University of Macerata From the SelectedWorks of Luca De Benedictis Fall October, 2012 Schools of Thought and Economists' Opinions on Economic Policy Luca De Benedictis Michele Di Maio Available at: https://works.bepress.com/luca_de_benedictis/35/
SCHOOLS OF THOUGHT AND ECONOMISTS OPINIONS ON ECONOMIC POLICY Luca De Benedictis 1 Michele Di Maio 2 1 Università di Macerata - Italy 2 Università Parthenope Napoli - Italy October 18-20, 2012 SIE, Matera
Outline of the talk Grouping and separating Economists - as humans - tend to gather together in communities. We look at economists as communities, and we ask the data: are economists individual opinions related to the School of thought they declare to belong to? Opinions on what? Economic Policy (in Italy). Alternative grouping: (1) space; (2) knowledge (e.g. field); (3) individual characteristics (e.g. gender); (4) ideology and more. Dataset: Survey of Italian economists representativeness a bird-eye view of the questionnaire Schools of Thought Schools of Thought and confounders Econometrics: Baseline: Weighted Ordered Probit Regression Placebo test The role of Schools of Thought: Pluralism The role of Schools of Thought: Mainstream vs TRotW Summing-up
Part I Grouping and separating
The survey We sent (April 27, 2007) an invitation to fill an online questionnaire to 1511 Italian economists both academics and professionals, working in Italy and abroad. The mailing list included: members of the Italian Economic Association (SIE); the International Association of Italian Economists (AIDEI); participants to economic conferences in Italy (last 3 years). In the e-mail message it was clearly stated that no specific knowledge was required to fill in the questionnaire; that anonymity was guaranteed; that filling in the questionnaire would take around 20 minutes. No worry about selection. 496 economists responded to the questionnaire. Reply rate 33%. In this paper we focus only on economists whose characteristics could be matched with a known population (MIUR): 335 observations.
The questionnaire (and follow-up info) The questionnaire included questions about: Individual characteristics (Age, Gender, Region of birth, Region of work, Job, University position, School of Thought, Field of research, Theory/ Empirics, Ideology, Pro-Market orientation, Optimism; the definition of economic decline and the applicability to Italy; the causes of the Italian economic decline (Propositions); the policy proposals to put in place (Propositions); evaluation and participation to the academic/public debate; quantification of personal worry about how things are evolving". We collected further information through online CVs: Graduation (type of Laurea, University, supervisor (many missing)); Master; Dottorato; PhD; Department of affiliation (in April 2007).
(Some) descriptive statistics Variable # % Total number 335 100 Gender Male 255 76 Female 80 24 Age 18-30 5 1 31-40 52 16 41-50 107 32 51-60 98 29 61-70 63 18 71-10 3 Place of work North West 103 31 North East 62 19 Center 119 36 South/Islands 51 14 Academic position Assistant 41 12 Associate 100 30 Full Professor 194 58 total 335 100
(Some) descriptive statistics (2) Variable # % JEL 424 a 100 JEL1 - Schools of Economic Thought and Methodology 20 5 JEL2 - Mathematical and Quantitative Methods 22 5 JEL3 - Microeconomics 37 9 JEL4 - Macroeconomics and Monetary Economics 71 17 JEL5 - International Economics 44 10 JEL6 - Financial Economics 13 3 JEL7 - Public Economics 37 9 JEL8 - Health, Education, and Welfare 6 1 JEL9 - Labor and Demographic Economics 25 6 JEL10 - Law and Economics 8 2 JEL11 - Industrial Organization 52 12 JEL12 - Business Admin. and Business Econ.; Marketing; Account. 4 1 JEL13 - Economic History 10 2 JEL14 - Economic Development, Tech. Change, and Growth 48 11 JEL15 - Economic Systems 3 1 JEL16 - Agricul. and Natural Resource Econ; Env. and Ecol. Econ. 10 2 JEL17 - Urban and Regional Economics 14 3 Note (a): Some respondents declared more than one primary field. Research 335 100 Empirical 21 6 101 30 107 32 70 21 Theoretical 36 11
Part II Schools of Thought
The distribution of Italian economists across SofT SCHOOL OF THOUGHT Number of respondents % Eclectic 95 28.35 Neoclassical/Mainstream 59 17.61 I do not refer to any specific SofT 56 16.71 Keynesian/Post-Keynesian 36 10.74 Keynesian/Neo-Keynesian 27 8.06 Institutionalist/Neo-Institutionalist 23 6.87 Evolutionary 17 5.07 Marxist/Sraffian/Neo-Marxist 13 3.88 Behavioral 5 1.49 Regulationist 3 0.89 Austrian/Neo-Austrian 1 0.29 Total 335 100 NOTE: The table includes the responses to the question: How will you define your methodological orientation? Multiple answers were possible: respondents that indicated three or more options (7 cases out of 335) were classified as Eclectic; double responses (32 cases) were randomly assigned to one of the two options indicated; specific (open) responses (15 cases) have been reclassified according to similarity (e.g. Marxist-Leninist have been classified as Marxist). The list of SofT is the one in Axarloglou and Theoharakis (JEEA, 2003) we also included the Eclectic category. Moreover, all the SofT discussed in Colander et al. (Rev.Pol.Ec., 2003) are included in our list.
Age distribution and SofT Neoclassical/Mainstream Evolutionary Regulationist No-school Eclectic Post-Keynesian Neo-Keynesian Marxist/Sraffian/Neo-Marxist Austrian/Neo-Austrian Institutionalist/Neo-Institutionalist Behavioral (17%) (32%) (29%) (19%) (3%) 18-40 41-50 51-60 61-70 71-
Pro-market distribution and SofT Marxist/Sraffian/Neo-Marxist Post-Keynesian Neo-Keynesian Institutionalist/Neo-Institutionalist Evolutionary Eclectic Regulationist No-school Austrian/Neo-Austrian Neoclassical/Mainstream Behavioral (2%) (9%) (45%) (38%) (12%) State Mixed Economy Market
Ideology distribution and SofT Marxist/Sraffian/Neo-Marxist Behavioral Regulationist Post-Keynesian Evolutionary Neo-Keynesian Eclectic Institutionalist/Neo-Institutionalist No-school Neoclassical/Mainstream Austrian/Neo-Austrian (22%) (57%) (16%) (7%) (3%) Left Center Right
Part III Econometrics
Benchmark: Weighted Ordered Probit Regression We estimated 58 weighted ordered probit regressions: one for each proposition i. The dependent variable opinion i is the detection of a continuous latent variable, opinion i, that can be divided into J = 4 ordinal categories, where the thresholds separating the different categories are going to be estimated. The J = 4 categories correspond to the four ordered options of the Likert scale used by the respondents to express their opinions on each of the 58 propositions: Strongly Disagree (SD)=1; Partially Disagree (PD)=2; Partially Agree (PA)=3; Strongly Agree (SA)=4. Regression: opinion i = β 1SofT i + c i β c + u i, (1) c i : confounders weighting: MIUR (SecS P01-P02-P03-P04-P05-P06-P012) University of graduation fixed-effects clustering: University/Departments
Benchmark (Causes) - Eclectics vs Mainstream Cause Eclectics 3 Dumping and unfair international competition -0.601 ** 5 Low attraction of FDI -0.440 7 Adoption of the Euro -0.951 ** 8 European Commission economic policy -0.312 9 BCE monetary policy -0.384 10 Primary commodity world price dynamics -1.234 *** 11 Difficult international political situation -1.547 *** 14 Small firms size (more difficult the adoption of new technologies) 0.275 15 Small firms s size (more difficult the innovation activity) 0.194 18 Low risk propensity of entrepreneurs -0.824 *** 19 Excessive protection of large domestic firms -0.645 ** 21 Quantity and quality of infrastructures -0.588 * 22 Quality of immaterial infrastructures (justice, authority, etc.) -0.740 23 Low competition level, the existence of barriers to entry -0.614 24 Bureaucratic impediments to private entrepreneurship -0.627 ** 25 Difficulties to gaining access to credit -0.764 *** 26 Persistence of the Italian North-South economic divide -0.018 27 Mezzogiorno b issue (crime) -0.259 28 Mezzogiorno b issue (infrastructures) 0.097 30 Low labor market flexibility -1.397 *** 32 Demographic dynamics -0.845 *** 34 Low human capital supply -1.764 *** 36 Labor Union behavior -0.580 * 37 Public debt level and composition -0.754 ** 38 Type of policies adopted to reduce public debt -0.285 39 Low efficiency of the Public Administration -0.470 * 40 Low efficiency of the bureaucracy -0.521 Notes: Confounders (included but coefficients are not shown): AGE, GENDER, MIGRANT, WORRY, OPTIMISM, REGION OF BIRTH, STUDIED ABROAD [1 AND 2], ACADEMIC POSITION, REGION OF WORK, MASTER, DOTTORATO/PhD, FIELD OF RESEARCH, ITALY, DEBATE, MARKET, SOCIAL MOBILITY and RIGHT. We include fixed-effects for the University where the respondent received her B.A (UNIVERSITY BA). Standard error are robust and clustered at the level of the University where the respondent is currently employed. Weights are used. Luca De Benedictis and Michele Di Maio
Benchmark (Causes) - Institutionalists vs Mainstream Cause Institutionalists 3 Dumping and unfair international competition 0.495 5 Low attraction of FDI -0.216 7 Adoption of the Euro 0.166 8 European Commission economic policy 0.724 * 9 BCE monetary policy 0.994 *** 10 Primary commodity world price dynamics -0.313 11 Difficult international political situation -0.512 14 Small firms size (more difficult the adoption of new technologies) -0.236 15 Small firms s size (more difficult the innovation activity) 0.098 18 Low risk propensity of entrepreneurs -0.467 19 Excessive protection of large domestic firms -0.302 21 Quantity and quality of infrastructures 0.066 22 Quality of immaterial infrastructures (justice, authority, etc.) -0.348 23 Low competition level, the existence of barriers to entry -0.812 ** 24 Bureaucratic impediments to private entrepreneurship -0.735 * 25 Difficulties to gaining access to credit -0.132 26 Persistence of the Italian North-South economic divide 0.414 27 Mezzogiorno b issue (crime) -0.188 28 Mezzogiorno b issue (infrastructures) -0.200 30 Low labor market flexibility -0.213 32 Demographic dynamics -0.192 34 Low human capital supply -1.436 *** 36 Labor Union behavior -0.362 37 Public debt level and composition 0.071 38 Type of policies adopted to reduce public debt 0.742 * 39 Low efficiency of the Public Administration -0.483 40 Low efficiency of the bureaucracy -0.356 Notes: Confounders (included but coefficients are not shown): AGE, GENDER, MIGRANT, WORRY, OPTIMISM, REGION OF BIRTH, STUDIED ABROAD [1 AND 2], ACADEMIC POSITION, REGION OF WORK, MASTER, DOTTORATO/PhD, FIELD OF RESEARCH, ITALY, DEBATE, MARKET, SOCIAL MOBILITY and RIGHT. We include fixed-effects for the University where the respondent received her B.A (UNIVERSITY BA). Standard error are robust and clustered at the level of the University where the respondent is currently employed. Weights are used. Luca De Benedictis and Michele Di Maio
Benchmark (Causes) - Neo-Keynesians vs Mainstream Cause Neo-Keynesians 3 Dumping and unfair international competition 0.837 * 5 Low attraction of FDI 0.202 7 Adoption of the Euro -0.540 8 European Commission economic policy -0.148 9 BCE monetary policy 0.774 10 Primary commodity world price dynamics -0.399 11 Difficult international political situation -0.680 14 Small firms size (more difficult the adoption of new technologies) 0.767 * 15 Small firms s size (more difficult the innovation activity) 1.106 *** 18 Low risk propensity of entrepreneurs -1.362 *** 19 Excessive protection of large domestic firms -0.893 ** 21 Quantity and quality of infrastructures -0.790 * 22 Quality of immaterial infrastructures (justice, authority, etc.) -0.758 23 Low competition level, the existence of barriers to entry -1.029 ** 24 Bureaucratic impediments to private entrepreneurship -0.481 25 Difficulties to gaining access to credit -0.318 26 Persistence of the Italian North-South economic divide 0.094 27 Mezzogiorno b issue (crime) 0.196 28 Mezzogiorno b issue (infrastructures) 0.278 30 Low labor market flexibility -0.519 32 Demographic dynamics -0.163 34 Low human capital supply -1.410 *** 36 Labor Union behavior -0.521 37 Public debt level and composition -0.079 38 Type of policies adopted to reduce public debt -0.160 39 Low efficiency of the Public Administration -0.749 * 40 Low efficiency of the bureaucracy -0.769 * Notes: Confounders (included but coefficients are not shown): AGE, GENDER, MIGRANT, WORRY, OPTIMISM, REGION OF BIRTH, STUDIED ABROAD [1 AND 2], ACADEMIC POSITION, REGION OF WORK, MASTER, DOTTORATO/PhD, FIELD OF RESEARCH, ITALY, DEBATE, MARKET, SOCIAL MOBILITY and RIGHT. We include fixed-effects for the University where the respondent received her B.A (UNIVERSITY BA). Standard error are robust and clustered at the level of the University where the respondent is currently employed. Weights are used. Luca De Benedictis and Michele Di Maio
Benchmark (Causes) - Post-Keynesians vs Mainstream Cause Post-Keynesians 3 Dumping and unfair international competition -0.302 5 Low attraction of FDI -0.973 ** 7 Adoption of the Euro -0.431 8 European Commission economic policy 0.309 9 BCE monetary policy 0.593 10 Primary commodity world price dynamics -0.298 11 Difficult international political situation -1.141 *** 14 Small firms size (more difficult the adoption of new technologies) 0.182 15 Small firms s size (more difficult the innovation activity) 0.668 * 18 Low risk propensity of entrepreneurs -0.763 ** 19 Excessive protection of large domestic firms -0.350 21 Quantity and quality of infrastructures -0.252 22 Quality of immaterial infrastructures (justice, authority, etc.) -2.204 *** 23 Low competition level, the existence of barriers to entry -1.112 ** 24 Bureaucratic impediments to private entrepreneurship -0.469 25 Difficulties to gaining access to credit 0.021 26 Persistence of the Italian North-South economic divide 0.394 27 Mezzogiorno b issue (crime) 0.478 28 Mezzogiorno b issue (infrastructures) 1.007 *** 30 Low labor market flexibility -1.342 *** 32 Demographic dynamics -0.528 34 Low human capital supply -1.154 ** 36 Labor Union behavior -0.857 * 37 Public debt level and composition -0.765 ** 38 Type of policies adopted to reduce public debt 0.170 39 Low efficiency of the Public Administration -0.995 ** 40 Low efficiency of the bureaucracy -0.847 * Notes: Confounders (included but coefficients are not shown): AGE, GENDER, MIGRANT, WORRY, OPTIMISM, REGION OF BIRTH, STUDIED ABROAD [1 AND 2], ACADEMIC POSITION, REGION OF WORK, MASTER, DOTTORATO/PhD, FIELD OF RESEARCH, ITALY, DEBATE, MARKET, SOCIAL MOBILITY and RIGHT. We include fixed-effects for the University where the respondent received her B.A (UNIVERSITY BA). Standard error are robust and clustered at the level of the University where the respondent is currently employed. Weights are used. Luca De Benedictis and Michele Di Maio
Benchmark (Causes) - Marxists vs Mainstream Cause Marxists 3 Dumping and unfair international competition -1.576 ** 5 Low attraction of FDI -2.100 *** 7 Adoption of the Euro -0.597 8 European Commission economic policy 1.749 ** 9 BCE monetary policy 1.962 10 Primary commodity world price dynamics -0.844 11 Difficult international political situation -0.788 14 Small firms size (more difficult the adoption of new technologies) -0.895 * 15 Small firms s size (more difficult the innovation activity) -0.571 18 Low risk propensity of entrepreneurs -0.400 19 Excessive protection of large domestic firms -0.966 * 21 Quantity and quality of infrastructures -1.189 * 22 Quality of immaterial infrastructures (justice, authority, etc.) -3.273 *** 23 Low competition level, the existence of barriers to entry -3.260 *** 24 Bureaucratic impediments to private entrepreneurship -1.681 *** 25 Difficulties to gaining access to credit -0.588 26 Persistence of the Italian North-South economic divide 1.184 ** 27 Mezzogiorno b issue (crime) 0.134 28 Mezzogiorno b issue (infrastructures) -0.316 30 Low labor market flexibility -10.037 *** 32 Demographic dynamics -1.535 *** 34 Low human capital supply -3.535 *** 36 Labor Union behavior -1.442 *** 37 Public debt level and composition -2.101 *** 38 Type of policies adopted to reduce public debt 1.150 39 Low efficiency of the Public Administration -1.488 * 40 Low efficiency of the bureaucracy -2.246 *** Notes: Confounders (included but coefficients are not shown): AGE, GENDER, MIGRANT, WORRY, OPTIMISM, REGION OF BIRTH, STUDIED ABROAD [1 AND 2], ACADEMIC POSITION, REGION OF WORK, MASTER, DOTTORATO/PhD, FIELD OF RESEARCH, ITALY, DEBATE, MARKET, SOCIAL MOBILITY and RIGHT. We include fixed-effects for the University where the respondent received her B.A (UNIVERSITY BA). Standard error are robust and clustered at the level of the University where the respondent is currently employed. Weights are used. Luca De Benedictis and Michele Di Maio
Summing up (1) Economists within specific Schools of Thought express different opinions on the Causes of the Italian economic decline (with some exceptions)...... even controlling for several possible confounders. Different opinions (we are very conservative): To test whether SofT i is a determinant of economists opinions, we run a likelihood ratio test between the ordinal regression model in equation (1) with and without the variable SofT i. Results indicate that SofT i is significant in 27 out of 40 Causes. The same is done on Policy prescriptions. The likelihood test ratio indicates that SofT i is significant in 12 out of 18 Policy proposals.
Placebo Are Schools of Thought always meaningful? Placebo test 1: we run the same regression having region of birth i instead of opinion i as a dependent variable. Placebo test 2: we run the same regression having migration i instead of opinion i as a dependent variable. SofT i is never significant.
Part IV From Pluralism to Dichotomies
Mainstream vs TRofW Mainstream group Outside-Mainstream group Residual group 1 Mainstream(M) Non-Mainstream (N-M) Neo-classical/Mainstream Institutionalist, Neo-Keynesian, No-school, Eclectic Marxist/Sraffian, Austrian, Regulationist, Behavioralist, Post-Keynesian, Evolutionary 2 MainstreamLarge (ML) HeterodoxJEL (HJEL) Neo-classical/Mainstream Marxist/Sraffian, Institutionalist, No-school, Eclectic, Neo-Keynesian Evolutionary, Austrian Post-Keynesian, Behavioralist, Regulationist 3 MainstreamLarge (ML) Non-MainstreamLarge (N-ML) Neo-classical/Mainstream Marxist/Sraffian, Institutionalist No-school, Eclectic Neo-Keynesian Evolutionary, Austrian Post-Keynesian, Regulationist Behavioralist
Regressions: Mainstream vs TRofW Causes Policy proposals Dumping and unfair competition Low attraction of FDI Euro EU Commission policy BCE monetary policy Commodities world prices International political situation Small firms low technology Small firms innovation Low risk propensity of entrepreneurs Protection of large firms Physical infrastructure Immaterial infrastructure Low competition Firms cost of bureaucracy Credit constraint North-South divide Mezzogiorno - crime Mezzogiorno - infrastructures Low labor market flexibility Demography Low human capital supply Labor unions behavior Public debt level Public debt policy Low efficiency - Public Administration Low efficiency - Bureaucracy More liberalizations More privatizations Increase Public Administration efficiency Foster SME consortia Strengthen firm-territory link Public Investment in strategic sectors Increase firms' Investments in ITC Funding academic research Increase investments in infrastructures More flexibility in the Labor market Reduce jobs' precariousness Reduce Labor Unions power M vs N-M ML vs HJEL ML vs N-ML -3-2 -1 0 1
Summing up (2) and conclusion Economists within specific Schools of Thought express different opinions on several propositions regarding the Causes of the Italian economic decline and the Policies to adopt. Making bilateral comparisons always having the Mainstream as a reference, we can notice that over the 39 propositions (Causes+Policy) where SofT i is significant, Eclectics show a different opinion with respect to Mainstream 20 times, Institutionalists 10 times, Neo-Keynesians 14 times, Post-Keynesians 20 times, Marxists 26 times, Evolutionary economists 20 times, and the No-school 18 times. From this rough metric one can see that Institutionalists are the most similar to Mainstream economists as far as their opinions on normative issue are concerned. If classified in a dichotomic way, Schools of Thought lose much of the explicative power...... unless if the proposition is characterized by an original polarization of individual opinions. Manicheism is (largely) statistically insignificant.