Strategic compromise, policy bundling and interest group power
2.4. EMPIRICAL CASE STUDY: CALIFORNIA, ILLINOIS AND TEXAS
Table2.6:ReformBundling (1)(2)(3) Dep:Allreforms Interaction:AllreformsDep:Allreforms Interaction:School&teacherreformsDep:School&teacherreforms Interaction:School&teacherreforms CoefficientsMarginsCoefficientsMarginsCoefficientsMargins PanelA.Dependentvariable:Qualityreformsinfinaldraft Teachersshareofpopulation-10.64**-9.927**-10.47**-9.827*-11.15**-10.54** [-19.11,-2.162][-18.24,-1.614][-18.82,-2.117][-18.16,-1.491][-19.04,-3.252][-18.41,-2.675] Accessreformsinfirstdraft-0.111*-0.00779-0.151*0.0399-0.1320.0475** [-0.211,-0.0109][-0.0405,0.0249][-0.298,-0.00332][-0.000550,0.0803][-0.266,0.00129][0.00892,0.0862] Teachersshareofpopulation×Accessreformsinfirstdraft10.75*19.86*18.75** [0.445,21.06][2.931,36.78][3.422,34.07] ControlsYesYesYesYesYesYes YearandStateFEYesYesYesYesYesYes Observations682682682682682682 R-squared0.06530.06570.0668 p-valueofH0:Teachershare+Interaction=00.98690.40200.4522 PanelB.Dependentvariable:Accessreformsinfinaldraft Teachersshareofpopulation0.4830.7780.5360.709-3.534-3.394 [-8.555,9.520][-8.062,9.619][-8.461,9.533][-8.125,9.542][-8.105,1.037][-7.876,1.088] Qualityreformsinfirstdraft-0.0802-0.0314-0.0807-0.0454*-0.0398-0.0111 [-0.200,0.0398][-0.0668,0.00402][-0.224,0.0625][-0.0857,-0.00506][-0.100,0.0206][-0.0352,0.0130] Teachersshareofpopulation×Qualityreformsinfirstdraft5.0883.6822.987 [-7.308,17.48][-11.55,18.92][-3.243,9.217] ControlsYesYesYesYesYesYes YearandStateFEYesYesYesYesYesYes Observations682682682682682682 R-squared0.03390.03460.0396 p-valueofH0:Teachershare+Interaction=00.47630.65400.8878
teachers are more sensitive to bundling when the reforms that are bundled into the bill are more targeted to them. The sum of the coefficients of teacher share and the interaction in these regressions is also insignificant (with ap-value of 0.04). Having access reforms in the first draft essentially nullifies the negative influence of teachers on quality reforms. One can see this relationship more clearly in Figure 2.1, where the average marginal effect of teachers strength is plotted against access reforms in the first draft for the regressions with all reforms as the dependent variable. A couple of things can be observed from this figure. First, when access reforms are close to zero, the marginal effect of teacher share is negative and significant at 10%.
Second, as access reforms in the first draft increases, this negative relationship between interest group strength and quality reforms weakens. This can be gleaned from the upward-sloping marginal effects curve. Furthermore, at some threshold level of access reforms, teachers strength no longer has a significant impact on the amount of quality reforms in an enacted bill. Observe that the threshold level of access is closer to zero when school and teacher access reforms are bundled with quality. Such a result suggests that bundling access with quality weakens the opposition to the bill, particularly so when the bundled reforms affect them more directly.
When it comes to the question of whether bundling quality in the first draft influences the amount of enacted access reforms in the final draft, Panel B of Table 2.6 reveals no significant effects.
As a robustness check, we test whether our results are sensitive to changes in the proxy for interest group strength. Up to this point, the measure we have been using is the number of full time-equivalent teachers as a proportion of district population. Here, we look at the total counts of full-time equivalent teachers. The three states in our sample however tend to differ substantially in the size of the electoral districts. To correct for this discrepancy we standardize the number of teachers using the state-specific distributions. Estimating equation (2.9) with state-standardized teacher counts results in Table 2.7. The coefficients reveal that one standard deviation increase in the number of teachers reduces quality reforms in a bill by about 3 percentage points. The marginal effects of teachers remain significant after using the new teacher variable. However, the interaction term is now different from zero only when teacher counts are interacted with school and teacher reforms, and are now highly significant. The counteracting effect of access only works if the reforms being bundled are more closely linked to their welfare.
One can also observe that on the access equations of Panel B, the estimate of the interaction term has become positive and significant at 5% when the final draft includes all reforms. Indeed, for these regressions, the total effect of teachers on access reforms (the sum of the coefficient of teachers and the interaction) is now positive and statistically different from zero (withp-values of around 0.04 to 0.05). Figure 2.2 shows how the effect of teachers on access reforms varies with quality reforms. In contrast to Figure 2.1, teachers have a statistically significant impact on access only for large enough levels of quality reforms in the first draft. If the bill has little to no quality reforms, having more teachers in the district has no effect on access reforms. However,
2.4. EMPIRICAL CASE STUDY: CALIFORNIA, ILLINOIS AND TEXAS
Table2.7:RobustnessChecks (1)(2)(3) Dep:Allreforms Interaction:AllreformsDep:Allreforms Interaction:School&teacherreformsDep:School&teacherreforms Interaction:School&teacherreforms CoefficientsMarginsCoefficientsMarginsCoefficientsMargins PanelA.Dependentvariable:Qualityreformsinfinaldraft State-standardizedteachersindistrict-0.0323*-0.0308*-0.0325**-0.0316*-0.0391**-0.0383** [-0.0594,-0.00512][-0.0579,-0.00375][-0.0597,-0.00531][-0.0587,-0.00446][-0.0656,-0.0127][-0.0647,-0.0119] Accessreformsinfirstdraft-0.0183-0.01800.005940.006300.01610.0164 [-0.0445,0.00790][-0.0443,0.00827][-0.0282,0.0401][-0.0279,0.0405][-0.0170,0.0492][-0.0167,0.0495] State-standardizedteachersindistrict×Accessreformsinfirstdraft0.02140.0281***0.0262*** [-0.000679,0.0434][0.0106,0.0456][0.00983,0.0425] ControlsYesYesYesYesYesYes YearandStateFEYesYesYesYesYesYes Observations682682682682682682 R-squared0.06470.06490.0676 p-valueofH0:Teachershare+Interaction=00.59910.81440.4756 PanelB.Dependentvariable:Accessreformsinfinaldraft State-standardizedteachersindistrict-0.00426-0.000793-0.00381-0.000670-0.00513-0.00374 [-0.0330,0.0245][-0.0291,0.0275][-0.0323,0.0247][-0.0289,0.0276][-0.0204,0.0102][-0.0190,0.0115] Qualityreformsinfirstdraft-0.0555***-0.0548***-0.0741***-0.0733***-0.0246**-0.0242** [-0.0894,-0.0216][-0.0886,-0.0209][-0.112,-0.0361][-0.111,-0.0354][-0.0422,-0.00709][-0.0418,-0.00667] State-standardizedteachersindistrict×Qualityreformsinfirstdraft0.0598**0.0667**0.0297 [0.0165,0.103][0.0202,0.113][-0.00217,0.0617] ControlsYesYesYesYesYesYes YearandStateFEYesYesYesYesYesYes Observations682682682682682682 R-squared0.03640.03720.0397 p-valueofH0:Teachershare+Interaction=00.05030.04190.2391
-.10.1.2.3 Marginal effect of state-standardized teacher counts
0 .5 1 1.5 2 2.5
Quality reforms in first draft Note: 90% confidence intervals
Dependent: Access reforms in final draft
-.10.1.2.3
Marginal effect of state-standardized teacher counts
0 .5 1 1.5 2 2.5
School & teacher quality reforms in first draft Note: 90% confidence intervals
Dependent: Access reforms in final draft
Figure 2.2: Marginal effect of state-standardized teacher counts on access reforms the effect of teachers on access reforms becomes positive and significant as quality reforms in the first draft increases. This could be indicative of the fact that as the interest group becomes stronger, having more quality requires bundling of more access for the bill to get enacted. Such an effect, however, is not robust to changes in the measure of interest group strength.
2.5 Conclusion
In this paper we set up a game-theoretic framework to explain policy formation under multiple reform dimensions. A policy-oriented government can endogenously propose a bundle of two policies which is then voted on in the legislature. We model this legislative process as a contest between the government and an interest group and find that the government compromises on his reform proposal to appease the opposition. In effect, he uses the two policy dimensions as tools of bargaining. The government proposes more of what the interest group supports and less of what it opposes. We also find that strategic bundling of policies may occur when the government prefers the status quo for one policy but proposes a positive value in equilibrium, to make the opposition less aggressive.
The education setting lends well to this theoretical framework because it has one well-organized interest group, the teachers unions, and many different kinds of education policies. Using legislative data on California, Illinois and Texas education bills from 2008-2013, we find that stronger opposition is associated with less quality reforms. Moreover, as predicted by the model, when bundling access reforms together with quality, the negative effect that we find is counteracted.
An extension of our analysis would be to see how our results hold in a generalized setting with more than one interest group andn-policy dimensions.
REFERENCES
References
Becker, G. S. (1983). A Theory of Competition Among Pressure Groups for Political Influence.
The Quarterly Journal of Economics, 98(3):371–400.
Bellani, L. and Scervini, F. (2015). Heterogeneous preferences and in-kind redistribution: Theory and evidence. European Economic Review, 78:196 – 219.
Berthold, M. R., Cebron, N., Dill, F., Gabriel, T. R., Kötter, T., Meinl, T., Ohl, P., Sieb, C., Thiel, K., and Wiswedel, B. (2007). KNIME: The Konstanz Information Miner. InStudies in Classification, Data Analysis, and Knowledge Organization (GfKL 2007). Springer.
Chaturvedi, A. and Glazer, A. (2005). Competitive proposals of policies by lobbies. Mimeo.
Corrales, J. (1999). The politics of education reform: Bolstering the supply and demand;
Overcoming institutional blocks. InThe Education Reform and Management Series, volume 2.
World Bank.
Cotton, C. and Dellis, A. (2016). Information lobbying and agenda distortion. Queen’s Economics Department Working Paper No. 1348.
De Donder, P., Le Breton, M., and Peluso, E. (2012). Majority voting in multidimensional policy spaces: Kramer-Shepsle versus Stackelberg. Journal of Public Economic Theory, 14:pp.
879�909.
Epstein, G. S. and Hefeker, C. (2003). Lobbying contests with alternative instruments.Economics of Governance, 4(1):81–89.
Epstein, G. S. and Nitzan, S. (2004). Strategic restraint in contests. European Economic Review, 48(1):201–210.
Epstein, G. S. and Nitzan, S. (2006a). Lobbying and compromise. Public Choice, 126:317–330.
Epstein, G. S. and Nitzan, S. (2006b). The politics of randomness. Social Choice and Welfare, 27:423–433.
Fabella, V. M. (2016). Political-economic determinants of education reform: Evidence on interest groups and student outcomes. European Journal of Political Economy. Forthcoming.
Felli, L. and Merlo, A. (2006). Endogenous Lobbying. Journal of the European Economic Association, 4(1):180–215.
Finger, L. (2016). Understanding Interest Group Influence: Teachers’ Unions and Education Reform Policymaking. Mimeo.
Glazer, A., Gradstein, M., and Konrad, K. A. (1998). The electoral politics of extreme policies.
The Economic Journal, 108(November):1677–1685.
Grindle, M. S. (2004).Despite the odds: The contentious politics of education reform. Princeton University Press.
Grossman, G. M. and Helpman, E. (1996). Electoral Competition and Special Interest Politics.
The Review of Economic Studies, 63(2):265–286.
Hillman, A. L. and Ursprung, H. W. (1988). Domestic Politics, Foreign Interests, and Interna-tional Trade Policy. The American Economic Review, 78(4):729–745.
Kang, K. (2014). Policy influence and private returns from lobbying in the energy sector. Mimeo.
Kang, K. (2016). Policy influence and private returns from lobbying in the energy sector.
Working paper. Review of Economic Studies, 83(1):269–305.
Lake, J. (2015). Revisiting the link between {PAC} contributions and lobbying expenditures.
European Journal of Political Economy, 37:86–101.
List, J. A. and Sturm, D. M. (2006). How Elections Matter: Theory and Evidence from Environmental Policy. The Quarterly Journal of Economics, 121(4):1249–1281.
McKelvey, R. D. (1976). Intransitivities in multidimensional voting models and some implica-tions for agenda control. Journal of Economic Theory, 12(3):472 – 482.
McKelvey, R. D. and Patty, J. W. (2006). A theory of voting in large elections. Games and Economic Behavior, 57(1):155 – 180.
Münster, J. (2006). Lobbying contests with endogenous policy proposals. Economics & Politics, 18(3):389–397.
Moe, T. M. (2012). Teachers Unions and American Education Reform: The Politics of Blocking.
The Forum, 10(1).
National Conference of State Legislatures (2014). Education bill tracking database.
Segal, J. A., Cameron, C. M., and Cover, A. D. (1992). A spatial model of roll call voting: Sena-tors, constituents, presidents, and interest groups in Supreme Court confirmations. American Journal of Political Science, 36(1).
Skaperdas, S. and Vaidya, S. (2012). Persuasion as a contest. Economic Theory, 51:465–486.
REFERENCES Tullock, G. (1980). Efficient rent seeking. In James M. Buchanan, Robert D. Tollison, G. T.,
editor,Toward a theory of the rent-seeking society, pages 97–112. College Station, TX: Texas A&M University Press.
Winkler, A. M., Scull, J., and Zeehandelaar, D. (2012). How Strong Are U.S. Teacher Unions?
A State-by-State Comparison. Technical report, Thomas B. Fordham Institute.