Lobbying successfully: Interest groups, lobbying coalitions and policy change in the European Union Heike Klüver Postdoctoral Research Fellow Nuffield College, University of Oxford Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 1 / 27
Structure of the Presentation 1 Introduction 2 Theory: Lobbying in coalitions 3 Research design 4 Data analysis 5 Conclusion Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 2 / 27
Introduction Interest groups and the European Union Multiple access points: EU is promising political opportunity structure Number of interest groups: Significant increase 1959: 100 1970: 400 1994: 1674 2001: 2000 Participatory engineering: Participation of interest groups to compensate for the democratic deficit Access Influence: Initiatives have provided broad access, but not all interest groups are able to translate their access into influence (Dür/de Bièvre2007, Persson2007) Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 3 / 27
Introduction Research Question Research question and contribution Research question Why are some interest groups able to influence policy-making in the European Union while others are not? State of the literature: Only few studies: Multitude of hypotheses and contradictory findings Problems: Methodological difficulties in measuring influence Concentration on specific group types or one or just a few policy issues Focus on individual group characteristics Contribution: Coherent theoretical framework New measurement approach Large-n analysis across numerous interest groups and issues Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 4 / 27
Introduction Research Question Research question and contribution Research question Why are some interest groups able to influence policy-making in the European Union while others are not? State of the literature: Only few studies: Multitude of hypotheses and contradictory findings Problems: Methodological difficulties in measuring influence Concentration on specific group types or one or just a few policy issues Focus on individual group characteristics Contribution: Coherent theoretical framework New measurement approach Large-n analysis across numerous interest groups and issues Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 4 / 27
Theoretical model Assumptions and propositions Theoretical model Starting point: Assumptions about actor preferences and logic of action (rationalist) Exchange model: Lobbying is conceptualized as an exchange relationship Exchange goods: Exchange goods are deduced from actor preferences taking into account the institutional setting Lobbying coalitions: The exchange of goods at the level of lobbying coalitions is decisive Issue context matters: The intensity of the exchange relationship varies with the issue context Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 5 / 27
Theoretical model Supply side: Interest groups Supply side: Interest groups Type of objective Associations Companies 1. Basic interest Survival Survival 2. Role-specific interest Acquisition of members Maximizing profitability 3. Means preference 1. Maximizing political influence 1. Acquisition of customers 2. Provision of services 2. Reducing costs 3. Maximizing political influence 4. Policy preference Policy position concerning Policy position concerning concrete policy initiative concrete policy initiative Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 6 / 27
Theoretical model Demand side: European institutions Demand side: European institutions Type of objective Commission MEPs Governments 1. Basic interest Survival Survival Survival 2. Role-specific interest Successful proposals Reelection Reelection 3. Means preference 1. Acquisition of policy-relevant information 2. Obtaining citizen support 3. Gaining support of powerful market actors 4. Policy preference Policy position concerning concrete policy initiative Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 7 / 27
Theoretical model The need for information The need for information European Commission Monopoly of legislative initiative Highly understaffed MEPs Evaluate consequences of proposal for their electorate Highly understaffed with extremely busy agenda National governments National ministries consult interest groups about stakeholder positions Interest groups provide expertise to preparatory Council bodies Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 8 / 27
Theoretical model The need for citizen support The need for citizen support MEPs and national governments Need support of voters to avoid electoral punishment Avoiding to be blamed by other actors for not listening to citizen demands European Commission Can exploit electoral dependence of MEPs and governments by gathering citizen support Can add legitimacy to its proposals by securing citizen support Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 9 / 27
Theoretical model The need for market power The need for market power MEPs and national governments Economic voting: Vote choice is driven by economic motives Governments and MEPs therefore secure the support of powerful market actors European Commission Exploits the dependence of MEPs and governments on market players Strategically forms alliances with market actors Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 10 / 27
Theoretical model Lobbying coalitions Lobbying coalitions Lobbying Coalition A Lobbying Coalition B IG1 IG2 IG3 IG4 LEG2 LEG1 IG5 IG6 Issue Dimension IG = Interest Group, LEG = Legislator Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 11 / 27
European Institutions Theoretical model Issue Context Hypotheses Lobbying Coalitions Theoretical model INFLUENCE INFORMATION CITIZEN SUPPORT MARKET POWER Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 12 / 27
Research design Influence conceptualization Influence conceptualization Outcome Convergence of policy output with actor s policy preference Reason Influence Luck Explanation Actor properties Factors not related to actor Observable implication Systematic effect of actor properties No systematic effect of actor properties Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 13 / 27
STAGE ACTORS Research design Influence measurement Influence measurement POLICY FORMULATION DECISION-MAKING a b c 1 2 4.5 5 Draft Proposal Commission Proposal Interest Group Legislative Act COMM COUNCIL, EP, COMM Policy formulation: IG influential as b < a + b Decision-making: IG influential as c < a + b Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 14 / 27
Research design Measuring preferences Measuring preferences Quantitative text analysis Technique: Wordfish (Slapin/Proksch AJPS 2008) Idea: Estimating policy positions based on relative word frequencies Advantage: No prior information about texts is required Textual data Idea: Using Commission Consultations to measure policy preferences Interest groups: Submissions in Commission consultations Legislative output: Consultation paper, EP summaries of legislative proposal and final legislative act Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 15 / 27
Research design Example Example: Proposal on CO 2 emissions from cars Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 16 / 27
Research design Dataset and operationalization Dataset and operationalization Dataset Issues: 56 policy proposals Interest groups: 2,696 associations and companies Data structure: Hierarchical dataset Multilevel modeling Operationalization of explanatory variables Issues: Prelex, EurLex, Policy proposals, Text analysis Interest groups: Website coding, Survey, Consultation submissions Coalitions: Aggregation of interest group characteristics Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 17 / 27
Data analysis Policy formulation Policy formulation: Multilevel analysis Variables Information supply Citizen support Market power Odds rat. Std. err. Odds rat. Std. err. Odds rat. Std. err. Fixed effects Lobbying coalition level Rel. information supply 1.027*** 0.003 Rel. citizen support 1.047*** 0.005 Rel. market power 1.047*** 0.007 Controls: Interest group level Type: Sectional group 1.062 0.125 1.215 0.295 1.045 0.278 Type: Cause group 0.924 0.139 1.021 0.302 Rel. IG characteristic 0.985 0.023 0.971*** 0.010 0.989 0.012 Controls: Issue level Salience 1.321 0.476 1.104 0.511 1.531 0.653 Complexity 1.452 0.446 1.508 0.580 1.553 0.529 Conflict 0.409 0.298 0.530 0.491 0.531 0.446 Existence of EU SQ 0.839 0.378 0.971 0.549 1.139 0.586 Legislative procedure: Codec. 2.788 2.479 5.995 7.722 1.173 1.191 Voting rule: QM 2.718 3.149 4.037 6.483 11.993 19.421 Member state support 0.979*** 0.005 0.990 0.009 0.989 0.012 Random effects Issue level variance 1.930 2.503 1.711 Model fit N / Issues 2696 / 56 1066 / 54 594 / 56 Log likelihood -1539-539 -339 AIC 3104 1104 701 BIC 3181 1168 754 LR Test, Prob > Chi2 0.000 0.005 0.187 ***p 0.01,**p 0.05,*p 0.10 Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 18 / 27
Data analysis Issue context: Policy formulation Interaction between complexity and information supply Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 19 / 27
Data analysis Decision-Making Decision-Making: Multilevel analysis Variables Information supply Citizen support Market power Odds rat. Std. err. Odds rat. Std. err. Odds rat. Std. err. Fixed effects Lobbying coalition characteristics Rel. information supply 1.009*** 0.004 Rel. citizen support 1.016*** 0.005 Rel. market power 1.014** 0.007 Controls: Interest group level Type: Sectional group 0.993 0.152 1.098 0.345 1.079 0.363 Type: Cause group 1.159 0.225 1.505 0.528 Rel. IG characteristic 1.038 0.027 0.988 0.011 0.999 0.013 Controls: Issue level Salience 2.319* 1.018 1.983 1.077 1.884 0.857 Complexity 1.138 0.413 1.339 0.569 1.366 0.460 Conflict 0.430 0.385 0.337 0.365 0.590 0.522 Existence of EU SQ 0.685 0.361 0.616 0.386 0.530 0.271 Legislative procedure: Codecision 2.877 2.852 1.270 1.685 3.928 4.044 Voting rule: QM 1.248 1.839 2.870 5.135 0.582 0.985 Member state support 1.049*** 0.008 1.051*** 0.013 1.037*** 0.014 Random effects Issue level variance 2.140 2.516 1.248 Model fit N / Issues 1793 / 42 717 / 40 389 / 42 Log likelihood -940-364 -213 AIC 1907 753 449 BIC 1978 813 497 LR Test, Prob > Chi2 0.000 0.004 0.047 ***p 0.01,**p 0.05,*p 0.10 Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 20 / 27
Data analysis Issue context: Decision-Making Interaction between complexity and information supply Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 21 / 27
Data analysis Free-riding Coalition effects: Free-riding? Effect Effect of lobbying coalitions Explanation Sum of characteristics of all coalition members Characteristics of few powerful groups Oberservable Implications 1. Coalition characteristics: statistically significant 2. Characteristics of powerful groups: not statistically significant 1. Coalition characteristics: statistically significant 2. Characteristics of powerful groups: statistically significant Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 22 / 27
Data analysis Free-riding Analysis of strong groups (Policy formulation) Variables Information Citizen support Market power Relative information supply 0.981 (0.037) Relative citizen support 0.982* (0.010) Relative market power 0.993 (0.009) Random effects Issue level variance 3.174 3.238 0.093 Model fit N / Issues 456 / 56 428 / 54 139 / 56 Log likelihood -268-229 -96 AIC 542 463 197 BIC 554 475 206 LR Test, Prob > Chi2 0.610 0.067 0.427 Pearson s r -0.003-0.175-0.071 ***p 0.01,**p 0.05,*p 0.10 Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 23 / 27
Data analysis Free-riding Analysis of strong groups (Decision-making) Variables Information Citizen support Market power Relative information supply 0.983 (0.048) Relative citizen support 0.995 (0.011) Relative market power 0.982 (0.013) Random effects Issue level variance 5.054 4.469 0.446 Model fit N / Issues 316 / 42 356 / 40 94 / 42 Log likelihood -171-178 -63 AIC 348 361 133 BIC 359 373 140 LR Test, Prob > Chi2 0.717 0.634 0.157 Pearson s r -0.075-0.120-0.166 ***p 0.01,**p 0.05,*p 0.10 Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 24 / 27
Conclusion Findings Findings Exchange relationship: EU institutions exchange influence for information, citizen support and market power Lobbying is a collective enterprise: Provision of these goods on the lobbying coalition matters Coalition effects: No free-riding on few powerful groups Context matters: Effect of information supply varies with issue complexity The earlier, the better: Lobbying during the policy formulation stage is on average more successful than during the decision-making stage Electoral resources most important: Citizen support and market power have a stronger effect than information supply Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 25 / 27
Conclusion Implications Implications Explaining policy outcomes Previous explanations: Legislative politics scholars predict policy outcomes solely drawing on preferences and bargaining power of EU institutions Findings: Interest group pressure has an important effect on EU policy-making Democratic legitimacy of the EU Output legitimacy: Interest groups need to supply information to affect policy-making Input legitimacy: Biased representation, but business interests are not favored over citizen interests Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 26 / 27
Conclusion Implications Thank you for your attention! Heike Klüver (University of Oxford) Copenhagen Business School Seminar 10 May 2011 27 / 27