The 2015 edition of Caritas Europa’s Crisis Monitoring Report suggests that, after six years of crisis, people in the seven countries analysed (Cyprus, Greece, Ireland, Italy, Portugal, Romania, and Spain) are still suffering from the consequences of the adopted measures, namely because the number of unemployed people or people living in poverty or at risk of poverty is still huge, whereas debt levels continue to be enormous with little economic growth.

The report of Caritas Europa also states that the economic crisis stemmed from economic theories and market liberalism policies that arose in the 1970s. These policies led to a crisis in private finance, which was transformed into a public finance crisis thus transferring the burden of its costs to the citizens (Zacune, 2013). Indeed, the dominant view in the European Union was that the crisis resulted from the profligacy of the peripheral countries and that, consequently, austerity should be prescribed. No consideration was given to the fact that these countries had growing trade deficits that would only be worsened by austerity measures which would impede economic recovery (Blankenburg, King, Konzelmann, & Wilkinson, 2013).

Indeed, prior to the crisis, the three countries (Greece, Portugal and Italy) which are the object of present analysis, experienced different situations because of numerous national specificities. The levels of growth in Italy and Portugal were stagnating, whereas there were high levels of government debt in Italy and Greece and, to a lesser extent, in Portugal (Caritas, 2015).

In 2010, Greece was the first country to receive financial assistance under supervision of Troika (the European Commission, the European Central Bank and the International Monetary Fund), followed in 2011 by Portugal, whereas in 2011 Italy was going through a process aiming at correcting its excessive deficit. In Italy, the adjustment measures involved more tax increases than cuts in expenditure, in contrast to the measures undertaken in Greece and Portugal (Caritas, 2015).

Despite the fact that high levels of debt were not the cause of the lack of economic growth, the cuts in government budgets were justified by the need to promote business confidence through reduction in expenditure (Quiggin, 2011). The austerity measures, which ignored social priorities, imposed increases in direct and indirect taxation, reductions in public wages and in public administration staff, reductions in social benefits and cuts in the education and health sectors (Caritas, 2015).

The economic impact of these measures was that government debts showed a continuous yearly increase (Krugman, 2013), confirming that the harsher the austerity, the higher the increase in debt-to-GDP ratio (Cavero & Poinasamy, 2013). Their social impact was a rise in poverty and social exclusion, greater precariousness with temporary and part-time employment increasing and permanent and full-time employment decreasing (Caritas, 2015), as well as historically high unemployment rates (Eurostat, 2015a). The massive cuts in education and health not only had negative effects for social justice but will also have long-term consequences for the economy (Schraad-Tischler & Kroll, 2014).

Lay people’s representations of the economic crisis

When a complex phenomenon produces such significant consequences, lay people seek to understand it. For this purpose, they watch and read the news, look at statistics and listen to the opinions of specialists that are diffused by the mass media. The information gathered is then shared in conversations that take place in social groups: points of view are confronted, judgments are compared and social representations are formed (Moscovici, 1976; see also Poeschl, 2003). These representations constitute in fact lay theories which, in the case of the economic crisis, include opinions about the causes and consequences of the crisis, for example, and the ways to deal with the situation. These representations have an impact on people’s beliefs and attitudes and orientate their behaviours (Jodelet, 1984).

Because of the importance of these representations in understanding citizens’ reactions to the situation created by the financial and economic crisis, authors from various domains became interested in knowing what opinions had been formed on the crisis in several countries or social groups (Papastamou et al., 2017, this issue), and some studies have already been undertaken in the theoretical framework of social representations.

A study conducted by means of a free association task compared the discourses of small retailers, bankers and economics students from Greece, Italy, France and Romania (Galli et al., 2010). Despite the economic and sociocultural differences between the countries, the results revealed globally similar representations structured around the word “unemployment”. In addition to “unemployment”, the central elements of the representation (i.e. the elements that generate and organize social representations (Abric, 1989, for example)) included in Greece the word “poverty” and in Italy, they included “poverty” as well as “money” and “work”. The representations of economics students were more theoretical and more structured, whereas the representations of small retailers, the group most affected by the crisis, included more elements of a psychological nature (e.g., “fear”) and economics (e.g., “company” or “savings”).

Another study using a free association task compared the discourses of Portuguese students based on their family income (Poeschl, Valentim, & Silva, 2015). The word most frequently associated with the crisis was “poverty”. The other most frequent words were “unemployment” and “money”, confirming the central status of these words, as well as “difficulties”, confirming the central status of these words. A comparison of the global discourses evoked by the students according to their family income revealed that the consequences of the crisis were described in terms of misery, difficulties, or change, depending on family income, whereas words that highlighted the factors responsible for the crisis, such as debts, mismanagement, Troika, or banks were common to all groups.

In turn, a study conducted in Greece by means of a questionnaire composed of opinion scales examined the impact of several factors on the willingness of the population to participate in actions against austerity measures (Chryssochoou, Papastamou, & Prodromitis, 2013). Underlining the specificity of the economic crisis that affects large parts of the population, albeit differently, the authors analysed the extent to which the reactions to the crisis were dependent on people’s actual financial position, their sense of grievances, feelings of vulnerability and their emotions about the crisis. The results revealed that respondents differed in their degree of willingness to participate in the different types of actions analysed by the authors: they were favourable to national products consumption and individual reactions (e.g., increasing one’s skills through education and training to be able to cope with the demands of the labour market); they were not clearly positioned with regard to activism (e.g., occupying public buildings and ministries) and the usual forms of participation (e.g., taking action through participation in political parties); and they rejected seeking financial security (e.g., taking one’s savings out of Greece to foreign banks in order to protect them), using the internet to call for political disobedience (e.g., publishing texts on the internet approving acts of political violence) and, above all, resorting to violence (e.g., placing incendiary devices in public buildings). Results also suggested that people with the lowest family incomes were more willing to participate in traditional forms of political action, such as joining political parties or labour unions, and to seek financial security, but were not more willing to take part in violent action than others.

Using some of the opinion scales developed by the Greek authors (Chryssochoou, Papastamou, & Prodromitis, 2013; Papastamou et al., 2017), the above-mentioned Portuguese study revealed that the differences in students’ family income not only produced differences in the evoked consequences of the crisis but also differences in emotions and the sentiments of vulnerability and relative deprivation that resulted from comparisons with others. Respondents with a lower family income experienced less positive emotions and more negative emotions toward the crisis, as well as stronger feelings of vulnerability and relative deprivation that resulted from comparison with others (Poeschl, Valentim, & Silva, 2015). However, respondents did not differ in their expectations about their future situation, which they foresaw without great changes, and, unlike the Greek respondents, declared a similar unwillingness as the other respondents to participate in protest actions. Indeed, the students rejected the three types of actions analysed in the study: economic resistance (such as refusing to pay any new taxes, taking one’s savings out of Portugal to foreign banks in order to protect them, or emigrating to a prosperous country to find a job); civic participation (e.g., taking action through participation in labour unions or political parties, or signing petitions), and, above all, participation in violent actions (such as burning cars of politicians, physically assaulting politicians, or destroying public property).

Perceived efficacy of actions against austerity measures

The differences between the results of the Greek and the Portuguese studies led us to examine how respondents perceive the efficacy of the proposed actions. Indeed, authors working on social movements have observed that people’s concerns about social and economic issues do not always lead them to engage in protest actions, and the broad literature about the factors likely to predict willingness to take part in collective action includes the perceived efficacy of actions (Brunsting & Postmes, 2002; Hornsey et al., 2006; Tausch et al., 2011). However, because the research on the relationship between perceived efficacy and intention to engage in collective action has led to inconsistent results (Hornsey et al., 2006), most models which were developed as an attempt to understand why people take part in public protests include, in addition to perceived efficacy of actions, identification with social categories struggling to achieve change, perception of grievances, and emotions (Gamson, 1991; Van Stekelenburg & Klandermans, 2010).

Although the literature on social movements enables us to assume that the willingness to engage in protest actions depends, at least partially, on the perception of the efficacy of actions, to our knowledge little attention has been given to the conditions that lead people to view specific actions as effective, motivating them to engage in collective action or to adopt individual solutions. It seems obvious, however, that people’s assessment of the efficacy of alternative forms of action depends on a great number of factors. The various forms of protest actions that may possibly be taken against austerity measures implemented in the present financial and economic crisis offered us the opportunity to analyse some of those factors. In our exploratory analysis of the predictors of the perceived efficacy of actions against austerity measures, we included societal, representational and individual determinants to understand people’s way of thinking, in line with the perspective of Doise (1986, 1992; Doise & Valentim, 2015).

For the analysis of societal determinants, we took into consideration the socio-economic situation of the countries, which presented great differences in 2011, when data began to be collected, first in Portugal, and then in Greece and Italy in 2012. Thus, 2011 indicators showed, for example, differences in (i) the unemployment rate (Greece: 17.9%; Portugal: 12.9% and Italy: 8.4%) (Eurostat, 2015a), (ii) the general government gross debt as a percentage of the gross domestic product (Greece: 172%; Portugal: 111.4%; Italy: 116.4%) (Eurostat, 2015b), or (iii) the human development index, a composite statistic based on life expectancy, education, and per capita income (Greece: 0.854; Portugal: 0.819; Italy: 0.872) (UNDP, 2014). In sum, the socioeconomic situation of Greece, in 2011, appeared to be the worst of the three countries. Portugal had, however, the lowest score of human development, and Italy, although hit by the crisis, appeared to be most well-off of the three countries.

For the analysis of representational determinants, we considered that the perceived efficacy of the actions against austerity measures might be predicted by two types of crisis-related variables: (i) the attribution of responsibility for the crisis and (ii) the predicted date for the crisis to come to an end. More precisely, we assumed that the perceived efficacy of different types of actions should depend on whom they are directed at; we also assumed that respondents’ perception of the importance of the sacrifices citizens would be asked to make because of austerity measures implemented to address mainly private problems would influence the way they look at possible responses. In other words, the perception that the crisis would last longer should lead them to give greater support to all types of actions.

Finally, for the analysis of individual determinants, we considered that the perceived efficacy of the actions might be modulated by respondents’ (i) political orientation, (ii) feelings of depression felt in the four weeks preceding the study and (iii) sex. We considered that, because the austerity measures were adopted by a right-wing government at the time data were collected, leftists would be more prone to support all types of protest actions; that, in accordance with Chryssochoou, Papastamou and Prodromitis (2013), a depressive state should be associated with inaction and, therefore, the more depressed the respondents, the more they should view all actions as ineffective; and, finally, that sex should be taken into account since women are overrepresented in the vulnerable groups which are the hardest hit by austerity (Hendra, 2013), and because they are viewed as using a less aggressive and competitive negotiating style than men (Dobrijevic, 2014).

In order to analyse the perceived efficacy of the actions listed in the questionnaire, we began by extracting different types of actions to cope with the crisis. Then, after verifying that they were related to the willingness to take action (cf. Chryssochoou et al., 2010; Poeschl et al., 2015), we looked for the relative weight of the societal, crisis-related and individual-related variables for the assessment of their efficacy.

Method

Respondents

Four hundred and fifty students took part in the study, 161 men and 289 women aged between 18 and 60 years (21.82 years on average). There were 135 Greek respondents, 68 men and 67 women, aged between 18 and 34 years (21.54 years on average), 154 Italian respondents, 39 men and 115 women, aged between 18 and 29 years (20.76 years on average), and 161 Portuguese respondents, 54 men and 107 women, aged between 18 and 60 years (23.18 years on average). The original sample of 453 Portuguese students was randomly reduced to correspond to the size of the Greek and Italian subsamples.

Questionnaire

There were several parts to the questionnaire, each focussing on different aspects of the crisis. In this paper, we analyse the perceived efficacy (1 = not at all; 7 = very much) of 32 types of actions (see Table 1) that people could adopt to react against austerity measures taken in the context of the current economic crisis and verify that they are correlated with their willingness to participate in them. To test the effect on the rating of their efficacy of the three types of above-mentioned determinants (societal, representational and individual), in addition to the country as the societal determinant, we used the following variables:

Table 1

Efficacy of the proposed actions to cope with the crisis. Factorial solution after varimax rotation. Percentage of explained variance and values of Cronbach’s alpha.

F1 16.53% F2 12.07% F3 11.36% F4 10.75% F5 6.91% F6 5.29%

Violence (α = .90)
Physical assault on politicians .82 –.05 .08 .19 .14 .05
Placing incendiary devices in public buildings .81 .02 .11 –.01 .01 .07
Burning politicians’ cars .78 –.04 .18 .16 .20 .07
Destroying public property .73 .10 .07 .08 .12 –.05
Publication of texts on the internet approving acts of political violence .64 .05 .11 .37 –.05 .11
Throwing eggs and yogurt at politicians .62 –.04 .26 .16 .25 .04
Attacking anti-riot police forces during demonstrations .58 .15 .05 .51 .08 –.01
Publication of texts on the internet inciting public disobedience .56 .08 .29 .50 –.05 .12
Civic participation (α = .83)
Acting through participation in unions .04 .77 .24 .15 –.00 .00
Acting through participation in political parties –.07 .74 –.06 –.14 .02 .13
Acting through participation in NGOs –.08 .73 .14 .07 .10 .06
Participating in popular assemblies in neighbourhoods .07 .71 .24 .27 –.05 .13
Sending political e-mails .17 .67 .20 .06 .10 .02
Constructing a website/blog .07 .57 .36 .06 .04 –.15
Political resistance to government measures (α = .85)
Striking .09 .27 .73 .14 .01 .02
Blocking roads, ports and airports .29 .11 .64 .25 .04 –.01
Occupation of public buildings and ministries .34 .12 .62 .35 .01 .02
Participation in public demonstrations against austerity measures .11 .29 .61 .27 .07 .23
Participating in rallies of Indignant Citizens .19 .39 .61 .20 .05 .08
Signing a petition .05 .47 .61 –.08 .18 –.12
Economic resistance to government measures (α = .84)
Refusal to pay any new taxes .21 .10 .38 .71 .22 .12
Refusal to pay tolls .20 .08 .40 .70 .23 .13
Illegally reconnecting power to those that could not pay electricity bills .47 .06 .23 .59 –.10 .15
Collectively withdrawing all money from banks .23 .07 .12 .56 .38 –.09
Stealing food from supermarkets and distributing it to the poor .53 .12 .09 .53 –.02 .04
Individual financial protection (α = .66)
Taking one’s savings out of the country to foreign banks to protect them .11 .07 –.03 .16 .78 .05
Emigrating to a prosperous country to find a job .04 –.02 .20 –.07 .73 .14
Keeping money at home to be able to cope with tough moments .17 .13 –.01 .16 .67 .04
Protectionism (α = .66)
Buying only national products –.01 .10 –.03 –.04 .15 .88
Boycotting foreign products .21 .05 .13 .22 .07 .74

Representational Determinants

The representational determinants likely to affect the assessment of the efficacy of the proposed actions were measured by the following variables:

(i) Attribution of responsibility for the crisis. Respondents were asked to express to what degree they thought 19 elements were responsible for the crisis in their respective country (1 = not at all; 7 = very much). The principal component factor analysis, with varimax rotation, performed on the attribution of responsibility, extracted four factors, with eigenvalue higher than one, which explained 52.90% of the total variance (respectively, 16.76%, 16.46%, 11.80% and 7.88%). The factors referred to: people in general (employees, trade unionists, citizens in general, public sector employees, the extreme leftists and immigrants), external factors (the European Union, the USA, the powerful countries of the European Union, the international factor, globalization and capitalism); internal factors (respective countries’ State; past governments; current government; mentality and employers) and, finally, financial organizations (banks and the financial-credit organizations). Because the values of Cronbach’s alpha were satisfactory (respectively .79, .77, .67, and .68), we aggregated the items to form four attribution scales.

(ii) Predicted date for the crisis to come to an end. Respondents had to assess in which year they thought the economic crisis would come to an end in their country. The measure was a single 7-point scale, on which the respondents had to indicate a year between “2012” and “after 2021”.

Individual Determinants

The individual determinants likely to affect the assessment of the efficacy of the proposed actions were measured by the following variables:

(i) Political orientation. Respondents’ political orientation was measured by a single 10-point scale (1 = left; 10 = right).

(ii) Feelings of depression. Respondents’ feelings felt during the four weeks preceding the study were assessed by 12 items (for example, feeling sad and despairing; feeling worthless; feeling that life is not worth living), measured on a 7-point scale (1 = not at all; 7 = all the time), which were aggregated in a unique scale, in conformity with the procedure used by Chryssochoou, Papastamou and Prodromitis (2013). For our sample the internal consistency was satisfactory, α = .87.

(iii) Sex. Respondents’ sex was taken into account.

Results

Efficacy of the Proposed Actions to cope with the Crisis

After removing two items which did not saturate on any of the factors, “increasing one’s skills through education and training to be able to cope with the demands of the labour market” and “signing a petition to exit the Eurozone”, the principal component factor analysis, with varimax rotation, performed on the assessment of the efficacy of the proposed actions extracted six factors with eigenvalue higher than one. These explained 62.91% of the total variance and pointed, respectively, to violence, civic participation, political resistance to government measures, economic resistance to government measures, individual financial protection, and protectionism (cf. Table 1).

Following a classical distinction (Wright, Taylor, & Moghaddam, 1990), we may consider that civic participation and political resistance to government measures represent two different types of normative collective actions, that violence and economic resistance to government measures are two different types of non-normative collective actions, and that individual financial protection and protectionism constitute two different types of individual actions.

A comparison of the global means of the six types of actions indicated differences in the assessment of their efficacy, F (5, 2175) = 187.33, p < .001, all means significantly different according to the LSD test for multiple comparisons, p < .05, with the exception of political resistance to government measures and individual financial protection, which did not have different means. Thus, protectionism was the action judged most effective, followed by civic participation, political resistance to government measures and individual financial protection, economic resistance to government measures and, finally, violence (see Table 2). We may observe that, globally, the assessment of the efficacy of actions follows an order that goes from more pacific to more combative behaviours: protective individual or collective normative actions are viewed as more effective than individualist actions or confrontational collective non-normative actions.

Table 2

Means, standard deviations and Pearson’s r correlations between the perceived efficacy of the six types of actions and the willingness to participate in them.

r M (SD) Efficacy M (SD) Participation

Protectionism .80*** 4.29 (1.68) 4.15 (1.92)
Civic participation .74*** 3.85 (1.34) 3.19 (1.57)
Political resistance to government measures .84*** 3.63 (1.45) 3.24 (1.60)
Individual financial protection .72*** 3.58 (1.42) 3.53 (1.56)
Economic resistance to government measures .83*** 3.06 (1.55) 2.71 (1.57)
Violence .87*** 2.03 (1.25) 1.79 (1.15)

***: p < .001.

Relationship Between Perception of Efficacy and Willingness to take Action

Because the analyses performed on the willingness to take action had led to different factorial solutions in the Greek study (where seven factors were extracted, see Chryssochoou et al., 2013) and the Portuguese study (which proposed three factors, see Poeschl et al., 2015), we constructed analogous groups of items from the scale of the willingness to take action and found that Cronbach’s alpha values were high: protectionism (α = .75), civic participation (α = .87), political resistance to government measures (α = .87), individual financial protection (α = .64), economic resistance to government measures (α = .85), and violence (α = .90). The correlations between the perceived efficacy of the six types of actions and the willingness to participate in them are presented in Table 2.

As may be seen in Table 2, there is a strong correlation between perceived efficacy and willingness to engage in each type of actions. We may note, moreover, that the willingness to engage in the six types of actions basically follows the same order as the assessment of their efficacy, and that respondents were more likely to agree with the efficacy of the actions than be willing to participate in them (all differences significant, p < .015). There was, however, one notable exception to this rule: there was no difference between the assessment of the efficacy of individual financial protection and the readiness to execute this action. Indeed, respondents declared themselves more willing to adopt individual actions than to engage in normative collective actions and, above all, in non-normative collective actions.

Differences Between Countries in the Assessment of the Efficacy of Actions

The one-way analysis of variance performed on the six types of actions in order to test whether the respondents of the three countries differed in the assessment of their efficacy revealed differences with regard to almost all types of actions, with the exception of civic participation: F (2, 440) = 2.16, ns (protectionism: F (2, 446) = 22.89, p < .001; political resistance to government measures: F (2, 446) = 21.15, p < .001; individual financial protection: F (2, 446) = 5.28, p = .005; economic resistance to government measures: F (2, 446) = 68.78, p < .001; violence: F (2, 442) = 38.20, p < .001).

As may be seen in Table 3, Greek respondents considered all types of actions more effective than did the other respondents, except for civic participation (presenting no differences between countries) and individual financial protection. Indeed, Portuguese respondents agreed with Greek respondents about the efficacy of individual financial protection, differing from Italian respondents, who viewed this type of actions as less effective. Portuguese respondents also considered protectionism more effective than Italian respondents did, but agreed with them about the (low) efficacy of political and economic resistance to government measures, and about the (lack of) efficacy of violence.

Table 3

Efficacy of the six types of actions depending on countries. Means and standard deviations (in parentheses).

Greece Portugal Italy F

Protectionism 4.94a (1.62) 4.33b (1.57) 3.67c (1.63) 22.89***
Civic participation 4.05 (1.09) 3.80 (1.55) 3.73 (1.30) 2.16
Political resistance to government measures 4.28a (1.30) 3.35b (1.46) 3.35b (1.38) 21.15***
Individual financial protection 3.62ab (1.43) 3.81a (1.43) 3.30b (1.37) 5.28**
Economic resistance to government measures 4.22a (1.40) 2.62b (1.32) 2.52b (1.37) 68.78***
Violence 2.77a (1.32) 1.69b (1.00) 1.76b (1.15) 38.20***

***: p < .001; **: p < .01.

Different letters on different lines indicate significantly different means.

Because the perceived efficacy of the proposed actions is correlated with the willingness to participate in these actions, the differences observed here in the perceived efficacy of actions might help to explain the differences in the willingness to take action found in the two above-mentioned papers (Chryssochoou et al., 2013; Poeschl et al., 2015).

Predictors and Differences Between Countries with Regard to Predictors

A global analysis of the crisis-related variables revealed that there were differences in the attribution of responsibility for the crisis to the different entities, F (3, 1293) = 501.38, p < .001. Respondents primarily blamed financial organizations (M = 5.69), then internal factors (M = 5.23), external factors (M = 4.97) and refused to blame the people in general (M = 3.44), with all means significantly different according to the LSD test for multiple differences, p < .05. On average, the date for the end of the crisis was set at 2019 (M = 4.81) varying between 2017 and 2021, which means that respondents were expecting to have to face the crisis for another ten years.

The one-way analysis of variance performed on the crisis-related variables likely to influence the assessment of the efficacy of actions indicated significant differences between countries for almost all variables. As may be seen in Table 4, the responsibility of the financial organizations was consensual, F (2, 445) = .86, ns, but Italian respondents, who belong to the least affected country, were more reluctant to blame the other entities (internal factors: F (2, 438) = 27.71, p < .001; external factors: F (2, 441) = 13.24, p < .001; people in general: F (2, 440) = 16.88, p < .001). In addition, Greek respondents considered the end of the crisis more distant than Portuguese respondents, who also considered it more distant than Italian respondents, F (2, 441) = 41.50, p < .001. Thus, the socioeconomic situation of the countries influenced the way respondents analysed the crisis: the more affected the countries, the more they were likely to blame different entities and to present pessimistic expectations about the duration of the crisis and of the efforts citizens will be asked to make.

Table 4

Crisis-related and individual-related variables depending on countries. Means and standard deviations (in parentheses).

Greece Portugal Italy F

Crisis-related variables
Attribution to people in general 3.67a (1.23) 3.68a (1.16) 2.99b (1.14) 16.88***
Attribution to external factors 5.13a (0.97) 5.14a (0.95) 4.65b (0.93) 13.24***
Attribution to internal factors 5.57a (0.79) 5.37a (0.89) 4.81b (1.01) 27.71***
Attribution to financial organizations 5.78 (1.03) 5.68 (0.98) 5.62 (1.16) 0.86
Date for the crisis to come to an end 5.78a (1.31) 4.60b (1.69) 4.19c (1.48) 41.50***
Individual-related variables
Political orientation (left-right) 4.30b (2.15) 5.17a (1.99) 5.16a (2.30) 5.55**
Feelings of depression 2.58ab (0.82) 2.45b (0.97) 2.80a (1.02) 5.36**

***: p < .001; **: p < .01.

Different letters on different lines indicate significantly different means.

In relation to individual-related variables, there were differences in political orientation and in the intensity of the feelings of depression felt toward the crisis. Indeed, the Greek respondents declared, in general, a more left-wing political orientation than the other respondents, F (2, 344) = 5.55, p = .004, and, curiously, the Italian respondents declared a higher intensity of feelings of depression than the Portuguese respondents, F (2, 438) = 5.36, p = .005.

Predictors of the Assessment of Efficacy of Actions

Because the differences in the perceived efficacy of the types of actions may stem from the socioeconomic situation of the countries as well as from a whole set of other determinants, the effects of which should be disentangled, we performed six regression analyses (method: enter), one for each type of actions, introducing as independent variables the countries, crisis-related and individual-related variables. As may be observed in Table 5, all types of variables contribute to some extent to predicting the perceived efficacy of at least some types of actions, although one should keep in mind that the statistical significance of the small beta values of some predictors may be due to the large number of cases in the analyses.

Table 5

Societal, crisis-related, and individual-related variables as predictors of perceived efficacy of actions.

Protectionism Civic participation Political resistance to government measures Individual financial protection Economic resistance to govern-ment measures Violence

Societal variables (reference least affected country)
More affected +.13* +.12 +.03 +.09 +.06 –.04
Most affected +.27*** +.08 +.18** –.08 +.39*** +.21***
Crisis-related variables
Attribution to people in general +.03 –.16* –.29*** –.04 –.14* +.03
Attribution to external factors +.12 –.02 +.11 +.05 +.08 +.09
Attribution to internal factors +.14* +.07 +.16** +.28*** +.21*** –.03
Attribution to financial organizations +.05 +.09 +.04 –.01 +.04 +.00
Date for the crisis to come to an end –.05 +.02 +.03 +.09 +.01 +.12*
Individual-related variables
Political orientation (left-right) +.15* –.11 –.19*** +.04 –.20*** –.15**
Feelings of depression +.10 +.11 +.16** +.13* +.21*** +.23***
Sex (reference male) –.00 +.07 –.06 –.16** –.08 –.23***

***: p < .001; **: p < .01; *: p < .05.

Countries were represented as two dummy variables with Italy, the least affected country, serving as the reference group. Portugal (vs. Greece and Italy) is a more strongly affected country and Greece (vs. Italy and Portugal) the most strongly affected.

First, we can see that the efficacy of protectionism, the action judged most effective and one of the two individual actions, is predicted above all by the situation of the two more affected countries, by a crisis-related variable (the attribution of responsibility to internal factors) and by an individual-related variable (political orientation) R = .41, R2adj. = .14, F (10, 314) = 6.46, p < .001. Thus, the more the situation of the country worsens, the more internal factors are blamed and the more right-wing people’s political position is, the more protectionism is viewed as effective.

The efficacy of civic participation, R = .29, R2adj. = .05, F (10, 311) = 2.75, p = .003, is only predicted by one crisis-related variable, the (lack of) attribution of responsibility for the crisis to the people in general. So, the less the people are blamed, the more civic participation is considered effective.

The efficacy of political resistance to government measures, R = .52, R2adj. = .24, F (10, 314) = 11.42, p < .001, is predicted above all by two crisis-related variables (the attribution of responsibility for the crisis to internal factors and (not) to the people) and also by two individual-related variables (respondents’ political orientation and intensity of depressive feelings in relation to the crisis) and the situation of the country. So, when the crisis is attributed to internal factors and not to the people, when the feelings of depression in relation to the crisis are more intense, the participants more left-wing, and the situation of the country is worse, the more likely political resistance to government measures is viewed as effective.

The efficacy of individual financial protection, R = .40, R2adj. = .13, F (10, 314) = 5.86, p < .001, is above all predicted by one crisis-related variable (the attribution of the responsibility for the crisis to internal factors) and then by two individual-related variables (respondents’ feelings of depression and sex). In this case, the more internal factors are viewed as responsible for the crisis and the more intense the feelings of depression, the more individual financial protection, the second of the two individual actions, is viewed as effective, especially by male respondents.

The efficacy of economic resistance to government measures, R = .64, R2adj. = .39, F (10, 313) = 21.34, p < .001, is also predicted, like political resistance, by all types of variables: first, and more than political resistance, by the situation of the country and two individual-related variables (respondents’ political orientation and depressive feelings about the crisis) and then by two crisis-related variables (the attribution of responsibility for the crisis to internal factors and (not) to the people). So, when the situation of the country is worse, the feelings of depression in relation to the crisis more intense, the participants more left-wing, and the crisis is attributed to internal factors and not to the people, the more likely economic resistance to government measures is viewed as effective.

Finally, the efficacy of violence, R = .54, R2adj. = .27, F (10, 312) = 12.85, p < .001, is above all predicted by all individual-related variables (respondents’ political orientation, intensity of feelings, and sex) by the situation of the country, and a crisis-related variable (the prediction of how long the crisis would last). So, especially when the respondents are male, when the feelings of depression are more intense, and the respondents more left-wing, when the situation of the country is worse and there is a weaker hope of seeing citizens better off by getting out of the crisis, the more the respondents are likely to view violence as effective. It may thus be noted that violence (generally viewed as the least effective action (see Table 3)) is predicted by a particularly pessimistic view of the crisis, with a long expected duration, as well as strong feelings of depression.

In Table 5, looking horizontally at the influence of the selected determinants to predict the perceived efficacy of actions, we can see that, in line with what was reported with the analyses of variance (Table 3), the respondents from the country in the most difficult situation find all types of actions more effective, with the exception of civic participation and individual financial protection. Protectionism seems to be an effective behaviour especially when people begin to feel the impact of the crisis more strongly.

The socioeconomic situation of the countries is not the only predictor of the assessment of the efficacy of actions. Indeed, the way in which respondents perceive the crisis also contributes to this assessment. Thus, the degree to which people attribute responsibility to internal factors (namely State, past and current governments, employers and the mentality of the country) and to the people in general (public employees, extreme leftists and immigrants) are particularly good predictors of the assessment of the efficacy of actions: holding internal factors responsible for the crisis leads people to consider individual actions and pressure on government to be effective, whereas putting less blame on the people leads them to view collective actions, such as civic participation and political and economic resistance to government measures as effective. Moreover, perceiving a distant date for the crisis to come to an end predicts considering the use of violence, an act of desperation, more effective.

Finally, all three types of individual-related variables also contribute to the assessment of the efficacy of actions. Respondents’ political orientation influences the perceived efficacy of almost all types of actions: left-wingers are more likely to value collective pressures on government, either in the form of political or economic resistance to government measures or even in the form of violent action, whereas right-wingers are more likely to favour individual protectionist behaviours. The intensity of feelings of depression also contributes to the assessment of the efficacy of almost all actions, except protectionism and civic participation: the more desperate the respondents are, the more likely they are to find the actions effective. Lastly, violence, like individual financial protection, which includes emigrating, seems to be viewed as more effective by male respondents, more disposed to ponder the use of radical collective or individual behaviours.

In summary, the best predictors of the assessment of the efficacy of the actions against austerity measures are the gravity of the socioeconomic situation of the countries, the attribution of the responsibility for the crisis to internal factors and not to the people, the respondents’ political orientation and the feelings of depression related to the crisis.

Conclusion

To analyse the determinants of the assessment of the efficacy of actions is important if we want to understand better why people do, or do not, engage in specific forms of protest actions, because the perceived efficacy of actions at least partially explains people’s willingness to engage in those forms of actions (Hornsey et al., 2006; Van Stekelenburg & Kandermans, 2010). The variety of actions that may be taken against the austerity measures implemented in the present financial and economic crisis offered the opportunity to analyse the contribution of a set of factors for their perceived efficacy.

Our results have to do with types of actions that may be classified, in line with the categorization of Wright, Taylor, & Moghaddam (1990) into normative collective actions (civic participation and political resistance to government measures), non-normative collective actions (violence and economic resistance to government measures) and individual actions (individual financial protection and protectionism). They indicate that normative collective actions are considered more effective than non-normative collective actions, in line with Tausch et al. (2011), but that the efficacy of individual actions might depend on their beneficiary – when they contribute to the collective well-being, individual actions are viewed as more effective than collective actions.

Having verified that, in line with the literature (Van Stekelenburg & Klandermans, 2010), the perceived efficacy of actions was correlated with the willingness to engage in them, even if the respondents were more inclined to adopt individual actions than to participate in collective actions, we turned to an aspect that we thought was less explored by the literature: to understand why some actions are considered effective and others not.

Our results indicate that the socioeconomic situation of the countries influences the assessment of efficacy of actions that might be taken against austerity measures. Thus, the Greek respondents considered almost all actions more effective than the other respondents did (with the exception of civic participation), and the Italian respondents considered almost all actions to be less effective. The Portuguese respondents, who were in a situation better than the Greek but worse than the Italian respondents, agreed with the former about the efficacy of individual financial protection and with the latter about the (low) efficacy of collective actions. Because the assessment of efficacy is linked to the willingness to engage in actions, this result may help to explain the difference in the results obtained by Chryssochoou, Papastamou and Prodromitis (2013) and Poeschl, Valentim and Silva (2015), showing a greater unwillingness of Portuguese citizens to engage in actions in comparison with Greek citizens. More generally, this result also suggests that people plan to adopt individual strategies to resolve a situation before engaging in collective actions, and that the worse the situation, the more people view all types of actions as effective.

The analyses of regression indicate, however, that the socioeconomic situation of the countries is not the only factor that influences the assessment of the efficacy of actions. Respondents’ representations of the situation also contributed to this assessment. Indeed, when respondents held internal factors responsible for the crisis they considered individual actions as well as pressures on government to be effective, and when they refused to blame the people, they viewed particularly normative collective actions, such as civic participation and political resistance to government measures, as effective. Moreover, the perception that the crisis would last longer, suggesting greater sacrifices from citizens to address mainly private problems, predicted that violence, a desperate action, would be considered more effective.

Finally, individual-related characteristics also contributed to explain the assessment of the efficacy of actions. In particular, a deeper feeling of depression did not lead respondents to view all types of actions as ineffective, as hypothesized by Chryssochoou, Papastamou and Prodromitis (2013), who linked the feeling of depression to inaction. On the contrary, it led to the view that almost all types of actions were effective (except protectionism and civic participation, the most consensual actions), supporting the opinion that when people think they have nothing to lose they are more likely to consider that everything may be worth trying (Tausch et al., 2011). Thus, emotions do not only influence the decision to participate in actions, as defended by some authors (Van Stekelenburg & Klandermans, 2010), but also influence the perceived efficacy of actions that may be taken.

With regard to political orientation, left-wingers were more likely to value collective pressure on government, either in the form of political or economic resistance to government measures or even in the form of violent actions, as was predicted by the fact that the austerity measures were implemented by a right-wing government. In turn, right-wingers were more likely to favour individual protectionist behaviours, likely to support the implemented measures. Male respondents found violence a more effective action than female respondents did, a result that seems to be more in line with the research on sex differences, which shows that sex differences in aggression are more pronounced for physical aggression (Hyde, 1984), than with the use of a less aggressive and competitive negotiating style by women than by men (Dobrijevic, 2014).

In sum, our results reveal that the assessment of the efficacy of actions is modulated by a whole set of determinants, which may be uncovered by exploring different levels of analysis. Some limitations of our study must, however, be underlined. First, the beta values of some of the chosen predictors are small, due to the large number of cases in the analyses, and there are certainly other societal and individual specificities, other elements of the representations formed about the crisis, which have contributed to the assessment of the efficacy of the actions against the austerity measures. Moreover, using a sample of students in our analysis might have contributed to shaping our results. Finally, the relationships identified here might not be simply linear but more complex, as it is the case for the prediction of intentions to take action (see Mari et al., 2017, this issue).

Further research is thus needed to understand what kind of factors are likely to predict the perceived efficacy of protest actions, as well as to understand better people’s reactions to the financial and economic crisis.