It was in May of 2011 that the Portuguese minister of finances and public administration signed the Memorando de Entendimento sobre as Condicionantes de Política Económica (Memorandum of Agreement on Conditioning of Economic Politics) with the Troika, an entity composed of the European Commission, the European Central Bank and the International Monetary Fund. Its objective was to implement measures aimed at reducing the over-indebtedness of the public and private sectors, which stood at 90% and 260% of the gross domestic product (GDP) respectively, at the end of 2010 (Quelhas, Reference Quelhas and Paz Ferreira2013).

The need for austerity is patent in this text. Austerity is defined by Blyth (2013, p. 2), as “a form of voluntary deflation in which the economy adjusts through the reduction of wages, prices and public spending to restore competitiveness, which is (supposedly) best achieved by cutting the state’s budget, debts, and deficits” (see Quelhas, Reference Quelhas and Paz Ferreira2013). In fact, many of these measures were related to the reduction of State spending, such as reductions in social spending, in the costs of labor, in unemployment subsidies, in retirement pensions, reductions in the sectors of education, health, public transport, and reductions in investment spending, among others.

Among all the consequences that have resulted from these measures, there are two that should be pointed out. First, while most of the citizens and companies were negatively affected, bank shareholders did, in fact, benefit from these measures. The measures included the use of public funds for the recapitalization of over-indebted private banks, as well as for issuing bank bonds guaranteed by the State (Quelhas, Reference Quelhas and Paz Ferreira2013). The most vulnerable citizens, those who could not be blamed, were consequently forced to bear the consequences of past excesses, and the austerity measures, accentuating the social inequalities, made Portugal the most unequal country in the Euro Zone in 2011 (Cavero & Poinasamy, Reference Cavero and Poinasamy2013).

The level of inequalities may be illustrated by means of the data provided by two reports. The first report, the Crisis Monitoring Report from Caritas Europa (2014), revealed a global increase in unemployment, namely, in long-term unemployment and in unemployment among young people between 15 and 24 years old (30.1% in 2011; 37.7% in 2012, with an average value for Europe of 23% in 2012). It also revealed an increase in the risk of poverty, reaching 25.3% of the population in 2012, in spite of the threshold of poverty having been lowered from an annual income of 5,207 Euros in 2010 to an annual income of 4,994 Euros in 2012 because of the global fall in incomes. The second report, the World ultra-wealth report 2013, drawn up by the consulting firm Wealth-X (2013), revealed that, in that year, there were 870 millionaires in Portugal, i.e. people with fortunes over 25 million Euros. Thus, the number of millionaires increased 10.8% between 2012 (where there were 785 millionaires) and 2013, a higher increase than the European average. In addition, the amount of their assets increased 11.1%, passing from 67 to 75 billion Euros, which was also a larger increase than the European average.

Furthermore, the austerity measures did not succeed in reducing the debt and boosting the economy. For example, whereas in 2010 the public sector debt had reached 90% of the GDP, in 2011 it had almost reached 108.3%, at the end of 2013 it had reached 128.9% and at the end of the first semester of 2014, it had reached 134% (Banco de Portugal, 2014). These numbers show that by curbing economic growth, austerity measures prevented the reduction of the public debt, leading some authors to conclude that the harsher the austerity, the higher the increase in debt-to-GDP ratio (Cavero & Poinasamy, Reference Cavero and Poinasamy2013). Thus, the increase in inequality does not only occur within the countries, but also between the countries, and the breach between central and peripheral countries within the Euro Zone is becoming wider.

The importance of the consequences of the financial and economic crisis for most citizens explains the great number of studies dedicated to analyzing its different aspects, and research conducted in the social representations framework have already been published (e.g. Galli et al., Reference Galli, Bouriche, Fasanelli, Geka, Iacob and Iacob2010). In a broad sense, social representations are socially constructed and shared forms of knowledge, common sense knowledge, that contribute to interpret reality, justify and guide behaviors (Jodelet, Reference Jodelet and Jodelet1989). In the positional approach of Doise (Reference Doise1985), which aims at analyzing the causes, conditions and consequences of variations in the representations, the study of social representations may consist in showing the relationship between representations and social positions. Indeed, because shared social insertions imply particular interactions and experiences, specific representations are associated to specific social memberships, and different representations are linked to different social positions.

The studies undertaken from the beginning of the 1980s on social representations of economic phenomena have in fact revealed the relationship between the way these phenomena are learned and the social insertions, as well as the impact of the representations formed on the way people behave (Roland-Lévy, Reference Roland-Lévy2004). Among the numerous works conducted in this domain, it is worth mentioning those that focused, for example, on the representations of economic inequalities (Emler & Dickinson, Reference Emler and Dickinson1985), of money (Viaud, Reference Viaud1999) or of consumption (Viaud & Roland-Lévy, Reference Viaud and Roland-Lévy2000), and also on representations of related phenomena such as unemployment (Milland, Reference Milland2002), globalization (Poeschl & Viaud, Reference Poeschl and Viaud2008) or corruption (Poeschl & Ribeiro, Reference Poeschl and Ribeiro2010).

With regard to the financial and economic crisis, Galli et al.’s (Reference Galli, Bouriche, Fasanelli, Geka, Iacob and Iacob2010) study compared the discourses, gathered by means of a free association task, of small retailers, bankers and economics students from Italy, France, Greece and Romania. In spite of the economic and sociocultural differences between the countries, the results globally revealed very similar representations, structured around the word “unemployment”. In Greece, in addition to “unemployment”, the central elements of the representation – i.e. the elements that generate and organize social representations (Abric, Reference Abric and Jodelet1989, for example) – included the word “poverty”; in Italy, they included “poverty” as well as “money” and “work”. In Romania, the central elements included “money”, “bankrupt” and “economy”, and, in France, “banks”, “money”, “bankrupt”, “recession” and “purchasing power”. Economics students’ representations 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 (for example, “fear”) and economics (for example, “company”, “savings”).

In turn, a study conducted in Greece by means of a questionnaire composed of opinion scales examined the impact of several factors on the intention of the population to participate in actions against the austerity measures imposed by the successive governments (Chryssochoou, Papastamou, & Prodromitis, Reference Chryssochoou, Papastamou and Prodromitis2013). The authors underlined the strong opposition expressed to these measures, which took the form of collective as well as individual actions: from 2009 to 2012, Greece registered more than 3000 suicides, which corresponds to a 37% rise in the suicide rate in that country.

Observing that the economic crisis impacts large parts the population, albeit differently, the authors examined to what extent the social groups most affected by the crisis were also those who were the most willing to act and protest. Their results suggested that the people with the lowest family incomes were more willing to undertake traditional forms of political action (i.e. participating in political parties, labor unions, NGOs, etc.) and to look for financial security (i.e. taking one’s savings out of Greece to foreign banks or keeping money at home), but they were not more willing to take part in violent actions than others.

The present study aims at analyzing the representations of the financial and economic crisis among Portuguese students and at examining to what extent their financial situation produces variations in these representations. It is based on the assumption that the positions occupied by groups in society, defined by their economic capital, shape their objective living conditions and thus, their way of looking at the world, forming a lasting interpretative framework that partially influences the actions and judgments of their members (Bourdieu, Reference Bourdieu1979; Doise, Reference Doise1985). We considered that the variations in the representations formed by the groups would be particularly important given that austerity measures affect social groups differently. This has implications not only for the theories they develop about the causes or factors that are responsible for the crisis (Valentim, Poeschl, & Forte, in preparation), but also for their emotions, the feelings of vulnerability and deprivation that stem from the comparisons they make with other persons or situations and, probably, their readiness to act in response to the austerity measures.

The present study took place in December 2011, using a questionnaire based on the instrument developed by the above-mentioned Greek authors. In addition to the opinions scales contained in that instrument, our questionnaire included a free association task, a technique commonly used in studies on social representations that makes it possible to reveal the content of the lay theories constructed about socially relevant objects (see, for example, Poeschl & Ribeiro, Reference Poeschl and Ribeiro2010; Valentim, Reference Valentim2011).

Method

Participants

Four hundred and fifty-three Portuguese adults took part in this study, 252 from the city of Coimbra and 201 from the city of Porto. They were students from social and human sciences, consequently unequally distributed by sex: there were 58 male students and 395 female students. Respondents were aged between 17 and 60 years and were subsequently distributed into four age groups of unequal amplitude because of the strong positive asymmetry in the distribution of ages. These age groups were as follows: 17–18 years (n = 150), 19–20 years (n = 198), 21–25 years (n = 79), 26 years and over, group composed of clearly older individuals (n = 25), (one non-response). Politically, 105 respondents declared themselves to be left-wing, 78 right-wing, and 180 center-oriented; 83 respondents declared they had no political orientation and 7 did not answer this question.

With regard to their monthly family income, the distribution was the following: less than 485 Euros (minimum salary from January 1st, 2011 according to the data published by the Fundação Francisco Manuel dos Santos, 2011): 34 respondents; from 486 to 1000 Euros: 169; from 1001 to 1500 Euros: 95; from 1501 (amount fixed for the salary cuts introduced on January 1st, 2011) to 2000 Euros: 70; more than 2001 Euros: 68 respondents (17 students did not report their family income). It is noteworthy that the small subgroup of male respondents was mostly made up of students residing in the city of Porto (Porto: 39; Coimbra: 19). On average, these respondents declared higher family incomes than the female students and were much more numerous in the over 2001 Euros income group (34.48% of the male students; 12.15% of the female students).

Questionnaire and procedure

The questionnaire that was used began with a brief introduction stating that the survey was part of an international research project. Respondents were then asked to indicate five words or expressions that came to their mind spontaneously when they heard the expression “economic crisis” and to rate the degree of negativity/positivity of each of the mentioned responses associated to the crisis (1 = very negative; 5 = very positive).

The questionnaire included several other tasks, from which only those that are relevant for the present study will be described.

Vulnerability

The perception of vulnerability was assessed by means of nine items that were inspired by the work of Staerklé, Delaine, Gianettoni, and Roux (Reference Staerklé, Delay, Gianettoni and Roux2007). Participants had to indicate how likely (1 = very unlikely; 7 = very likely) they thought it was that in one or two years from that moment they would find themselves in any of nine situations of vulnerability (see Table 1).

Table 1. Likelihood of finding oneself in a situation of vulnerability. Saturations, means (1 = very unlikely; 7 = very likely) and Cronbach’s alphas

Note: The original items were reversed so that higher values correspond to a greater feeling of vulnerability.

Emotions

The emotions felt during the current crisis (1 = not at all; 7 = very much) were measured by means of fourteen emotions that were described by Jasper (1998; see Table 2).

Table 2. Emotions felt in relation to the crisis. Saturations, means (1 = not at all; 7 = very much) and Cronbach’s alphas

Relative deprivation

Relative deprivation (Runciman, Reference Runciman1966) was examined by means of the responses given by the respondents (1 = totally disagree; 7 = totally agree) to nine comparisons, some with other persons, others over time (see Table 3). All items were recoded so that the highest degree of agreement corresponded to a comparison which was detrimental to the respondents.

Table 3. Social comparison. Saturations, means (1 = totally agree; 7 = totally disagree) and Cronbach’s alphas

Note: The item “At my age my parents had had more success in life than me” was withdrawn as it did not saturate in any of the two factors.

Inclination to take action

Respondents’ inclination to participate in actions (1 = not at all; 7 = very much) in response to the austerity measures was assessed using the thirty two items proposed by Chryssochoou et al. (Reference Chryssochoou, Papastamou and Prodromitis2013).

Data analysis

Firstly, the reduction rules generally used in free association tasks (Rosenberg & Jones, Reference Rosenberg and Jones1972) were applied to the set of responses produced on the economic crisis. To describe the magnitude and the nature of the information gathered from the open items, three statistics were calculated (cf. Deconchy, Reference Deconchy1971): (1) fluidity (the total number of words), which reflects the ease with which people talk about an object; (2) amplitude (the number of different words), which reflects the number of elements that become accessible when people talk about an object; (3) richness (the ratio between amplitude and fluidity), which provides a measure of the integration of the information about the object.

Secondly, the different dimensions of the representations of the economic crisis, the way in which they are inter-related and the extent to which they are representative of the various groups of respondents were looked into. For this purpose, the Alceste program of textual data analysis was used. This program makes it possible to study the formal structure of co-occurrence of the words in a particular corpus, by executing a downward hierarchical classification using the chi-square distance from a numeric table registering the set of the lemmatized (reduced to the root) forms that constitute the discourses produced (see, for example, Reinert, Reference Reinert1993).

Then, a principal component factor analysis was performed on the responses to the items assessing the feeling of vulnerability, the emotions and the feeling of relative deprivation. Finally, a One-Way analysis of variance was executed to examine the effect of family income on the positivity of the words evoked, on the scales constructed from the factorial solutions, and on the inclination to take part in action against the austerity measures.

Results and discussion

Firstly, the results obtained from the free association task are presented, followed by the results of the analysis of the answers on the feelings aroused by the crisis in the students according to the different family income groups and, finally, the inclination to take part in actions against the austerity measures is analyzed.

Free associations with the crisis

Globally, 2339 words were registered, among which 175 different words with a frequency varying between 1 (34 words) and 197 (“poverty”) were found. Thus, it can be observed that the richness statistics, which measures the ratio number of different words/total number of words, was small (.07). This small value indicates that the discourse on the economic crisis is highly consensual. In addition to “poverty”, the other words registered with a high frequency of evocation were “unemployment” (178 occurrences), “money” (163 occurrences) and “difficulties” (106 occurrences). “Poverty”, “unemployment” and “money” were also mentioned with a high frequency by the respondents of Galli et al.’s (Reference Galli, Bouriche, Fasanelli, Geka, Iacob and Iacob2010) study, reinforcing their status as central elements of the social representations of the economic crisis.

The analysis of the corpus by means of the Alceste program of textual data analysis was performed on the 453 elementary units of context (EUC) that correspond to the groups of words provided by the 453 respondents. The program took into consideration all the words with a minimum frequency of four occurrences and thus classified 86.75% of the collected material into six classes of words. The relationship between the classes is illustrated in the dendrogram in Figure 1.

Figure 1. Relationship between classes of words and relative weight of the classes.

Class 1 – Misery

Class 1, one of the two biggest classes, was composed of 86 EUC, i.e. 21.88% of the elementary units of context, and was organized around “misery” (χ² = 60.72). It also included the words “hunger” (χ² > 30.00), “exclusion”, “sadness”, “solidarity” (χ² > 20.00), “dissatisfaction”, “anguish”, “unemployment”, “school dropouts”, “decadence”, “impotence” (χ² > 10.00). This class grouped words that were mainly evoked by the students from Coimbra (χ² = 5.37) and whose family income was under 485 Euros (χ² = 5.34). It was also evoked by students who stated they were politically center-oriented (χ² = 3.76).

Class 2 – Difficulties

Class 2 was composed of 72 EUC, i.e. 18.32% of the units of context, and was structured around “difficulties” (χ² = 89.66). It also comprised the words “savings” (χ² > 50.00), “stress” (χ² > 20.00), (lack of) “money”, “isolation”, “politics”, “despair” and “recession” (χ² > 10.00). The words tended to be more frequently evoked by the students with monthly family income between 1001 and 1500 Euros (χ² = 2.35) and with no political orientation (χ² = 2.20). This class differs from Class 1 by the clearly less dramatic connotation of the words mentioned.

Class 3 – Quality of life

Closely associated to Class 2, the small Class 3 includes 46 EUC, which corresponds to 11.70% of the global corpus. Around “quality of life” (χ² = 50.00), the words “unemployment”, “bankruptcy”, “precariousness” (χ² > 30.00), “crisis”, “State” (χ² > 20.00), and “austerity” (χ² > 10.00) were found. This class was mainly evoked by students aged over 26 years (χ² = 2.40).

Class 4 – Change

Class 4, with 25 EUC, i.e. 6.36% of the units of context, was the smallest of all extracted classes. It was strongly structured around “change” (χ² = 198.00), to which the words “contestation”, “effort” (χ² > 70.00), “opportunities” (χ² > 30.00) and “collapse” (χ² > 20.00) were associated. This class, which seems to warn of the need to find solutions to overcome the crisis, was mainly evoked by the students who had a family income superior to 2001 Euros (χ² = 5.53), aged between 21 and 25 years (χ² = 2.14). It revealed that a reference to opportunities is above all possible for persons who do not have to find ways to survive first.

These four classes reveal variations in the representations of the economic crisis that stem from the different categories of family income, in other words and according to the positional perspective, variations that reflect the social positions and specific experiences of the respondents. In addition to these four classes, there are two classes that described the economic and geopolitical situation of Portugal.

Class 5 – Debts

Class 5, the second biggest class, included 85 elementary units of context, which corresponds to 21.63% of the global corpus. This class was organized around “debts” (χ² = 68.80) and also comprised the words “mismanagement” (χ² > 60.00), “inflation”, “deficit”, “purchasing power”, “irresponsibility” (χ² > 30.00), “incompetence”, “corruption” (χ² > 20.00). These words were evoked by the whole set of students and point to the state of the country and some of the factors that led to the situation of crisis now faced.

Class 6 – Portugal

Class 6, with 79 EUC, included 20.10% of the global corpus. It was organized around the term “Portugal” (χ² = 73.60), to which the words “Troika” (χ² > 70.00), “banks” (χ² > 60.00), “government” (χ² > 50.00), “Greece” (χ² > 50.00), “European Union”, “Angela Merkel”, “Euro”, “economy” (χ² > 30.00), “José Sócrates”, “IFM”, “scholarship” (χ² > 20.00), “capitalism”, “finances” (χ² > 10.00) were associated. These responses point to the actors who intervene in the economic crisis and were particularly evoked by the male students (χ² = 12.92), aged over 26 years (χ² = 4.91), from the city of Porto (χ² = 4.80), with left-wing political orientation (χ² = 4.01). Contrary to the words grouped in Class 5, which give an insight into the country, the words contained in Class 6 situate the country within a network of influences that goes beyond the borders of Portugal.

Thus, it can be observed that the first four classes, which mainly associated the consequences of the crisis to individual ways of living, are linked to two big classes that focus not on the persons but on the country. The first of these two classes refers to the debt, which results from the mismanagement of the country, due to incompetence or corruption. It is noteworthy that no allusion to the moralistic interpretation of the crisis as “living beyond our means”, which marked the discourses of some politicians (see, for example, the Jornal de Notícias, of May 6, 2011) was found. In the second class, the situation of Portugal is viewed in the light of its political and economic partners. This last class is the only one that revealed a politically involved discourse, which does not only point to a problem arising from management by the national leaders but looks at the situation from a broader perspective.

As could be expected, the rating of the responses was very negative: 1.68 on a scale in which 1 = very negative and 5 = very positive. In spite of the differences in the words associated to the crisis, the family income groups did not differ in the assessment of their valence, F(4, 431) = .87, ns (see also Table 4).

Table 4. Positivity attributed to the words evoked, vulnerability, emotions, and relative deprivation. Means according to the five positions of family income and analysis of variance

Note: In each line, different letters indicate significantly different means according to the LSD test for multiple comparisons, p < .05.

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

In order to better understand how the students from the different income levels experienced the situation of crisis, their perception of vulnerability, their emotions and the feeling of deprivation that stemmed from the comparisons made with other persons or over time was analyzed.

Vulnerability

The principal component factor analysis with Varimax rotation performed on the responses to the nine items measuring the probability that the respondents may find themselves in one or two years from that moment in situations of vulnerability, extracted two factors with eigenvalue greater than one, which, together, explained 55.98% of the total variance. The first factor, which explained 35.73% of the variance, reflected financial vulnerability, and the second, which explained 20.25% of the variance, pointed to vulnerability felt in the domain of personal fulfillment (see Table 1).

It should be noted that the students considered it more likely to find themselves in a situation of financial vulnerability than in a situation of vulnerability regarding personal fulfillment, t(452) = 14.78, p < .001. The feeling of vulnerability concerning personal fulfillment was not very strong, which seems to highlight the opinion that the crisis mainly affects the economic dimension of the students’ life.

Emotions

The principal component factor analysis with Varimax rotation performed on the ratings of the fourteen emotions felt by the respondents during the crisis extracted two factors with eigenvalue greater than one, which, together, explained 50.43% of the total variance. As may be seen in Table 2, the first factor, which explained 33.51% of the variance, included the negative emotions whereas the second factor, which explained 16.92% of the variance, grouped the positive emotions.

As could be expected, the students felt more negative emotions than positive emotions, t(452) = 5.47, p < .001. Among the negative emotions, indignation predominated, whereas culpability was viewed as irrelevant and, among the positive emotions, solidarity was more experienced than trust.

Social comparison and feeling of relative deprivation

All items concerning the comparisons that the respondents were asked to make were recoded so that the larger value of the scale would correspond to the highest degree of disagreement with the positive formulation of the items or, in other words, so that the highest degree of agreement would correspond to a comparison detrimental to the respondent, likely to originate a feeling of relative deprivation.

The principal component factor analysis with Varimax rotation performed on the nine comparisons also extracted two factors with eigenvalue greater than one, which together explained 61.23% of the total variance. The first factor, which explained 36.80% of the variance, encompassed comparisons with others, and the second, which explained 24.44% of the variance, encompassed comparisons over time (see Table 3).

The result from the comparisons likely to generate a feeling of relative deprivation was more negative when respondents compared themselves with other people than when they compared situations over time, t(452) = 15.09, p < .001. The students disagreed in particular with the idea that the Portuguese are in a better position than other Europeans. However, it may be observed that the lowest rating (3.79) failed to reveal disagreement with the expectation of being in a better social position in the future, which, in spite of everything, may reflect some hope for a possible improvement of the situation of the country and of Portuguese citizens.

Effect of family income

In order to uncover the extent to which different levels of family income affect the perception of the probability of finding oneself in situations of vulnerability, the emotions felt, and the feeling of relative deprivation that stems from comparisons with others and over time, a One-Way analysis of variance was performed on the scales that aggregated the items grouped on the extracted factors. The means calculated according to the five positions of the monthly gross family income are registered in Table 4.

Table 4 reveals a very consistent pattern of results. A lower income led to the perception of a greater probability of finding oneself in situations of financial vulnerability, F(4, 431) = 5.06, p = .001, η² = .05, and of vulnerability in personal fulfillment, F(4, 431) = 5.49, p < .001, η² = .05, to more negative emotions, F(4, 431) = 6.23, p < .001, η² = .06, and less positive emotions, F(4, 431) = 3.61, p = .007, η² = .03, as well as more unfavorable comparisons with others, F(4, 431) = 6.65, p < .001, η² = .06. In other words, as family income increased, pessimism decreased, with the sole exception of the comparisons over time that, in addition to revealing some cautiousness with regard to making predictions about the future, did not differ according to the income groups, F(4, 431) = 1.65, ns.

As could be expected, the variations in the representations of the economic crisis were associated to the specific experiences of the respondents that result from the differences in family income. It remains to examine whether the differences in family income and their subsequent differences in the representation of the crisis gave rise to differences in inclination to act in response to the austerity measures taken by the government.

Inclination to take action

The principal component factor analysis on the answers concerning the readiness to take action revealed that students envisaged three types of possible participation (Valentim, Poeschl, & Forte, in preparation): civic participation (participating in labor unions, in political parties, in NGOs, signing petitions, etc.), participation in violent actions (burning politicians’ cars, physically assaulting politicians, destroying public property, etc.) and economic resistance (refusing to pay any new taxes, taking one’s savings out of Portugal, keeping money at home, emigrating, etc.). It also revealed that, globally, the students were rather unwilling to participate in any actions in the context of the economic crisis. The single action with which they agreed was emigrating, and participating in violent actions seemed out of the question.

In order to understand to what extent different levels of family income, which significantly affect how students experience the crisis, shaped their inclination to take action against the austerity measures, a One-Way analysis of variance was performed on the scales that aggregated the items grouped on the factors pointing to the three possible types of actions (civic participation, violent participation and economic resistance). The means calculated according to the five positions of the monthly gross family income are shown in Table 5.

Table 5. Inclination to take action. Means according to the five positions of family income (1 = not at all; 7 = very much) and analysis of variance

Table 5 shows that the vulnerability felt and the negative comparisons were not accompanied by any behavioral intention, not even by any inclination to participate in civic actions. In other words, in spite of the differences in the way they are affected by the crisis, the groups with different family incomes did not vary in their (lack of) readiness to act in response to the austerity measures taken by the government, a result which contrasts with the conclusions of the study conducted in Greece by Chryssochoou et al. (Reference Chryssochoou, Papastamou and Prodromitis2013).