Journal of Economic Psychology 19 (1998) 775±796
Attitudes towards European Monetary Union in Greece: Antecedents, strength and consequences Flora Kokkinaki
1
London Business School, Sussex Place, Regent's Park, London NW1 4SA, UK Received 20 May 1998; received in revised form 24 July 1998; accepted 25 July 1998
Abstract The objective of the present study was twofold: First, to examine Greek people's attitudes towards European Monetary Union (EMU) and to explore the antecedents, strength and consequences of these attitudes on intentions to support or oppose the single European currency. Second, and from a more theoretical perspective, to investigate the structure of attitude strength and the eects of strength dimensions on attitude±intention consistency. A total of 644 Greek citizens participated in the study. The ®ndings indicate that Greeks are fairly supportive of the monetary union and are quite optimistic about its consequences. The anticipated outcomes of EMU for Greece and for Europe were found to be signi®cant determinants of attitudes. In addition, attitudes were determined by the strength of national identity and by the more general attitudes towards the membership of Greece in the European Union. On the whole, attitudes were fairly strong and had a substantial relation with intentions. Attitude strength was found to consist of three dimensions: Embeddedness, Conviction and Internal Consistency. Only Conviction and Internal Consistency were found to have a signi®cant moderating eect on the attitude±intention relation. Further, the moderating eect of these factors was interactive, in that either high Conviction or high Internal Consistency suced to make an attitude in¯uential. Ó 1998 Elsevier Science B.V. All rights reserved. Keywords: European Monetary Union; Attitudes; Attitude strength
1
Tel.: +44 171 262 5050; fax: +44 171 724 1145; e-mail:
[email protected].
0167-4870/98/$ ± see front matter Ó 1998 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 7 0 ( 9 8 ) 0 0 0 3 7 - 3
776
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
1. Introduction Europe now stands at a historic threshold as the process of European integration has moved beyond the goals of the Single European Act towards the goals of political and economic integration enunciated in the Maastricht Treaty. Public opinion is bound to play an increasingly important role in the integration process (Rattinger, 1994). In certain cases (e.g. the Danish referenda), public opinion has been taken into account in the decisions to participate in the new scheme. In most countries, however, public opinion has played a rather peripheral role and most decisions have been made by central governments. Nonetheless, understanding public opinion towards issues such as the European Monetary Union (EMU) is an essential step in the implementation of the Maastricht reforms. Politicians and policy makers need to understand the discrepant views that people within and across countries might hold, in order to avoid the social con¯icts that the integration process might engender. 1.1. Greek people's attitudes towards EMU Greece is a southern European country and, in terms of population and geographical size, is among the smaller countries in the European Union (EU). Its economic position among EU member states is also rather low. The issue of EMU has been at the centre of public attention in Greece after the Maastricht Treaty. The Greek government launched a strict austerity programme in its 1997 budget, in order to meet the economic convergence criteria that would enable Greece to join the single currency. 2 During this year, signi®cant changes occurred in the Greek economy, such as increasing growth rates and decreasing in¯ation rates and budget de®cits (Financial Times, 25 November 1997). These positive changes, however, implicated short-term economic costs for Greek citizens (e.g. freeze on wages). Despite protests against the government's ®scal policy, mainly from public sector unions, and although public opinion has not been overtly expressed in a referendum, it has been assumed that Greek citizens provide a ``permissive consensus'' that enables the government to pursue the Maastricht reforms. Most public opinion surveys (e.g. Eleftherotypia, 8 June 1997) do not provide a sucient, in-depth account of people's attitudes towards the single 2
The present study was conducted between June and August 1997.
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
777
European currency and do not examine what factors underlie and determine them (see MuÈller-Peters et al., this issue). The ®rst objective of the present study is to explore Greek people's attitudes towards EMU and to investigate their antecedents. The study further examines the strength of these attitudes and their consequences on intentions to support or oppose the single currency. More speci®cally, the study investigates the network of attitudinal beliefs related to the anticipated economic and political consequences of EMU: It is examined whether people expect the introduction of the single currency to have consequences on their personal income, the value of their savings and their job security, on government debt, the level of unemployment, taxation, the rate of in¯ation and the prices of goods and whether people will lose money by the very conversion to the euro. The potential consequences of EMU on the cooperation and trade between European countries, on the ease of price comparisons across countries and on the competitiveness of the euro, in relation to the dollar and the yen, are also examined. These anticipated outcomes of EMU have been included in the standard questionnaire used across countries, although, in certain cases, they are not treated as attitudinal beliefs, but rather as indicators of other constructs (see MuÈller-Peters et al., this issue). In addition, a number of country speci®c outcomes have been included in the Greek questionnaire: It is examined whether the EMU will have any consequences on the political independence and security of Greece and on its relation with Turkey and the Balkan countries. A number of other constructs, and speci®cally, national identity, European identity, life satisfaction, attitude towards the participation of Greece in the EU and perceived fairness of the membership of Greece in the EU, are also examined as potential determinants of attitudes (see MuÈller-Peters et al., this issue). Understanding attitudes and their determinants is important to the extent that these attitudes in¯uence people's decisions and actions. Accumulated empirical evidence indicates that whereas some attitudes are consequential, others have few, if any, eects on decisions and behaviour (see Kraus, 1995). Attitude strength is thought to underlie and moderate this relationship between attitudes and behaviour: 3 The stronger an attitude, the more likely it is to in¯uence behaviour (Krosnick and Petty, 1995). Attitude strength is therefore an important construct and it is essential that measures of attitude
3 Attitude strength is also thought to determine the temporal stability and resistance of attitudes to change and to in¯uence the processing of attitudinal information (Krosnick and Petty, 1995).
778
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
strength are included in attitude research. In the case of public opinion surveys, attitude strength can provide a valid indicator of the power of attitudes to aect individual's subsequent intentions and actions. Following this rationale, and as will be described in the following section, several variables are included in the present study as indicators of the strength of people's attitudes towards the single European currency. 1.2. The structure and moderating role of attitude strength The concept of attitudes provides an appropriate platform for the study of public opinion since it encompasses the range of beliefs, feelings and behaviours that are associated with the issue in question. The utility of attitudes and attitude theories is evidenced by their extensive application in the study of social, political and economic behaviour. Typically the ®ndings of such applications are not used as input for theory evaluation and development. Nonetheless, natural settings are the domain to which attitude researchers wish to generalise and evidence from such settings can be used to reveal the theories' strengths and limitations (Eagly and Chaiken, 1993). The second objective of the study is to use the empirical data to address a more theoretical issue. Speci®cally, the study explores the structure of attitude strength and its moderating role in the attitude±intention relation. Empirical investigations of attitude strength have employed several attitudinal qualities as indicators of this latent construct. For example, variables such as the importance of an attitude (Krosnick, 1988), attitude certainty (Fazio and Zanna, 1978) and structural consistency (Norman, 1975) have been found to moderate the behavioural eects of attitudes (see Kraus, 1995). Although strength indicators are conceptually and operationally distinct, they are interrelated. The nature of these interrelations is an issue that has only recently started to receive attention. Findings from con®rmatory factor analyses indicate that attitude strength is a multi-dimensional construct (e.g. Krosnick et al., 1993). However, research on its exact structure is less conclusive as ®ndings are inconsistent across studies. For example, Krosnick et al. (1993) failed to ®nd a clear pattern in the relations amongst 10 strength indicators and conclude that to view attitude strength as a single latent construct re¯ected by these indicators is inappropriate. However, Pomerantz et al. (1995) found a two-factor structure among the seven strength indicators they used. The ®rst factor, which the authors call Embeddedness, re¯ects the centrality and importance of the attitudinal object for the individual, while the second factor, Commitment, is de®ned by attitude extremity and cer-
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
779
tainty. More recently, Prislin (1996) used 13 strength indicators to predict attitude stability. These variables clustered into three dimensions: the General Strength Index, de®ned by variables such as certainty, importance, frequency of thought and knowledge; Internal Consistency, de®ned by evaluative±cognitive and evaluative±aective consistency; and Extremity, de®ned by evaluative and aective extremity. Prislin also found that stable attitudes are not necessarily high on all three dimensions and that one dimension is sucient to make an attitude stable. Unlike previous research, the present study presents an opportunity to explore the structure and consequences of attitude strength in a naturalistic context involving a non-student sample and an issue of great social relevance. The study examines the interrelations among 12 strength indicators commonly used in attitude research: attitude certainty, personal involvement, vested-interest, frequency of thinking about the object, frequency of talking about the object, attitude importance, subjective knowledge about the attitude object, attention to information related to the attitude object, attitude ambivalence, attitude extremity, evaluative±cognitive and evaluative±aective consistency (for a review of strength related variables, see Kraus, 1995; Petty and Krosnick, 1995; Prislin, 1996; Raden, 1985). The study further investigates the consequences of attitude strength and of strength dimensions on individuals' propensity to act on their attitudes. Although, this part of the research is also exploratory, consistent with previous ®ndings, it is expected that attitude strength is a multi-dimensional construct. Previous research has not examined the consequences of strength dimensions on attitude±behaviour consistency. However, on the basis of Prislin's ®ndings (Prislin, 1996), it is expected that attitudes need not necessarily be high in all strength dimensions in order to have an impact on behavioural intentions. 2. Method Participants: A snowball sampling procedure was followed in recruiting respondents for the study. Speci®cally, a number of questionnaires were distributed, either personally or by mail, to an initial sample of individuals who expressed interest in the study. It was ensured at this stage that people from dierent age groups and geographical regions were targeted, as it was thought that in this manner a fairly representative sample would emerge. This initial sample of individuals were asked to complete the questionnaire and to distribute a number of additional questionnaires to other people who they
780
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
thought might be interested in the study. In every case, questionnaires were accompanied by a pre-paid, self-addressed envelope, to ensure anonymity of responses. A total of 1110 questionnaires were distributed, 695 of which were returned. Fifty-one of these were excluded from the analysis due to the large number of missing values. The ®nal sample consisted of 644 individuals (328 male, mean age 35.49). Table 1 presents a description of the sample. As can be seen in the table, the sample was fairly representative of the Greek population in terms of gender and area. However, high educational levels and ages between 20 and 40 years were overrepresented. Questionnaire design and measures: The cover letter and the ®rst page of the questionnaire explained the purpose of the study and provided instructions on the use of 5-point scales, as most questionnaire items employed such a format. In most cases, only the endpoints of each scale were presented in the questionnaire, with ®ve unlabelled boxes between them. Respondents were asked to tick the box that represented their opinion or feelings. Most items included a separate ``don't know'' option. The items used for the assessment of attitudes towards EMU, attitudes towards the participation of Greece in the EU, behavioural intentions to support or oppose the introduction of the euro in Greece, outcome beliefs, national and European identity and perceived fairness of the Greek membership in the EU, are presented in Appendix A. All measures were coded on a 1±5 scale and in a way such that that higher scores represent more positive attitudes, intentions and beliefs and stronger national and European identity. Also, in the case of perceived fairness, higher scores indicate that respondents think that Greece receives more than it pays to the EU. ``Don't know'' responses were treated as missing values. Most indicators of the strength of attitudes towards EMU (i.e. attitude certainty, personal involvement, attitude importance, vested interest, frequency of thinking, frequency of talking, subjective knowledge, attention to information and attitude ambivalence) were directly assessed in the questionnaire (see Appendix A). Attitude scores were recoded to provide a measure of attitude extremity. The most extreme scores, on both sides of the scale, were recoded as 3, moderate scores were recoded as 2 and scores around the median were recoded as 1. Evaluative±cognitive consistency and evaluative±aective consistency were operationalised as the absolute dierence between the evaluative (good/bad, very much for/very much against) and the cognitive (reasonable/unreasonable, insecure/secure, controllable/ uncontrollable) and aective (calm/worried, tense/relaxed) items used in the computation of the attitude index. All strength indicators were scored so that
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796 Table 1 Description of the sample and population statistics
a
Variable
Frequency
Valid percent
Population percent
Gender Male Female Missing
328 299 9
52.3 47.7
49.2 50.7
Age 16±19 years 20±29 years 30±39 years 40±49 years 50±59 years 60±69 years 70±79 years Missing
33 199 200 106 51 25 15 15
5.2 31.6 31.8 16.9 8.1 4.0 2.4
7.4 14.7 13.7 12.0 12.7 10.7 6.2
Area Rural areas or village Small or middle sized town Large town or city Missing
157 84 357 46
26.3 14.0 59.7
28.3 12.7 58.8
129 32 292 45 31 71 32
20.4 5.1 46.2 7.1 4.9 11.2 5.1
Occupational status Self-employed Retired Employed Housewife Unemployed Student None of the above (e.g. in military service) Missing
12
Educational level b Primary school Lower secondary school Higher secondary school Further, non-University education University education Missing
41 45 193 110 238 17
6.5 7.2 30.8 17.5 38.0
Personal income Less than 185 000 drachmas 185 000±275 000 drachmas 275 000±650 000 drachmas More than 650 000 drachmas Missing
216 207 144 23 54
36.6 35.1 24.4 3.9
39.6 10.7 20.6 0.6 10.7
781
782
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Table 1 (Continued) Variable
Frequency
Valid percent
Household income Less than 200 000 drachmas 200 000±300 000 drachmas 300 000±650 000 drachmas More than 650 000 drachmas Missing
74 174 259 97 40
12.3 28.8 42.9 16.1
Population percent
a
Population statistics are based on the 1991 census. Statistics on personal income are not available. Statistics on occupational status and household income are not applicable as they are based on dierent classi®cation systems. However, it should be noted that according to the declared income statistics of the ®scal year 1996, 47.5% of the tax payers earn less than 2 000 000 Greek drachmas per annum, 17.5 earn less than 3 000 000, 23.2% earn less than 5 500 000 and 11.9% earn more than 5 500 000. It should also be noted that the Greek drachma-ECU rate at the time was approximately 309 drachmas 1 ECU. b The population distribution of education level is based on people aged 10 years and above and includes illiterate people and children still in primary education, not represented in the table. It should be noted that University level education frequencies include people with post-graduate degrees, degrees from polytechnics and current University/Polytechnic students.
higher numbers represent greater strength, with the exception of the consistency measures, where higher scores indicate larger dierences between the components of the attitude and therefore less internally consistent attitudes. 3. Results 3.1. Greek people's attitudes towards EMU Mean attitude towards the single currency was relatively positive 3.29 (SD 1.03), indicating that, on the whole, Greeks are in favour of EMU. The majority of respondents (57.1%) considered EMU to be a ``good'' thing and 55.1% were ``for'' a single European currency. Only 14.5% thought that EMU is a ``bad'' thing and 20.8% were ``against'' a single currency. Similarly, intentions to vote for a single currency were positive (mean 3.67, SD 1.40). 53.1% of the respondents considered it likely that they would vote for a single currency versus 21.2% who considered it unlikely. Attitudes were strongly related to intentions (r .63, p < .001). The results indicate that Greeks are optimistic about the consequences of EMU. For 10 of the 18 listed consequences respondents were rather on the positive side, in terms of the changes they expected the single currency to
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
783
Table 2 Attitudinal beliefs: Means, standard deviations, correlations with attitude and factor loadings Beliefs Income Debt Savings Job security Unemployment In¯ation Tax Price of goods Personal money loss Cooperation Competitiveness Trade Price comparisons Feel more European Political independence Political security Relation with Balkan countries Relation with Turkey % of explained variance
Mean
SD
r
Factor 1
2.77 2.83 2.55 3.21 2.83 3.15 2.60 2.86 3.25 3.73 3.88 3.90 3.91 2.90 2.76 3.44 3.63
1.16 1.08 1.14 1.21 1.10 1.09 1.15 1.21 1.08 0.95 0.95 0.94 0.88 1.19 1.08 1.03 1.12
.38 .37 .31 .41 .35 .48 .31 .28 .18 .40 .46 .41 .31 .39 .32 .26 .39
.14
.17 .10 .25 .12 .80 .78 .76 .74 .62 .15 .21 .32
3.21
1.23
.30
.24 30.70
.82 .79 .73 .72 .71 .71 .70 .61 .59 .23 .23 .11 .29 .28
Factor 2
Factor 3
.23
.29
.24 .19
.33 .32
.21 ).18 .12 .32
Factor 4 .35 ).11 .33 .43 .11 .19 .23 .23 .25
.29 .85 .79
.19
.24
.12
.82
15.00
9.00
7.40
.10 .85
p < .05.
bring about (see Table 2). More speci®cally, respondents were slightly pessimistic about the consequences of EMU on their personal income, the value of their savings and the prices of goods. Also, they did not expect the introduction of the single currency to improve government debt, taxation and unemployment level. However, they were optimistic about the consequences of EMU on job security and in¯ation rate and they did not expect to lose money by the introduction of the euro per se. Respondents were also optimistic about the consequences of EMU on the cooperation and trade between countries, the facilitation of price comparisons across Europe and the strengthening of European identity. Although they expected EMU to decrease the political independence of Greece, they expected it to increase political security and to improve the relation with Turkey and the Balkan countries. Outcome beliefs were moderately related to attitudes in most cases. As can be seen in Table 2, attitude±belief correlations ranged from .18 to .46. Correlations between attitudinal beliefs ranged from .01 to .70, although most
784
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Fig. 1. A conceptual model of the relation between attitudes, attitude antecedents and intentions.
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
785
beliefs were moderately to strongly intercorrelated. Attitudinal beliefs were submitted to a principal component analysis with an oblique rotation. An oblique, instead of a varimax rotation was chosen because of the high item intercorrelations. The rotated factor solution yielded four weakly intercorrelated factors (all factor correlations were below .18) which accounted for 62.1% of the total variance. The obtained factor structure is easily interpretable: The ®rst factor represents the economic consequences of EMU for the individual and the country; the second factor re¯ects the consequences for Europe as a whole; the third factor represents the political consequences for Greece; and ®nally, the fourth factor represents, more speci®cally, the political consequences for the relation between Greece and its neighbour countries (see Fig. 1). On the basis of factor loadings, indices of each outcome factor were computed as the mean of responses to the corresponding items. To examine the value of these outcome factors as predictors of attitudes, attitudes were regressed on these indices. As can be seen in Table 3, all outcome factors were signi®cant predictors of attitudes. As can be seen in Fig. 1, in addition to outcome beliefs, a number of other constructs were used as predictors of attitudes. Speci®cally, national identity, European identity, attitude towards the participation of Greece in the EU, life satisfaction and perceived fairness of the participation of Greece in the EU were also entered into the equation. The inclusion of these variables, rendered insigni®cant the contribution of the outcome factors related to the political consequences of EMU and to the consequences on the relation with neighbour countries (Table 4, see also Fig. 1). National identity and attitudes towards the membership of Greece in the EU were found to be signi®cant
Table 3 Regression of attitudes on belief factors Variables
B
b
t
Economic consequences for Greece and the individual Consequences for Europe Political consequences for Greece Consequences for the relation with neighbour countries
.37
.27
6.86
.45 .10 .14
.34 .10 .15
7.98 2.54 3.08
p < .01;
p < .001.
R
R2
F
.63
.40
73.60
786
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Table 4 Regression of attitudes on belief factors and other variables Variables
B
Economic consequences for Greece and the individual Consequences for Europe Political consequences for Greece Consequences for the relation with neighbour countries Perceived fairness of the participation of Greece in the EU European identity National identity Attitude towards the membership of Greece in the EU Life satisfaction
p < .01;
b
t
.32
.25
5.73
.23 .06 .07
.18 .06 .08
3.62 1.44 1.80
.02
.03
0.70
.06 ).12 .28
.07 ).14 .31
1.69 )3.46 6.28
.04
.02
0.68
R
R2
F
.69
.48
35. 94
p < .001.
predictors of attitudes. However, the strength of identi®cation with Europe and its nations, the perceived fairness of the participation of Greece in the EU and life satisfaction did not contribute signi®cantly to the prediction of attitudes. In order to determine whether attitudes are signi®cant predictors of intentions to support or oppose EMU, intentions were regressed on attitudes. This analysis revealed that attitudes were strong predictors of intentions (b .63, t 19.39, p < .001). 3.2. The structure and moderating role of attitude strength The strong relationship between attitudes and intentions can be explained in terms of attitude strength. As can be seen in Table 5, with the exception of frequency of thinking and talking about the issue, subjective knowledge and evaluative±aective consistency, scores on all other strength indicators were quite high, suggesting that attitudes are in¯uential. Speci®cally, respondents perceived the issue as personally involving and their attitudes as important, they paid attention to relevant information and they were quite certain and non-ambivalent about their attitudes. They also thought they had vested interest in the issue. In order to examine the structure of attitude strength, a principal component analysis with an oblique rotation of the strength indicators was
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
787
Table 5 Factor loadings of strength indicators Variables
Mean
SD
Frequency of thinking Personal involvement Frequency of talking Amount of attention Attitude importance Vested interest Certainty Subjective knowledge Ambivalence Extremity Evaluative±aective consistency Evaluative±cognitive consistency
2.39 3.41 2.08 3.11 3.08 3.24 3.45 2.45 3.24 2.23 1.02 0.61
1.43 1.20 1.29 1.36 1.34 1.34 1.32 1.10 1.20 0.77 0.91 0.62
% of explained variance
Factor 1 .77 .73 .71 .69 .65 .63 .23 .30 ).10 31.70
Factor 2
Factor 3
).14
).15 .10 ).11 .14
).16 ).28 .14 .10 ).73 ).62 .55 ).40
12.40
.17 .16 ).36 .71 .70 8.40
conducted. An oblique rotation was again preferred over a varimax because most strength factors were intercorrelated (correlations ranged from ).01 to .66). This analysis revealed a fairly clear three component solution (all component correlations below .18). The three extracted factors explained 52.6% of the total variance. Table 5 presents factor loadings for the three factor solution. Frequency of thinking, personal involvement, frequency of talking, attention to information, attitude importance, and vested interest de®ned the ®rst factor which accounted for 31.7% of the total variance. This factor accounted for more than half of the total explained variance and was de®ned by six out of the 12 strength indicators. It therefore seems to be the most signi®cant dimension of attitude strength. This factor is related to Prislin's Generalised Attitude Strength (Prislin, 1996), not only because it is the strongest dimension of the construct but also because it is de®ned by similar indicators. However, at a conceptual level, the factor resembles more closely to Pomerantz et al. (1995) Embeddedness, as it seems to re¯ect the underlying importance and centrality of the issue for the individual. The term Embeddedness will be used therefore to describe it. Attitude certainty, subjective knowledge, ambivalence and extremity de®ned the second factor. It should be noted that subjective knowledge also contributed to the ®rst factor, a ®nding which indicates that the perceived personal importance of an issue is also related to how knowledgeable one feels about this issue. However, it is fairly clear that this factor represents the degree to which individuals are clear and certain about their evaluation and therefore the
788
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
term Conviction will be used to describe it (Abelson, 1988). It should be noted that this factor, although not conceptually identical, is related to Prislin's Extremity dimension and Commitment of Pomerantz et al. Consistent with Prislin's ®ndings, the third factor was de®ned by evaluative± aective and evaluative±cognitive consistency and will therefore be termed Internal Consistency. Attitude extremity also had a high loading on this factor. In order to examine the eect of strength factors on the consistency between attitudes and intentions, an index of each dimension was computed as the sum of scores on the main indicators and moderated regression analyses were performed. Speci®cally, Embeddedness was computed as the sum of frequency of thinking, personal involvement, frequency of talking, attention to information, importance, and vested interest. Conviction was operationalised as the sum of certainty, knowledge, ambivalence and extremity and Internal Consistency was operationalised as the sum of evaluative±aective and evaluative±aective consistency. Behavioural intention was regressed on the interaction of each strength factor with the attitude score. Table 6 presents the results of these analyses. Only Conviction and Internal Consistency had a signi®cant moderating eect on the attitude±intention relation. Attitudes of high Conviction were more likely to exert a strong in¯uence on intentions than attitudes held with low Conviction. Similarly, internally consistent attitudes were more likely to in¯uence intentions than less internally consistent attitudes. Because of the scoring of evaluative±aective and evaluative±cognitive consistency the negative sign of the b coecient indicates that the smaller the distance between these components of the attitudes, the more likely the attitude is to be in¯uential. These ®ndings support Prislin's suggestion that one dimension of attitude strength suces for an attitude to be strong.
Table 6 Coecients of the strength dimension ´ attitude terms Variables
R2
b
t
Attitude ´ Embeddedness Attitude ´ Conviction Attitude ´ Internal Consistency
.41 .41 .43
.08 .44 ).35
0.75 2.53 )3.32
p < .01;
p < .001.
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
789
To examine whether Conviction and Internal Consistency interacted in moderating the attitude±intention relation, the three-way interaction term (attitude by Conviction by Internal Consistency) was entered into the equation after the eect of all factors and their two-way interactions had been partialled out. A signi®cant eect of the three-way interaction term was found (b 1.02, t 2.39, p < .01). In order to unravel the interactive eect of these strength dimensions, median splits based on scores on each factor were employed and attitude±intention correlations were computed for each subgroup. This analysis further supported the notion that one strength component suces to make an attitude in¯uential. Speci®cally, it was found that only when both Conviction and Internal Consistency were low, did attitudes have a weak relation with intentions (r .25). When at least one dimension was high, attitude±intention correlations were very strong (r .73,.86,.61 for the low Conviction, high Consistency, High Conviction high Consistency and High Conviction low Consistency groups, respectively, all zs > 2.57, p < .01). 4. Discussion The ®rst objective of the present study was to investigate Greek people's attitudes towards the European monetary integration and to explore their antecedents and consequences on intentions to support or oppose it. On average, Greeks appear to be supportive of the single European currency and optimistic about its consequences. Also, they seem keen to support the participation of Greece in EMU in case of a referendum. Attitudes were found to be quite strong and were strongly related to behavioural intentions. In particular, respondents perceived the issue as important and personally involving. They also thought they had vested interest in the issue, and they reported attending relevant information. However, they did not think and talk about the issue very often and they did not feel particularly knowledgeable. Nevertheless, they were fairly clear and certain about their attitudes. As will be discussed in the following paragraphs, attitude strength moderated the relation between attitudes and intentions. Greek people expect EMU to encourage cooperation and to facilitate trade and price comparisons across countries. Also, they expect the adoption of euro to strengthen European identity, to increase the political security of Greece and to improve its relations with Turkey and the Balkan countries, although perhaps at the expense of political independence. This might
790
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
indicate that Greeks see other European nations as allies and that they believe that European nations understand the security concerns of Greece. This is particularly important in the context of the Greek±Turkish relations which for the past 30 years have been under the pressure of the consequences of the invasion in Cyprus and of territorial claims and disputes. It is possible that Greeks are prepared to somewhat compromise their political independence to a central European Government, in order to defend their sovereignty. However, Greeks are less optimistic about the economic consequences of EMU, a ®nding which contradicts the widespread belief that the EU will assist Greece in increasing economic growth rates. Although they expect the monetary integration to increase job security and to decrease in¯ation rates, they do not expect it to increase their income level, the value of their savings, to reduce the government debt and taxes and to improve unemployment rates. These ®ndings might re¯ect a pessimism related more to the country itself than to the consequences of the integration. Given the problems of the Greek economy in recent years, it is possible that Greeks are sceptical about the country's ability to bene®t from the potential for economic growth and stability that the monetary integration might oer. In any case, it seems that these negative beliefs are less important in determining attitudes towards EMU than the beliefs related to the anticipated consequences for the whole European community, such as, the facilitation of cooperation and trade across countries and the strengthening of European identity. This class of consequences was a stronger predictor of attitudes than the economic and political outcomes. However, the strongest predictor was people's attitudes towards the participation of Greece in the EU, indicating that people who are positive towards the uni®cation of Europe are also positive towards the monetary integration. Despite the fact that many national projects in the country were at the time funded by the EU and the common belief that southern European countries see the EU as a ``milking cow'', it is surprising that the perceived fairness of the participation of Greece in the EU did not contribute signi®cantly to the prediction of attitudes. The strength of national identity was found to be a strong predictor of attitudes indicating that people who strongly identify with the national group are less in favour of the integration. It should also be noted that the political consequences of the integration and those on the relation with neighbour countries, European identity and life satisfaction were not signi®cant predictors of attitudes. The second objective of the study was to explore the structure of attitude strength and, speci®cally, the interrelations among 12 strength indicators and to investigate its consequences on attitude±intention consistency. Consistent
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
791
with previous ®ndings, the results indicate that attitude strength is a multidimensional construct consisting of three dimensions: Embeddedness, Conviction and Internal Consistency. Embeddedness represents the degree to which the issue or object in question is perceived as personally relevant and important and is re¯ected in personal involvement, frequency of thinking and talking, attention to information, importance and vested interest. Conviction represents the degree to which individuals are convinced their attitude is right and is de®ned by certainty, subjective knowledge, ambivalence and extremity. Finally, Internal Consistency represents the extent to which the evaluative, aective and cognitive components of an attitude are consistent with each other. It is important to note that only Conviction and Internal Consistency moderated signi®cantly the attitude±intention relation and that these factors interacted in a way such that holding an attitude high in either factor suced to make the attitude consequential. These ®ndings indicate that in¯uential attitudes need not necessarily be high on all strength dimensions. It is possible, for example, that holding an attitude with high Conviction makes this attitude strong, irrespective of its personal importance and Internal Consistency. These ®ndings further imply that single indicators of attitude strength do not suce and attitude research needs to incorporate measures of all strength dimensions in order to capture the potential of attitudes to aect behaviour. On the whole, the ®ndings of the present study indicate that Greeks are positively pre-disposed towards EMU. The main antecedents of attitudes towards the single European currency are the attitudes towards the membership of Greece in the EU, the beliefs related to consequences of the integration for Europe, the beliefs related to the economic consequences of the integration for Greece and Greek people and national identity. Further, the results suggest a three-dimensional structure of attitude strength and support the notion that not all strength dimensions need to be high for an attitude to be strong and in¯uential. However, several limitations restrict the generalisability of ®ndings. The most important limitation concerns the use of selfreported measures and the use of a non-representative sample. Further research is therefore required before these ®ndings can be accepted. In particular, additional research examining the structure of attitude strength across dierent issues is necessary. Also, it would be interesting to examine shifts in Greek people's attitudes towards the single currency after the more recent devaluation of the Greek drachma, its entry into the EU's Exchange Rate Mechanism (March 1997) and the exclusion of Greece from the ®rst round of the monetary union.
792
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Acknowledgements The author would like to thank the European team of researchers who contributed to this project, the research assistants in Greece for their help with data collection, and two anonymous reviewers for their constructive comments on an earlier version of this article. Appendix A. Questionnaire items and alpha reliabilities Variable
Questionnaire item
Attitude towards EMU A single European currency is. . . When thinking about a single European currency I feel. . . Are you for or against the European Union having one European currency in all member states, including Greece? That is, replacing the drachma by the European currency? Behavioural intentions How likely is it that you would vote for the introduction of the single currency in our country, in case of a referendum on the participation of Greece in the EMU? Outcome beliefs
What eect do you think the introduction of a single European currency would have on each of the following? Unemployment: The level of unemployment in Greece Debt: Government debt in Greece. In¯ation: Rate of in¯ation in Greece. Political independence: The political independence of Greece. Political security: The political security of Greece in relation to external threats. . .
Endpoints
alpha
reasonable/ unreasonable bad good insecure/secure controllable/ uncontrollable calm/ worried tense/ relaxed
.85
very much for/ very much against
very likely/ very unlikely
greatly decrease/ greatly increase greatly decrease/ greatly increase greatly decrease/ greatly increase greatly decrease/ greatly increase greatly decrease/ greatly increase
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796 Variable
National identity
Questionnaire item In¯ation: Rate of in¯ation in Greece. What impact do you expect a single European currency to have on each of the following? Income: Your personal income will... Savings: The value of your savings will... Job security: Your job security will... Relation with Turkey: What eect do you think that the introduction of a single European currency will have on the relation between Greece and Turkey? Relation with Balkan countries: What eect do you think that the introduction of a single European currency will have on the relation between Greece and the other Balkan countries? Tax: The introduction of a single European currency will lead to tax increases in Greece. Personal money loss: I personally will lose a lot of money if a single European currency is introduced. Price comparisons: With a single European currency we shall be able to compare prices more easily across Europe. Cooperation: The goal of a strong European economy will encourage greater cooperation across Europe. Competitiveness: A single European currency will be better able to compete with the Dollar and the Yen and will therefore gain more in¯uence in the world. Trade: The introduction of a single European currency will make trade between European countries easier. Feel more European: A single European currency will make us feel more European. Price of goods: Do you believe that goods will become more or less expensive with the new currency? I feel attached to Greece and its people.
Endpoints
793 alpha
greatly decrease/ greatly increase greatly decrease/ greatly increase greatly decrease/ greatly increase very positive/ very negative very positive/ very negative
strongly disagree/ strongly agree
strongly disagree/ strongly agree strongly disagree/ strongly agree strongly disagree/ strongly agree strongly disagree/ strongly agree strongly disagree/ strongly agree
strongly disagree/ strongly agree much more expensive/ much less expensive strongly disagree/ strongly agree strongly disagree/strongly agree
794
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Variable
Questionnaire item
Endpoints
I feel attached to Europe and its people.
a very good thing/a very bad thing
European identiy
Attitude towards the EU Generally speaking, do you think that Greece's membership of the European Union is. . .? Perceived fairness Do you think that Greece pays more money into the European Union than it receives, or vice versa? Attitude certainty How certain are you about your attitude towards the single European currency? Personal involvement A single European currency is a subject that might matter more to one person than to another. How important is this subject to you personally? Please indicate the extent to which the issue of EMU is of concern to you? Attitude importance How important do you think is the issue of EMU, in relation to other political issues? Vested interest To what extent do you think that the introduction of the single European currency will have consequences for you personally? Frequency of thinking How often do you think about EMU? Frequency of talking a How often do you discuss the issue of EMU? Subjective knowledge How informed do you think you are about the single European currency? Compared to the average person in Greece, how well informed do you feel about the single European currency? Attention to information How much attention do you pay to information related to the single European currency?
alpha
certain/ uncertain
pays more than it receives/ receives more than it pays Not at all important/ very important of concern to me/ of no concern to me much more important/ much less important very much/ not at all
very often/ rarely
very often/ rarely very well informed/ not at all informed much less informed/ much better informed very much/ not at all
very much/ not at all
.62
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796 Variable Ambivalence
795
Questionnaire item
Endpoints
alpha
To what extent do you have both positive and negative opinions and feelings about the single European currency at same time? To what extent do you think that the introduction of the single European currency will have both positive and negative consequences at the same time?
very much/ not at all
.78
a
In some studies, frequency of thinking and talking about an issue have been treated as indicators of attitude accessibility (see, for example, Wegener et al., 1995). However, it is preferable to treat them as separate indicators given the reported low correlations with response latencies (e.g. Krosnick et al., 1993). b A number of measures of objective knowledge were also included in the questionnaire (e.g. the name and value of the new currency) and the sum of correct responses was calculated. However, scores were heavily skewed (62% of respondents did not provide any correct answers) and therefore this item was dropped from any subsequent analysis. It should be noted that this variable was weakly related to subjective knowledge (r .23).
References Abelson, R.P., 1988. Conviction. American Psychologist 43, 267±275. Eagly, A.H., Chaiken, S., 1993. The Psychology of Attitudes. Ft Worth, Harcourt Brace Jovanovich, TX. Fazio, R.H., Zanna, M.P., 1978. Attitudinal qualities relating to the strength of the attitude-behavior relationship. Journal of Experimental Social Psychology 14, 398±408. Kraus, S.J., 1995. Attitudes and the prediction of behavior: A meta-analysis of the empirical literature. Personality and Social Psychology Bulletin 21, 58±75. Krosnick, J.A., 1988. The role of attitude importance in social evaluations: A study of policy preferences, presidential candidate evaluations, and voting behavior. Journal of Personality and Social Psychology 55, 196±210. Krosnick, J.A., Boninger, D.S., Chuang, Y.C., Berent, M.K., Carnot, G., 1993. Attitude strength: One construct or many related constructs? Journal of Personality and Social Psychology 65, 1132±1151. Krosnick, J.A., Petty, R.E., 1995. Attitude strength: An overview. In: Petty, R.E., Krosnick, J.A. (Eds.), Attitude Strength: Antecedents and Consequences. Erlbaum, Hillsdale, NJ, pp. 1±24. M uller-Peters, A., Kiell, G., Pepermans, R., Battaglia, N., Beckmann, S., Burgoyne, C., van Everdingen, Y.M., Farhangmehr, M., Guzman, G., Kirchler, E., Koenen, C., Kokkinaki, F., Lambkin, M., Lassarre, D., Lenoir, F.-R., Luna-Arocas, R., Marell, A., Meier, K., Moisander, J., Ortona, G., Quintanilla, I., van Raaij, W.F., Routh, D., Scacciati, F., Uusitalo, L., Wahlund, R. Explaining attitudes towards the euro: Design of a cross-national study. Journal of Economic Psychology 19, 663± 680. Norman, R., 1975. Aective-cognitive consistency, attitudes, conformity, and behavior. Journal of Personality and Social Psychology 32, 83±91. Petty, R.E., Krosnick, J.A., 1995. Attitude Strength: Antecedents and Consequences. Erlbaum, Hillsdale, NJ. Prislin, R., 1996. Attitude stability and attitude strength: One is enough to make it stable. European Journal of Social Psychology 26, 447±477.
796
F. Kokkinaki / Journal of Economic Psychology 19 (1998) 775±796
Pomerantz, E.M., Chaiken, S., Tordesillas, R.S., 1995. Attitude strength and resistance processes. Journal of Personality and Social Psychology 69, 408±419. Raden, D., 1985. Strength-related attitude dimensions. Social Psychology Quarterly 48, 312±330. Rattinger, H., 1994. Public attitudes to European integration in Germany after Maastricht: Inventory and typology. Journal of Common Market Studies 32, 525±540. Wegener, D.T., Downing, J., Krosnick, J.A., Petty, R.E., 1995. Measures and manipulations of strengthrelated properties of attitudes: Current practice and future directions. In: Petty, R.E., Krosnick, J.A. (Eds.), Attitude Strength: Antecedents and Consequences. Erlbaum, Hillsdale, NJ, pp. 455±487.