Fratelli D’italia

Differences and similarities in social values between Italian macro-regions

Authored by: Giuseppe A. Veltri

The Routledge Handbook of Contemporary Italy

Print publication date:  May  2015
Online publication date:  May  2015

Print ISBN: 9780415604178
eBook ISBN: 9781315709970
Adobe ISBN: 9781317487555

10.4324/9781315709970.ch2

 

Abstract

According to many scholars, politicians and opinion makers, the social, economic and cultural differences within Italy are of such magnitude to speak of ‘three Italies’: the North, the Centre and the South. For example, the GDP per capita of the South is around 58 per cent of that of the North and Centre, with 36 per cent of the Italian population (Malanima and Zamagni, 2010).

 Add to shortlist  Cite

Fratelli D’italia

Introduction: a divided country

According to many scholars, politicians and opinion makers, the social, economic and cultural differences within Italy are of such magnitude to speak of ‘three Italies’: the North, the Centre and the South. For example, the GDP per capita of the South is around 58 per cent of that of the North and Centre, with 36 per cent of the Italian population (Malanima and Zamagni, 2010).

Debates about the disparity between the North and South have been present almost from Italy’s creation as a unified state in the nineteenth century. A hundred and fifty years later, during the national celebrations of 2011, the presumed cultural differences remain at the core of the public debate on Italian national identity. The later distinction between the North, Centre and South introduced the idea of ‘three Italies’. Surprisingly, since the publication of Edward Banfield’s famous ‘The moral basis of backward societies’ (1965) and Robert Putman’s Making Democracy Work (1993), there has been a lack of recent research on the nature and extent of these cultural differences. There have been some historical studies on the stereotypes of the South in Italian history and culture (Dickie, 1999; Lumley and Morris, 1997; McCrae et al., 2007), but most studies are from the social capital literature (Leonardi, 1995; Girlando et al., 2005) and little else has been published on the topic of cultural difference within Italy using more complete and larger sets of cultural indicators beside ‘social capital’ proxies. Recently, Tabellini (2010) investigated the role of cultural factors for economic development in European regions, including Italy, which he analysed at a regional level using the NUTS-2 classification. Similarly, De Blasio and Nuzzo (2009) studied the role of social capital on local economic development indicators, finding the presence of an effect. Yet, whilst there is a great deal of research on the differences in the economic performance and social structures of the Italian ‘macro-regions’, there are very few recent studies available on the issue of cultural differences that are based on more than one or two indicators.

Over 20 years after Putnam’s work, the differences within Italy continue to be exploited by political parties – in particular the Lega Nord (Northern League) – and have therefore assumed an even more central role in the political and cultural arena. The ostensible cultural differences between the North and South have become so politically charged that some scholars describe these as a form of ‘internal orientalism’ (Schneider, 1998). However, the attribution of regional characteristics to individuals is problematic and runs the risk of falling into the trap of the ecological fallacy (Robinson, 1950; Hofstede, 1980, 2001).

This study will analyse carefully the cultural differences between the Italian macro-regions within the context of ‘social values’. The most important step is to identify suitable definitions of cultural differences that can be used to identify and compare supposed geographical clusters. In this sense, the conceptualization of ‘culture’ is defined by the psychological and social construct of ‘values’. In the social sciences, values are conceptualized in different ways (Hitlin and Piliavin, 2004). Tsirogianni and Gaskell (2011) define social values as ‘socially collective beliefs and systems of beliefs that operate as guiding principles in life’ (p. 2). The sociological perspectives on social values follow the example of Zerubavel (1997) and his proposed cognitive sociology. The next section describes in more detail the theoretical aspects underpinning each set of indicators.

The aim of this study is twofold: first, to determine and assess the differences in social values between macro-regions in Italy, exploring the existence of culturally homogeneous areas distinguishable among them; second, to contribute to the debate in sociology about culture as a ‘latent variable’. Data from the European Value Studies (EVS) will be used to carry out a multi level variance components analysis to identify and determine regional differences.

Social values as cultural indicators

What constitutes ‘culture’ is a long-standing debate in the social sciences, and the use of this term outside academia appears to be even more vague and equivocal. However, when the aim is to compare different cultures, the emphasis is given to cultural traits that are trans-situational, long-standing and social. The notion of ‘values’ – and of social values, in particular – points precisely in this direction. It refers to the more abstract beliefs that are immune to sudden change and has a solid research tradition in the fields of social and cross-cultural psychology and sociology. At the same time, the choice of theoretical approaches is limited by the availability of empirical data specifically collected to measure dimensions of social values. This chapter applies different conceptualizations of the notion of ‘values’ drawn from the literature on cultural sociology, which implement differently the notion of ‘values’ as a way to measure ideological dimensions: economic conservatism-liberalism (Middendorp, 1978); a democratic/authoritarian dimension (Eckstein, 1966); generalized trust as the foundation of civic participation; and social capital (Uslaner, 2002; Putnam, 1993).

Economic attitudes are measured in the EVS (2008) in terms of economic liberalism and conservatism versus a progressive economic agenda (Middendorp, 1978). The items are designed to capture individual preferences regarding welfare, unemployment, competition, commercial freedom and income inequality. This set of indicators is of particular interest given the significant differences in GDP per capita between Italian regions and the differences in economic development between the more advanced ‘North’ and the rest of Italy (Malanima and Zamagni, 2010). Cultural ‘compatibility’ is often considered an independent variable of economic performance (Huntington and Harrison, 2000).

The set of indicators in the EVS (2008) that relate to the attitudes of Italian citizens towards political systems and democracy includes items on leadership, the role of experts, democracy as a system, democracy and the economy, and democracy and law and order. These indicators represent the conceptual legacy of Harry Eckstein’s congruence theory (Eckstein, 1966), which argues, in essence, that political systems tend to be based on authority patterns that are congruent with the authority patterns of other units of society. Two classic studies rooted in this approach are Almond and Verba’s The Civic Culture (1963) and Putnam’s Making Democracy Work (1993).

The fourth set of indicators of social values applied in this study concern the notion of ‘generalized trust’. Individual actors do something for the general good not because they know other actors but because they trust that their own action will be ‘rewarded’ by the positive development of communal relations. Trust is necessary when role expectations and family relationships no longer help us to anticipate the reactions of our individual or collective interaction partners. Generalized trust in a society is inextricably linked to the generalized trustworthiness of that particular society (Uslaner, 2002); and, unfortunately, to be trusting among a gallery of rogues is to be gullible (Ostrom and Walker, 2003; Cook et al., 2005). Naturally, trust is neither individually rational nor socially beneficial if betrayal is likely: ‘social trust is a valuable community asset if – but only if – it is warranted’ (Putnam, 2000: 135). Moreover, ‘generalized trust’ has the capacity to create positive feedback: trust creates reciprocity and voluntary associations, and reciprocity and associations both strengthen and generate trust (see Putnam, 1993: 163–85).

In summary, the sets of indicators include different approaches to social values: economic conservatism, a pro-democracy/authoritarian dimension and generalized trust, which have been used in the vast sociological literature on social capital and cultural variables to explain differences in economic development. Applying all sets, this study aims to address the following research questions: What are the macro-regional differences in Italy in terms of social values measured by four different sets of conceptual and empirical constructs? Given the large differences that exist in terms of socio-economic conditions, do the five Italian macro-regions represent distinct cultural entities in terms of social values endorsement?

Data and methodology

The analysis uses one data set: the European Value Studies (EVS) 2008 data on Italy. Social values indicators come from the 2008 EVS. Several preparatory steps on the data were needed before the main analysis.

The first step was to recode the regional classification of respondents to create a variable for macro-region classification that is the same for both data sets. Data were nested and analysed using the software MLwIN 2.27 (Rasbash et al., 2009). The second step is to identify the controlling variables to perform between-groups comparisons. Two demographic characteristics were selected as control variables: age and education. Initially the variable ‘income’ was also selected, but it was discarded because of the high number of missing values in both data sets, which reduced considerably the size of the samples for the subsequent analysis. Age and education were selected as control variables because of their impact on social values (Ferssizidis et al., 2010; for education, albeit in the context of social capital (including generalized trust), see Huang et al., 2009).

The third step is to compute a baseline two-level model (with the two controlling variables) to determine variances at both levels. The fourth is to perform a multilevel variance components analysis calculating the variance partition coefficient 1 that is the proportion of total variance due to level-2 (in this case macro-regional) differences. A VPC equal to 1 would tell us that all the people in a particular macro-region have an identical level of endorsement of a specific cultural indicator (that is, 100 per cent of the total individual differences are at the macro-regional level); a VPC equal to zero would indicate that the people do not share any macro-regional-related common level of endorsement. A high VPC value informs us that macro-regions are very important for understanding individual differences in social values. On the other hand, a VPC of zero would suggest that the macro-regions are similar to random samples taken from any location and would suggest that macro-regions are not relevant for understanding cultural differences. 2

Analysis

The sets of indicators from the EVS survey have been analysed to determine and quantify macro-regional differences. The Italian macro-regions used in this analysis are: the North-West or ‘NW’ (EVS, N = 406) which includes Piedmont, the Aosta Valley, Liguria and Lombardy; the North-East or ‘NE’ (EVS, N = 302) including the Veneto, Friuli Venezia Giulia, Trentino Aldo Adige (composed of the ‘Province Autonome of Trento and Bolzano’) and Emilia Romagna; the Centre or ‘C’ (EVS, N = 282) comprising Tuscany, Umbria, the Marche and Lazio; the South or ‘S’ (EVS, N = 358) consisting of Abruzzo, Molise, Campania, Apulia and Calabria; and the Islands (‘I’): Sardinia and Sicily (EVS, N = 171). This particular configuration of regions in macro-regions is that adopted by the Italian Statistical Office (ISTAT) and EUROSTAT. The analysis begins with the evaluation of economic conservatism, followed by authoritarianism and generalized trust.

This section presents the analysis carried out on four sets of survey items from the European Values Survey 2008, covering economic conservatism, authoritarianism, democratic values and generalized trust.

Economic conservatism and macro-regions

The first set, shown in Table 2.1, regards the dimension of economic conservatism measured by six indicators: individual responsibility (1) vs state responsibility (10); the unemployed should not refuse jobs (1) vs unemployed have a right to refuse jobs (10); competition is good for people (1) vs. competition harmful (10); state should give freedom to firms (1) vs. state should control firms (10); equalize incomes (1) vs. incentives for individual efforts (10); private ownership should be increased (1) vs. state ownership increased (10). Table 2.1 presents the mean scores of economic conservatism by macro-region. The Islands and the South, followed by the Centre, are the most supportive of the state having responsibility for ensuring that everyone is provided for. These regions are followed by the North-West and North-East respectively.

Table 2.1   Multilevel variance components analysis for each cultural set by macro-regions

Italian 5 Macro-regions

individual vs state responsibility for providing for people (Q58A)

take any job vs right to refuse job when unemployed (Q58B)

competition good vs harmful for people (Q58C)

state to give more freedom to firms vs to control firms more effectively (Q58D)

equalize incomes vs incentives f or individual effort (Q58E)

private vs state ownership of business (Q58F)

NW

Mean

5.51

3.65

4.60

5.74

6.07

4.60

N

393

395

391

380

392

361

Std. Deviation

2.533

2.561

2.480

2.708

2.709

2.220

NE

Mean

5.14

3.23

4.52

5.67

5.96

4.08

N

293

294

297

287

293

276

Std. Deviation

2.515

2.157

2.524

2.740

2.691

2.037

C

Mean

5.90

3.46

4.39

6.02

6.14

4.81

N

268

263

257

252

263

235

Std. Deviation

2.564

2.542

2.644

2.719

2.739

2.413

S

Mean

6.20

3.75

4.37

6.28

5.63

5.15

N

345

338

331

319

343

298

Std. Deviation

2.545

2.497

2.633

2.857

2.895

2.593

I

Mean

6.25

3.52

4.23

5.80

5.34

4.96

N

166

163

149

138

161

118

Std. Deviation

2.851

2.559

2.560

2.838

3.138

2.543

Total

Mean

5.75

3.54

4.45

5.91

5.87

4.69

N

1465

1453

1425

1376

1452

1288

Std. Deviation

2.605

2.469

2.563

2.771

2.815

2.368

Regarding the next indicator on the ‘unemployed having the right to accept or refuse any job’, there are almost no differences in the mean scores of the macro-regions: all the macro-regions lean towards the statement that the unemployed should accept any job available. The same can be said about the item ‘competition is good or harmful for people’: there is very little difference in the mean scores and all the macro-regions endorse the idea that competition is good for stimulating people to work hard and encouraging the development of new ideas.

The next item in the economic conservatism set concerns the freedom of firms and the regulatory role of the state. This topic offers a rather mixed picture: the South and the Centre were the most in favour of the state controlling firms, while the Islands, the North-West and North-East – the latter being the most entrepreneurial region of Italy – endorsed this the least. However, all the macro-regions leant towards endorsing more state control. The next item regards the polarities of a more equal vs a more meritocratic society. All the macro-regions are slightly inclined towards support for a meritocracy, with differences in the intensity of their endorsement. Differences between the macro-regions are not significant, as proved by the MANOVA analysis presented later. The last item concerns the role of the state in owning businesses and industry. The least supportive of state ownership and instead endorsing more private ownership is the North-East followed by the North-West and the Centre (M = 4.81, SD = 2.41). The South and the Islands are more supportive of state ownership.

Results were analysed using a one-way MANOVA between-groups design controlled for the respondents’ age and education. This analysis revealed a significant multivariate effect of macro-regional identity on values scores, F(23, 4166) = 3.254, p < 0.0001, Wilks’ λ 0.937. Subsequent univariate ANOVA of the response variables showed significance at the p < .05 differences between groups for several variables of economic conservatism values: ‘individual vs state responsibility’, F(4, 1201) = 7.706, p < 0.000; ‘take any job vs right to refuse when unemployed’, F(4, 1201) = 2.592, p < 0.05; ‘private vs state ownership of business’, F(4, 1201) = 7.262, p < 0.000. Differences were not statistically significant at the p < .05 for ‘competition is good vs harmful for people’, ‘more freedom vs control for firms’ and ‘equalize incomes vs incentives for individual efforts’.

Table 2.2 presents the amount of variance at the macro-regions level σu 2 and at an individual level. The VPC quantifies the proportion of variance at macro-regional level (or level-2) for each of the six indicators in the economic conservatism set. The proportion of variance due to differences at the macro-regional level for the first indicator – ‘individual vs state responsibility for providing for people’ – is 4 per cent. The second indicator regarding the right to refuse a job when unemployed has a 5.7 per cent variance due to differences between macro-regions. For the third indicator, regarding competition being helpful vs harmful for people, the level-2 variance amounts to 4 per cent. The fourth item, on having more freedom for firms or more state control, has a 3.4 per cent variance due to differences between macro-regions. The fifth item on the choice between a more equal vs a more meritocratic society has a level of variance at the macro-region level of 5 per cent. Similarly, the last indicator on the larger or smaller role of the state in the economy has a 5 per cent variance due to differences between macro-regions. Overall, the multilevel variance components analysis of the economic conservatism indicators has, on average, a 4.5 per cent variance due to differences between macro-regions.

Table 2.2   Means and standard deviations of economic conservatism indicators across macro-regions

Values

Macro-regional level variance (σu 2)

Individual level variance (σe 2)

VPC

VPC %

N

Individual vs state responsibility for providing for people

0.27

6.393

0.0405

4.05

1465

Take any job vs right to refuse job when unemployed

0.331

5.492

0.0568

5.68

1453

Competition good vs harmful for people

0.262

6.322

0.0397

3.97

1425

State to give more freedom to firms vs to control firms more effectively

0.26

7.372

0.0340

3.40

1376

Equalize incomes vs incentives for individual efforts

0.406

7.394

0.0520

5.20

1452

Private vs state ownership of business

0.291

5.334

0.0517

5.17

1288

Factor 1 Dissatisfaction towards democracy

0.037

0.886

0.0400

4.00

1054

Factor 2 Authoritarianism

0.005

0.965

0.0051

0.51

1054

Trust: People can be trusted vs you can’t be too careful

0.01

0.197

0.0483

4.83

1456

Trust: Most people try to take advantage of you vs try to be fair

0.284

4.754

0.0563

5.63

1481

People mostly look out for themselves vs mostly try to be helpful

0.377

4.644

0.0750

7.50

1500

Authoritarianism and pro-democracy values

The second and third sets 3 of indicators aim to capture the degree of authoritarianism and the support for a democratic system, and all the indicators of these two sets are analysed in terms of latent variables or factors. A first group of indicators concerns ‘leadership’ or preference for a strong leader (and therefore a reduced role for parliament and elections), the role of experts in decision-making, and a general endorsement of the overall merits of a democratic political system. Table 2.3 reports the mean scores and standard deviations of the three items by the five macro-regions.

Table 2.3   Means and standard deviations of authoritarianism indicators across macro-regions

Italian 5 Macro-regions

political system: strong leader (Q66A)

political system: experts making decisions (Q66B)

political system: army ruling (Q66C)

political system: democratic (Q66D)

NW

Mean

3.42

2.56

3.68

1.38

N

382

372

384

383

Std. Deviation

.815

.916

.626

.580

NE

Mean

3.42

2.72

3.68

1.43

N

286

268

284

291

Std. Deviation

.807

.919

.557

.609

C

Mean

3.43

2.63

3.68

1.34

N

255

251

256

259

Std. Deviation

.871

.905

.619

.564

S

Mean

3.27

2.72

3.63

1.40

N

329

305

316

325

Std. Deviation

.963

.951

.611

.567

I

Mean

3.41

2.73

3.69

1.47

N

162

142

156

159

Std. Deviation

.903

.981

.528

.549

Total

Mean

3.39

2.66

2.66

1.40

N

1414

1338

1338

1417

Std. Deviation

.871

.931

.928

.577

Regarding the first item, there are no statistically significant differences in the mean scores across the five macro-regions. The second indicator on ‘having experts making decisions, not the government’, reveals some minor differences, with the North-East, the South and the Islands finding the role of experts to be more negative in comparison with the Centre and the North-West, which is the least sceptical. The last item in this group is an evaluation of democracy as a political system: citizens of all the macro-regions consider democracy to be beneficial and there are only very marginal differences in that regard. Results were analysed using a oneway MANOVA between-groups design controlled for age and education. This analysis revealed a significant multivariate effect of regional identity on values scores using macro-regions, F(11, 3328) = 2.057, p < 0.05, Wilks’ λ 0.981. Subsequent ANOVA of the response variables showed that differences in two items – the role of leaders and the overall appraisal of democracy – were not statistically significant at p < .05. Differences in attitudes towards the role of experts across macro-regions were statistically significant F(4, 1262) = 2.542, p < 0.05.

Table 2.4 reports the means and standard deviations for the third set 4 of indicators on the endorsement of democracy. The first item concerns democracy as the best possible political system (although not without its pitfalls) and there are almost no differences in the mean scores: all respondents agree democracy to be the best possible system. The second item regards democracy as bad for the economy. Similarly, there are no large differences between the macro-regional mean scores and all of them tend to disagree with this statement, with the South and Islands disagreeing to a slightly lesser extent. The next item describes democracies as ‘indecisive and hav[ing] too much squabbling’; there are similar mean scores for all the macro-regions without any significant difference: all respondents mildly disagree with this statement. The last indicator states that ‘democracies aren’t good at maintaining order’. On this statement, once again, there are no large differences among Italian citizens from the five macro-regions. All of them tend to disagree with this statement; the South alone was slightly less in disagreement.

Table 2.4   Mean scores of three items measuring attitudes towards democracy across macro-regions

Italian 5 Macro-regions

democracy: best political system (Q67A)

democracy: causes bad economy (Q67B)

democracy: is indecisive (Q67C)

democracy: cannot maintain order (Q67D)

NW

Mean

1.52

2.81

2.50

3.02

N

376

362

370

372

Std. Deviation

.589

.649

.730

.684

NE

Mean

1.56

2.86

2.48

3.01

N

286

256

276

275

Std. Deviation

.628

.623

.690

.678

C

Mean

1.51

2.86

2.61

3.14

N

255

223

248

241

Std. Deviation

.560

.696

.817

.709

S

Mean

1.53

2.77

2.52

2.98

N

326

268

302

300

Std. Deviation

.541

.669

.768

.715

I

Mean

1.56

2.61

2.47

3.02

N

153

125

131

133

Std. Deviation

.572

.792

.871

.707

Total

Mean

1.53

2.80

2.52

3.03

N

1396

1234

1327

1321

Std. Deviation

.579

.675

.763

.698

Results were analysed using a one-way MANOVA between-groups design controlled for age and education. This analysis revealed a significant multivariate effect of macro-regional identity on values scores F(15, 3419) = 1.914, p < 0.05, Wilks’ λ 0.973. Subsequent ANOVA of the response variables showed that differences in all four items on democracy were not statistically significant at p < .05.

Both sets of indicators were analysed using a factor analysis. A Principal Components Analysis (PCP) with a Varimax (orthogonal) rotation of the previous 7 Likert scale questions was conducted on data gathered from 1,054 participants. An examination of the Kaiser-Meyer Olkin measure of sampling adequacy suggested that the sample was factorable (KMO = 0.803). The results of an orthogonal rotation of the solution are shown in Table 2.5. When loadings of less than 0.25 were excluded, the analysis yielded a two-factor solution with a simple structure (factor loadings = > 0.25).

Table 2.5   PCP: rotated two-factor solution

Rotated Component Matrix a

Component

1

2

democracy: is indecisive (Q67C)

.836

democracy: causes bad economy (Q67B)

.781

democracy: cannot maintain order (Q67D)

.690

.282

political system: experts making decisions (Q66B)

.262

political system: democratic (Q66D)

–.751

democracy: best political system (Q67A)

–.737

political system: the army ruling (Q66C)

.664

political system: strong leader (Q66A)

.429

.480

The two-factor solution is interpreted as follows: four items are loaded mainly into one factor and it is clear from Table 2.5 that all the items – especially the first three – relate to a dimension of dissatisfaction with democracy; three items define a second factor and all relate to a form of authoritarianism. Two items – ‘democracy cannot maintain order’ and the preference for strong leaders – contributed to both factors.

Results were analysed using a one-way MANOVA between-groups design controlled for age and education. This analysis revealed a significant multivariate effect of macro-regional identity on the factor scores F(4, 2092) = 2.365, p < 0.05, Wilks’ λ 0.989. Subsequent ANOVA of the response variables showed that only in one factor – the authoritarianism dimension – are the macro-regional differences statistically significant at the level of p < .05.

Table 2.2 shows the amount of variance of both factors due at the macro-regions level σ2 u and the variance at the individual level. The proportion of variance due to differences at the macro-regional level for the first factor – ‘dissatisfaction towards democracy’ – amounts to 4 per cent. In the second factor on the dimension of authoritarianism, level-2 variance is at 0.5 per cent and this is confirmed by the fact that a level-two model is not statistically superior to a one level solution, ‘Factor 2 Authoritarianism’ 2, N = 1482) = 0.942, p = 0.331.

Generalized trust and macro-regions

The fourth set of indicators concerns one of the most heavily debated aspects of cultural differences between the macro-regions in Italy: ‘generalized trust’. As discussed previously, generalized trust has often been associated with economic development as a precondition of healthy economic transactions and relationships reducing their costs. Table 2.6 shows the mean scores and standard deviations.

Table 2.6   Means and standard deviations of generalized trust items across macro-regions

Italian 5 Macro-regions

people can be trusted vs you can’t be too careful (Q7)

most people try to take advantage of you vs try to be fair (Q8)

people mostly look out for themselves vs mostly try to be helpful (Q9)

NW

Mean

1.59

5.57

4.36

N

396

403

403

Std. Deviation

.492

2.154

2.171

NE

Mean

1.67

6.00

4.38

N

283

294

298

Std. Deviation

.469

2.119

2.167

C

Mean

1.67

5.36

4.06

N

272

273

278

Std. Deviation

.473

2.311

2.248

S

Mean

1.79

4.94

4.05

N

344

345

352

Std. Deviation

.407

2.356

2.346

I

Mean

1.80

4.69

3.67

N

161

166

169

Std. Deviation

.400

2.253

2.211

Total

Mean

1.69

5.37

4.16

N

1456

1481

1500

Std. Deviation

.462

2.274

2.240

The first item is about choosing the endorsement between the binary choices ‘most people can be trusted’ and ‘you can’t be too careful in dealing with people’. All respondents leaned towards the second option. The South and the Islands are stronger endorsers of the ‘not trusting’ option compared with the Centre and the North-East. The strongest endorser of generalized trust for this item was the North-West. The second trust indicator is constituted by two polarities: ‘most people try to take advantage of you’ and ‘most people try to be fair’. 5 Similarly, the Islands and the South mostly leaned towards the cynical pole compared with the more trusting Centre and North-West. The most trusting macro-region was the North-East. The third trust indicator concerns the choice between the two polarities ‘people mostly look out for themselves’ and ‘people mostly try to be helpful’. All respondents leaned towards the cynical pole. As with the previous item, the Islands and the South are the two macro-regions that are the least trusting, closely followed by the Centre. The most trusting respondents were from the North-West and the North-East.

Differences in scores were analysed using a one-way MANOVA between-groups design controlled for age and education. This analysis revealed a significant multivariate effect of regional identity on values scores using macro-regions, F(11, 3672) = 5.375, p < 0.000, Wilks’ λ 0.955. Subsequent ANOVA of the response variables showed that differences in all three items were statistically significant at the p < .05: ‘people can be trusted/you can’t be too careful’ F (4, 1392) = 9.110, p < 0.000; ‘most people try to take advantage/try to be fair’ F(4, 1392) = 9.615, p < 0.000; ‘people mostly look out for themselves/mostly try to be helpful’ F(4, 1392) = 2.629, p < 0.05.

Table 2.2 also shows the amount of variance for the political system values set of indicators at the macro-regional level and at an individual level. The proportion of variance due to differences at the macro-regional level for the first indicator, ‘people can be trusted/you can’t be too careful’, amounts to 4.8 per cent. The second indicator of generalized trust, ‘most people try to take advantage of you/try to be fair’ has a level-2 variance of 5.6 per cent. The third item or ‘people mostly look out for themselves/mostly try to be helpful’, has 7.5 per cent variance due to macro-regional differences. Overall, the average proportion of variance due to differences between the macro-regions for this set was 6 per cent.

Discussion

In the dimension of ‘economic conservatism’, the five Italian macro-regions are extremely similar without considerable differences. The same can be said of the factorial dimensions of authori tarianism vs support of democracy. Regarding the ‘generalized trust’ dimension, the South and the Islands are only slightly more distrustful than the North and, in particular, the North-East – which is the area in which respondents are the most willing to trust others.

A small regional cultural variance is in line with a recent study by Tabellini (2010), which found that the regional dummy variables for most of his cultural indicators explained around 6 per cent of the variance and that ‘regional distributions are clearly different, but the range of variation within each region remains large’ (p. 687). Tabellini interpreted this finding as ambiguous because of the small number of respondents in some regions and therefore questioned the representativeness of the samples. In this study, the analysis of five macro-regions indicated the same low proportion of variance explained at macro-regional level, although much larger samples were used.

Overall, however, the findings reveal the existence of small differences between the Italian macro-regions analysed. There are several potential interpretations and explanations for such findings but two aspects are particularly important: one is historical and one is related to social and psychological theory.

First, an interpretation of these results should consider Italian history and in particular the relationships between Northern, Central and Southern Italy. Following the unification of the Italian kingdom (in 1861), there was an almost continuous flow of migration from Southern Italy to the rest of the country, particularly after the Second World War (during the 1950s and 1960s), when an enormous number of people left the South to live in the Centre and the North of Italy. In other words, internal migration in Italy has been (and continues to be) a notable phenomenon. Hence, regional cultures have undergone many years of mutual contamination and influence.

Second, people implement and combine relational models depending on their values, position in society, groups, institutional and cultural contexts, historical processes, and relations to others, etc. In other words, people maintain distinctive and inconsistent action frames that are invoked in response to particular contextual cues (DiMaggio, 1997). For example, social values also compete strongly with social norms (Bardi and Schwartz, 2003) and may not be used as much by individuals who value conformity to the social context (Lönnqvist et al., 2006; Mellema and Bassili, 1995). In conclusion, the existing differences in collective behaviour between the Italian macro-regions should be explained more accurately considering the combined effect of the situational and ecological factors, rather than in terms of a coherent cultural expression.

Conclusions

The analysis indicates that cultural differences between macro-regions do exist but the situation is complex with some noticeable differences and many surprising similarities. However, it is difficult to judge whether such differences should be considered significant and what effects they might have. Ideally, a comparison with other European countries and their internal diversity would provide a reference point; for example Steel and Taras (2010) consider a variance of between 3 and 18 per cent at country level as small, highlighting the danger of using national averages to make assumptions about individual cultural values.

In general, the proportion of variance explained at the individual level (or differences between individuals) is much larger than that due to the differences at the macro-regional level for the large majority of indicators. Overall, macro-regions are not homogeneous and distinct cultural areas. It follows that there is little evidence of clearly dissimilar cultural macro-areas in terms of large differences in social values among Italian citizens.

The results presented in this study challenge the notion of fundamentally culturally diverse – in terms of individual social values – macro-areas in Italy. In other words, they reveal the ecological fallacy of the application of regional characteristics to individuals without distinguishing between individual correlations and ecological correlations (Robinson, 1950).

There are several other potential analytical routes for future research to follow in order to measure cultural differences in Italy and therefore the evidence is not conclusive. However, this analysis takes an important first step towards debunking a dangerous simplification in the study of Italian society with the hope that other scholars will complement it in the near future.

Notes

The formula for the VPC is:

For example, the level-two baseline model for one indicator with the controlling variables is:

For quantifying explained variance, R2 analogs are defined at each level as the difference between the variance components for the baseline (i.e. intercepts only) model and the variance component for the current model divided by the variance component for the baseline model (Kreft and DeLeeuw, 1998). For testing the significance of the variance between the macro-regions, the log likelihood of the model that includes σu 2 can be compared with the log likelihood of the almost identical model that does not include σ2 u. If the χ2 test with 1 degree of freedom rejects the null hypothesis of no difference at 0.05, then σu 2 is statistically significant. If p is noticeably larger than 0.05, say p > 0.10, then σu 2 is not statistically significant. All reported level-two models are significant compared with a level-one solution, unless stated otherwise.

In the second set, all items were measured with the following scale: -5 other missing; -4 question not asked; -3 nap; -2 na; -1 dk; 1 very good; 2 fairly good; 3 fairly bad; 4 very bad.

In the second set, all items were measured with the following scale: -5 other missing; -4 question not asked; -3 nap; -2 na; -1 dk; 1 very good; 2 fairly good; 3 fairly bad; 4 very bad.

‘Most people would try to take advantage of me’ was 1 and ‘most people would try to be fair’ was 10, variable 63 Q8, EVS 2008. Similarly, for the following item ‘ people mostly look out for themselves’ had value 1 and ‘people mostly try to be helpful’ had value 10, v64 Q9, EVS 2008.

References

Almond, G. A. and Verba, S. (1963). The civic culture: Political attitudes and democracy in five nations. Princeton, NJ: Princeton University Press.
Bardi, A. and Schwartz, S. H. (2003). Values and behavior: Strength and structure of relations. Personality and Social Psychology Bulletin, 29(10): 1207–20.
Cook, K. S. , Hardin, R. and Levi, M. (2005). Cooperation without trust? New York: Russell Sage Foundation.
De Blasio, G. and Nuzzo, G. (2009). Historical traditions of civicness and local economic development. Journal of Regional Science, 50(4): 833–57.
Dickie, J. (1999). Darkest Italy: The nation and stereotypes of the Mezzogiorno, 1860–1900. New York: St. Martin’s Press.
DiMaggio, P. (1997). Culture and cognition. Annual Review of Sociology, 23: 263–87.
Eckstein, H. (1966). Division and cohesion in democracy: A study of Norway. Princeton, NJ: Princeton University Press.
European Values Study (EVS) (2008) Integrated Dataset. GESIS Data Archive, Cologne. ZA4800 Data file version 3.0.0, doi:10.4232/1.11004.
Ferssizidis, P. , Adams, L. M. , Kashdan, T. B. , Plummer, C. , Mishra, A. and Ciarrochi, J. (2010). Motivation for and commitment to social values: The roles of age and gender. Motivation and Emotion, 34(4): 354–62.
Girlando, A. P. , Anderson, C. J. and Zerillo, J. W. (2005). An examination of Hofstede’s paradigm of national culture and its malleability. Journal of Transnational Management, 10(1): 23–36.
Hitlin, S. and Piliavin, J. A. (2004). Values: Reviving a dormant concept. Annual Review of Sociology, 30(1): 359–93.
Hofstede, G. H. (1980). Culture’s consequences: International differences in work-related values. Newbury Park, CA: Sage.
Hofstede, G. H. (2001). Culture’s consequences: comparing values, behaviors, institutions, and organizations across nations. Thousand Oaks, CA: Sage.
Huang, J. , Maassen van den Brink, H. and Groot, W. (2009). A meta-analysis of the effect of education on social capital. Economics of Education Review, 28: 454–64.
Huntington, S. P. and Harrison, L. E. (2000). Culture matters: How values shape human progress. New York: Basic Books.
Kreft, I. and DeLeeuw, J. (1998). Introducing multilevel modeling. Thousand Oaks, CA: Sage.
Leonardi, R. (1995). Regional development in Italy: Social capital and the Mezzogiorno. Oxford Review of Economic Policy, 11(2): 165–79.
Lönnqvist, J. , Leikas, S. , Paunonen, S. , Nissinen, V. and Verkasalo, M. (2006). Conformism moderates the relations between values, anticipated regret, and behavior. Personality and Social Psychology Bulletin, 32: 1469–81.
Lumley, R. and Morris, J. (1997). The new history of the Italian south: The Mezzogiorno revisited. Exeter: University of Exeter Press.
McCrae, R. R. , Terracciano, A. , Realo, A. and Allik, J. (2007). Climatic warmth and national wealth: Some culture-level determinants of national character stereotypes. European Journal of Personality, 21(8): 953–76.
Malanima, P. and Zamagni, V. (2010). 150 years of the Italian economy, 1861–2010. Journal of Modern Italian Studies, 15(1): 1–20.
Mellema, A. and Bassili, J. N. (1995). On the relationship between attitudes and values: Exploring the moderating effects of self-monitoring and self-monitoring schematicity. Personality and Social Psychology Bulletin, 21: 885–92.
Middendorp, C. P. (1978). Progressiveness and conservatism: The fundamental dimensions of ideological controversy and their relationship to social class. The Hague: Mouton.
Ostrom, E. and Walker, J. (2003). Trust and reciprocity: Interdisciplinary lessons from experimental research. New York: Russell Sage Foundation.
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New York: Simon & Schuster.
Putnam, R. D. , Leonardi, R. and Nanetti, R. (1993). Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press.
Rasbash, J. , Charlton, C. , Browne, W. J. , Healy, M. and Cameron, B. (2009). MLwiN Version 2.1. Bristol: Centre for Multilevel Modelling, University of Bristol.
Robinson, W. S. (1950). Ecological correlations and the behavior of individuals. American Sociological Review, 15(3): 351–7.
Schneider, J. (1998). Italy’s ‘Southern question’: Orientalism in one country. Oxford: Berg.
Steel, P. and Taras, V. (2010). Culture as a consequence: A multi-level multivariate meta-analysis of the effects of individual and country characteristics on work-related cultural values. Journal of International Management, 16(3): 211–33.
Tabellini, G. (2010). Culture and institutions: Economic development in the regions of Europe. Journal of the European Economic Association, 8(4): 677–716.
Tsirogianni, S. and Gaskell, G. (2011). The role of plurality and context in social values. Journal for the Theory of Social Behaviour, 41(4): 441–65.
Uslaner, E. M. (2002). The moral foundation of trust. Cambridge and New York: Cambridge University Press.
Zerubavel, E. (1997). Social mindscapes: An invitation to cognitive sociology. Cambridge, MA: Harvard University Press.
Search for more...
Back to top

Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.