.01 | 0.02 | [-0.04, 0.05] | .00 |
| Power | -.31 | -0.16** | 0.02 | [-0.20, -0.12] | .07 | .32 | 0.18** | 0.03 | [0.12, 0.23] | .06 |
| Achievement | .10 | 0.07* | 0.03 | [0.01, 0.12] | .01 | -.08 | -0.06 | 0.03 | [-0.12, 0.01] | .00 |
| Benevolence | .13 | 0.11** | 0.03 | [0.04, 0.17] | .01 | -.08 | -0.07 | 0.04 | [-0.15, 0.01] | .00 |
| Universalism | .04 | 0.03 | 0.03 | [-0.03, 0.09] | .00 | .01 | 0.01 | 0.04 | [-0.06, 0.08] | .00 |
| Conformity | -.06 | -0.05 | 0.03 | [-0.11, 0.01] | .00 | .11 | 0.10** | 0.04 | [0.02, 0.17] | .01 |
| Group security | .18 | 0.14** | 0.03 | [0.07, 0.20] | .02 | -.06 | -0.05 | 0.04 | [-0.13, 0.03] | .00 |
| Tradition | -.01 | -0.01 | 0.03 | [-0.06, 0.04] | .00 | .01 | 0.01 | 0.03 | [-0.06, 0.07] | .00 |
| ΔR2 | .26** | .16** | ||||||||
| F(11, 788) | 76.70** | 51.70** | ||||||||
| R2 Total | .52 | .42 | ||||||||
*p < .05. **p < .01.
The results shown in Table 1 are not fully consistent with our initial hypotheses, although the discrepancies are not so difficult to explain. First of all, intrinsic as well as extrinsic freedom was favored by self-direction at the same strength (H1). Secondly, extrinsic freedom had the strongest link to the self-enhancement motives of power, control and domination (H2). These are also the motives that most strongly differentiate individuals with an extrinsic sense of freedom from individuals with an intrinsic sense of freedom, as the latter is negatively related to power. Thirdly, intrinsic freedom has been positively predicted by benevolence (which belongs to self-transcendence values) and group security. Although, in theory, group security belongs to conservation values, our additional analyses have shown that it was more highly correlated with self-transcendence values (especially with benevolence) than with conservation values.
Moral Intuitions
Table 2 shows the results of another pair of regression analyses. In the first analysis the intrinsic freedom scale was regressed on moral intuitions, including the statistical control of extrinsic freedom (covariate). The effects of six moral intuitions contributed 17% variance to the total R2 over and above the covariate effect. In the second analysis the extrinsic freedom scale was regressed on moral intuitions including the statistical control of intrinsic freedom (covariate). In this case, the effects of moral intuitions contributed 8% variance to the total R2 over and above the covariate effect.
Table 2
Effects of Moral Intuitions on Intrinsic and Extrinsic Sense of Freedom – Regression Analysis
| Predictor | Intrinsic freedom
|
Extrinsic freedom
|
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | B | SE | 95% CI | η2 | β | B | SE | 95% CI | η2 | |
| Constant | 1.30** | 0.14 | [1.01, 1.63] | 0.66** | 0.18 | [0.31, 1.20] | ||||
| Intrinsic freedom | .49 | 0.55** | 0.04 | [0.47, 0.62] | ||||||
| Extrinsic freedom | .43 | 0.39** | 0.03 | [0.33, 0.44] | ||||||
| Care/harm | .28 | 0.24** | 0.04 | [0.16, 0.31] | .05 | -.14 | -0.15** | 0.04 | [-0.23, -0.05] | .01 |
| Fairness/reciprocity | .20 | 0.18** | 0.04 | [0.10, 0.26] | .02 | -.07 | -0.07 | 0.05 | [-0.19, 0.01] | .00 |
| Ingroup/loyalty | -.14 | -0.12** | 0.03 | [-0.19, -0.05] | .01 | .16 | 0.16** | 0.04 | [0.07, 0.25] | .02 |
| Authority/respect | -.10 | -0.08** | 0.03 | [-0.15, -0.02] | .01 | .10 | 0.09** | 0.04 | [0.01, 0.17] | .01 |
| Sanctity/degradation | .03 | 0.03 | 0.03 | [-0.05, 0.10] | .00 | .00 | 0.00 | 0.08 | [-0.08, 0.09] | .00 |
| Liberty | .05 | 0.04 | 0.03 | [-0.02, 0.10] | .00 | .20 | 0.20** | 0.04 | [0.12, 0.27] | .03 |
| ΔR2 | .17** | .08** | ||||||||
| F (7, 792) | 86.20** | 59.30** | ||||||||
| R2 Total | .43 | .34 | ||||||||
**p < .01.
As expected in H4, intrinsic freedom was positively predicted by moral intuitions of care/harm (β = .28, p < .001) and fairness/reciprocity (β = .20, p < .001). Moreover, it was also - though to a lesser extent - negatively predicted by moral intuitions of ingroup/loyalty (β = -.14, p = .001) and authority/respect (β = -.10, p = .009). In the second regression analysis, as expected (H5), the extrinsic freedom was positively predicted by moral intuitions of liberty (β = .20, p < .001). In addition, it was negatively predicted by care/harm (β = -.15, p = .002), and - quite unexpectedly - positively predicted by ingroup/loyalty (β = .16, p < .001) and authority/respect (β = .10, p = .020).
Preferences for Political Community
In this section we tested the predictive effects of intrinsic and extrinsic sense of freedom on the preferred type of political community. As we argued, preference for communitarian political community should be favored by an intrinsic sense of freedom, while the preference for liberal political community can be favored by both intrinsic and extrinsic freedom.
Table 3 shows the results of two regression analyses. In the first, liberal orientation (M = 4.28, SD = 0.64) was regressed on intrinsic and extrinsic sense of freedom, including the statistical control of communitarian orientation’s effect (covariate). The effects of both predictors contributed 8% variance to the total R2 over and above the covariate effect. In the second regression analysis the scores on the communitarian orientation scale (M = 4.57, SD = 0.62) were regressed on both types of freedom, including the statistical control of liberal orientation’s effect (covariate). In this case, the effects of both predictors contributed 10% variance to the total R2 over and above the covariate effect.
Table 3
Effects of Intrinsic and Extrinsic Sense of Freedom on the Preference for Liberal/Communitarian Political Community – Regression Analysis
| Predictor | Liberal orientation
|
Communitarian orientation
|
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | B | SE | 95% CI | η2 | β | B | SE | 95% CI | η2 | |
| Constant | 1.60** | 0.17 | [1.31, 1.92] | 1.20** | 0.16 | [0.83, 1.50] | ||||
| Liberal orientation | .50 | 0.48** | 0.03 | [0.42, 0.54] | ||||||
| Communitarian orient. | .51 | 0.53** | 0.03 | [0.47, 0.59] | ||||||
| Intrinsic freedom | -.23 | -0.26** | 0.04 | [-0.34, -0.18] | .05 | .35 | 0.37** | 0.04 | [0.30, 0.44] | .12 |
| Extrinsic freedom | .33 | 0.35** | 0.03 | [0.28, 0.41] | .11 | -.05 | -0.05 | 0.03 | [-0.11, 0.02] | .00 |
| ΔR2 | .08** | .10** | ||||||||
| F(3, 796) | 151.30** | 159.60** | ||||||||
| R2 Total | .36 | .38 | ||||||||
**p < .01.
Analysis of the β coefficients shows that the increase in intrinsic freedom predicted a decrease in preferences for liberal orientation (β = -23, p <.001), and – as expected (H6) – it was considerably associated with an increase in preferences for communitarian orientation (β = .35; p < .001). On the other hand, the increase in extrinsic freedom – as expected (H7) – considerably predicted an increase in preferences for liberal orientation (β = .33; p <.001) and was unrelated to the preference for communitarian orientation (β = -.05; p = .184).
Support for the Principles of Liberal Democracy
We expected that both intrinsic and extrinsic sense of freedom favor support for democratic principles. However, as it was argued, their potential influence should be different. The effect of intrinsic freedom was expected to be positive and direct, whereas the effect of extrinsic freedom was expected to be positive but mediated by the preference for a liberal orientation. Table 4 presents verification of both hypotheses.
Table 4
Direct and Indirect Effects of Intrinsic and Extrinsic Freedom on Support for Democratic Principles
| Direct effects | |||||||
|---|---|---|---|---|---|---|---|
| Predictor | r | βstep 1 | βstep 2 | B | SE | 95% CI | η2 |
| Constant | 0.75** | 0.24 | [0.27, 1.23] | ||||
| Intrinsic freedom | .33 | .18 | .15 | 0.16** | 0.06 | [0.04, 0.28] | .02 |
| Extrinsic freedom | .37 | .27 | .05 | 0.05 | 0.06 | [-0.06, 0.16] | .00 |
| Liberal orientation | .57 | .41 | 0.44** | 0.05 | [0.33, 0.54] | .14 | |
| Communitarian orientation | .51 | .22 | 0.23** | 0.06 | [0.12, 0.34] | .04 | |
| ΔR2 | .15** | .26** | |||||
| F(4, 395) | 67.80** | ||||||
| R2 Total | .41 | ||||||
| Indirect effects | |||||||
| Mediator | B | SE | Boot 95% CI | ||||
| Intrinsic freedom | |||||||
| 1.1 Liberal orientation | -0.15 | 0.03 | [-0.22, -0.09] | ||||
| 1.2 Communitarian orientation | 0.11 | 0.03 | [0.05, 0.19] | ||||
| Extrinsic freedom | |||||||
| 2.1 Liberal orientation | 0.19 | 0.03 | [0.12, 0.26] | ||||
| 2.2 Communitarian orientation | -0.03 | 0.02 | [-0.07, 0.01] | ||||
**p < .01.
At the level of Pearson's correlation intrinsic and extrinsic freedom to a similar degree favor support for democratic principles (r = .33 and .37, p < .001, respectively). In the first step of hierarchical regression analysis, support for democratic principles (M = 4.52, SD = 0.65) is still positively predicted by both freedoms when we control for their covariance (β = .18 and .27, p < .001; respectively). The second step of the analysis shows that both liberal and communitarian orientation favor support for democratic principles, although the effect of the former is clearly larger (β = .41 and .22, p < .001; respectively). The effect of intrinsic freedom remains statistically significant (β = .15, p = .008), while the effect of extrinsic freedom becomes insignificant. Thus, in terms of the mediation analysis, we found the expected direct effect for intrinsic freedom and no such effect for extrinsic freedom (H8).
The above results suggested mediation effects. To precisely determine their size and statistical significance, we used the PROCESS v 2.15 macro for testing mediation effects (Model 4) (Hayes, 2014). Of the four separate analyses, the first two concerned the effect of intrinsic freedom on support for democratic principles, separately mediated by liberal (1.1) and communitarian (1.2) orientation. In analysis 1.1, a mediating variable as well as a dependent variable were controlled for by two covariates: extrinsic freedom and communitarian orientation, whereas in analysis 1.2, the covariates were extrinsic freedom and liberal orientation. We found two significant indirect effects of intrinsic freedom: the negative effect mediated by liberal orientation (B = -0.15), and the positive one mediated by communitarian orientation (B = 0.11).
Similarly, two subsequent analyses concerned the effect of extrinsic freedom on support for democratic principles, separately mediated by liberal (2.1) and communitarian (2.2) orientation. In analysis 2.1, a mediating variable as well as a dependent variable were controlled for by two covariates: intrinsic freedom and communitarian orientation, whereas in analysis 2.2, the covariates were intrinsic freedom and liberal orientation. We found only one significant indirect effect of extrinsic freedom – the expected (H9) positive effect mediated by liberal orientation (B = 0.19).
In sum, the presented results suggest that extrinsic freedom is not a direct predictor of support for principles of liberal democracy. However, this form of freedom seems to considerably affect support for democracy by increasing preferences for the liberal vision of political community. In the case of extrinsic freedom, the picture of possible influences is more complex. On the one hand, as was expected, intrinsic freedom turned out to be directly supportive for liberal democracy. However, it can also generate two indirect effects that are antagonistic to one another, and largely cancel each other out. Since both liberal and communitarian orientation favor support for democracy, it seems that in the case of intrinsic freedom this support can, in a parallel manner, be weakened and strengthened – as intrinsic freedom decreases liberal orientation and increases communitarian orientation, respectively.
Discussion
Some normative theories of liberal democracy contrast the two forms of freedom: 'unconditional' (absolute, not limited) and 'conditional' (relative, limited). According to political and social philosophers like Fromm (1970/1941) and Berlin (1969), the essence of unconditional freedom (‘freedom from’ or ‘negative freedom’) is the priority of individual autonomy and independence over the achievement of other important social goals and values. Consequently, freedom becomes the most important political value. In the case of conditional freedom (‘freedom to’ or ‘positive freedom’), the priority is to enable the realization of people’s individual life potential, which goes far beyond maximizing individual autonomy and independence. Consequently, maximizing individual autonomy and independence must take into account a wider range of important social goals and values.
The aim of the research described in this paper was an attempt to capture the psychological sense of the philosophical distinction of unconditional and conditional freedom through the distinction of extrinsic and intrinsic experience of freedom. In the course of the analysis, it turned out that both hypothetical forms of experiencing freedom are quite strongly positively correlated, but empirically distinct. Moreover, our analyses showed that they are related to different sets of human values and moral intuitions. As for human values, the only similarity between extrinsic and intrinsic freedom lies in the expected positive relationship with self-direction (intellectual autonomy and independence of action). The most striking difference appeared in relation to the value of power, which turned out to be the most important but opposite predictor. The broadly understood motive of power, control and domination definitely seems to favor extrinsic freedom and disfavor intrinsic freedom. Just as with the Americans examined by Alford (2003), Poles who understand freedom as being free from external limitations, perceive the origin of freedom in having power and resources that allow the control and domination of others (see also. Hamilton, 1978; Doliński, 1993). In contrast, in the case of intrinsic freedom, pro-social motives of close ties with people and collective security are clearly noticeable.
Even more pronounced mental differences are seen in the related moral foundations. Intrinsic freedom is clearly related to moral intuitions based on care for others and fairness in social relations. These positive relationships with the ethics of individual welfare are accompanied by negative relations with the moral codes that make up the ethics of community, i.e., loyalty to the group and respect for authorities. In contrast, extrinsic freedom has a completely different pattern of moral correlates. Firstly, as one might expect, its strongest moral predictor is liberty. Secondly, extrinsic freedom tends to be negatively related to the ethics of individual welfare (especially care/harm) and positively related to the ethics of community (especially ingroup/loyalty). This pattern was not expected at all, but it seems to harmonize with the earlier finding that extrinsic freedom is considerably related to the value of power. It would also be in line with our previous research, in which support for authoritarian power was favored by unlimited freedom and significantly weakened by freedom limited by the law and respect for other people (Skarżyńska, 2018).
Many definitions of democracy postulate that the goal of creating political communities based on a democratic legal order is to pursue a 'common good'. The state, as a formal and legal emanation of such a community, arises as a result of self-organization of citizens who want specific forms of relations between themselves, and in relation to the institutions of the state. We assumed that the broadly understood model of a democratic political community is based on at least two different concepts of citizenship and relations between the state and citizens – a liberal or communitarian orientation. As could be expected, extrinsic freedom very clearly favors preferences of a political community based on the principles of liberal orientation, i.e., on the belief that the most important goal of the state is to provide every citizen with maximum freedom through appropriate legal rules (human and civil rights). Because society is nothing more than a network of interactions and objective-mercantile interdependencies between individuals, these rights cannot be subordinated to a society understood as an autonomous, real social existence. The extrinsic sense of freedom clearly determines preference for the liberal model of political community, while intrinsic freedom is related to the rejection of liberal community. Whereas intrinsic freedom definitely favors communitarian orientation - it advocates a vision of the society having its own developmental goals and guided by overriding collective ethics. Communitarians believe that personal freedom is something highly unstable and illusory if it is disconnected from communal values.
Our research proves that both models, liberal and communitarian, can support democracy. What is more, in terms of direct effects, the effect of liberal orientation turns out to be clearly stronger than the effect of communitarian orientation. However, paradoxically, the same extrinsic freedom, which is so favorable to liberal orientation, appeared as only indirectly conducive to democratic thinking.
This is not surprising if we take into account that unconditional freedom is rather grounded in striving for power and domination than in the motives of self-transcendence and concern for others. In this sense, norms and principles of the liberal political community seem to fulfill a dual function with respect to extrinsic freedom. On the one hand, they may provide ideological rationalization for its expansiveness, but on the other hand they may also impose strong limits on its selfish potential.
People who experience intrinsic freedom seem to first of all draw on the motives of self-transcendence and concern for others. Such values as power and domination certainly do not belong to their priorities. In contrast to extrinsic freedom, relationships between intrinsic freedom and democratic principles turned out to be much more complex. Based on our analysis, the total effect of intrinsic freedom on support for democracy can be decomposed into three components: 1) positive direct effect; 2) positive indirect effect (by strengthening communitarian orientation); and 3) negative indirect effect (by weakening liberal orientation).
Effects number 2 and 3 were not included in the set of initial hypotheses. However, in the light of other results presented in this paper, effect 2 in particular is not much of a surprise. As we argued and expected, intrinsic sense of freedom favored communitarian orientation towards the political community, and communitarian orientation, like the liberal one albeit weaker, favored support for democratic principles. In the case of effect 3, which is antagonistic to effects 1 and 2, a question arises about its meaning. Does it suggest a manifestation of a hidden authoritarian potential, or is it evidence for thoughtful use of freedom? We cannot give a fully satisfactory answer on the basis of the data we have. However, taking into account effects 1 and 2, we suppose that the second interpretation is more likely to be accurate. The negative relationship between intrinsic freedom and liberal orientation seems primarily to be a kind of reluctance towards fetishization of liberty at the expense of other social and political values.
Since intrinsic freedom turns out to be more pro-democratic than extrinsic freedom, it is worth considering the factors strengthening this form of experiencing freedom. We believe that, apart from personal values and moral intuitions, an important role is played by individual differences in cognitive and behavioral control. A strong sense of internal control should favor the experience of positive freedom (in Berlin's view), that is, freedom to pursue one's goals and values, limited by concern for the freedom and fair treatment of other people. It is already known that sense of internal control (cognitive and behavioral) – and thus also the experience of positive freedom – is associated with a specific individual self-image and image of the social world. Research, conducted since the mid-1990s by Carol Dweck and her colleagues, proves that internal control is greater the more people believe that personality traits, including intelligence, can be successfully developed (Dweck, 2000). In turn, belief in the developmental potential of the human mind is positively related to persistence, resilience, and deliberation in judging people. All these characteristics help in achieving planned goals and personal development. It can therefore be assumed that experiencing intrinsic freedom is not only pro-democratic, but gives people clear developmental benefits.
Finally, it is worth noting that the concept of intrinsic and extrinsic freedom is perhaps a reasonable answer to the ‘secret’ of a positive and strong relationship between liberal and communitarian orientation (which amounted to r = .53). Though, at the theoretical level, both orientations may seem contradictory, at the level of mental representation (and our measurement) they are not contradictory and even strongly coexist. Moreover, they both make their substantial contribution to support for democracy. It can be argued that this is because most people equally often experience freedom in its extrinsic as well as intrinsic sense.
Both orientations propose different concepts of building political community, but liberals as well as communitarians can similarly believe in democracy as a better model of social order than other political systems, having many more advantages than disadvantages. Liberals as well as communitarians can also share similar ideas about the standards of the democratic state, such as free elections, political pluralism, freedom of economic activity, the rule of law, equality of voting rights, etc. It can be assumed that this similarity is responsible for a large part of the common variance. Most disputes between advocates of liberal and communitarian orientation concern such important aspects of democracy as the scope and strength of respect for human rights, and the attitude towards authoritarian forms of government. Such differences have perhaps the greatest impact on the different political identification and preferences expected between liberals and communitarians.
If we refer our results to the theory of social identity (cf. Tajfel & Turner, 1986), the phenomenon of coexistence of extrinsic and intrinsic forms of freedom seems to correspond with the description of a general mechanism shaping human identity. It describes the relations between two spheres of personal identity: individual and social. Individual identity is the manifestation of one's need for autonomy and uniqueness - it is formed as a result of the processes of individuation, which leads to separation of the Self from non-Self. The basis for social identity lies in identification processes, thanks to which we are able to recognize ourselves as a part of a collective. The alternating experience of extrinsic and intrinsic freedom can be seen as parallel to the alternating experience of individualization and identification. This phenomenon can be explained in terms of the theory of self-categorization (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), which emphasizes that group categorization processes allow people to smoothly shift (self-categorization) from the level of individual identity (autonomous Self) to the level of group identity, as well as ‘switching’ different group identities. Giving greater importance to the individual or social identity depends on both personality factors and social context. In terms of personality characteristics, for example, it has been shown that increased level of anxiety leads to an increase in the tendency to identify with others (identification) and a decrease in individuation (e.g., Maslach, Stapp, & Santee, 1985). As for contextual factors, mental "switching" from one way of thinking about political community to another often results from changes in the sense of personal and/or group security perceived by large social groups. Another reason may be the phenomenon of surfeit (cf. Apter, 1984). In certain historical periods, one or the other orientation dominates in social discourse, reinforced by the policy of the ruling elite and the media narrative. In Poland, an individualistic narrative prevailed from the beginning of the systemic change. From 2015, after the change of the political power from rather liberal to conservative, a communitarian narrative became much more expressive and popular.