Title page 1

Contents 7

Abstract 5

Acknowledgements 6

Executive Summary 9

1. Introduction 11

2. Systematic evidence synthesis 14

Examples of meta-analyses in different fields 15

3. Data 18

Study characteristics and methodology 18

Policy-design features 19

Contextual country-level factors 19

4. Descriptive statistics 21

5. Meta-analysis methodology 24

Random effects meta-analysis 24

Bayesian Model Averaging 25

6. Results 26

Random effects model results 26

Bayesian model averaging estimates 29

7. Discussion and conclusion 32

References 34

Annex A. Examples of meta-analyses in different fields 42

Health sciences 42

Road safety 43

Education 43

Agriculture 44

Labour markets 44

Annex B. Standard error conversion methodology 45

Annex C. Policy amendments 46

Annex D. Descriptive statistics 47

Annex E. Tables of main estimation results 50

Annex F. Robustness checks 56

Table 1. Examples of evidence syntheses across fields 15

Table 2. Bayesian Model Averaging results 30

Figure 1. Systematic evidence synthesis and its use in the policymaking process 12

Figure 2. Average effects by study characteristics 22

Figure 3. Average effects by policy-design features 23

Figure 4. Coefficient plot of random effects meta-analysis 26

Figure 5. Coefficient plot of random effects meta-analysis with converted standard errors 29

Boxes 8

Box 1. History of systematic evidence syntheses and meta-analyses 14

Box 2. Examples of how the ongoing and future work of CPEN can inform policymaking 32

Table A B.1. Calculating the implied percentage difference in emissions vs. counterfactual and associated standard errors 45

Table A D.1. Descriptive statistics 47

Table A E.1. Meta regression results with imputed standard errors 50

Table A E.2. Random effects meta-analysis results with converted standard errors 53

Table A E.3. BMA analysis with country-by-instrument dummies 54

Table A F.1. Meta-regression model with significance dummy 56

Table A F.2. Robustness check with additional macro-economic controls 57

Table A F.3. Robustness check with Knapp-Hartung and truncated Knapp-Hartung adjustments to the standard errors 58

Figure A D.1. Correlation between study effects and number of policy amendments 48

Figure A D.2. Association between estimated effects and macroeconomic variables 49

Appendix Boxes 8

Box A C.1. Example of policy amendments considered 46