Environmental and Resource Economics

, Volume 68, Issue 2, pp 441–444 | Cite as

Erratum to: Renewable Energy Policies and Technological Innovation: Evidence Based on Patent Counts

Erratum
  • 547 Downloads

1 Erratum to: Environ Resource Econ (2010) 45:133–155 DOI 10.1007/s10640-009-9309-1

While replicating the study of Johnstone et al. (2010), Bruns and Kalthaus (2017, forthcoming) noted that a dummy variable for one country (New Zealand) is missing in the reported regressions for solar patents.1 The authors have re-estimated the models including the dummy variable, and the corrected results are reported alongside the original results in the table below. The first part of the table corresponds to the results presented in Table 5 in the original paper. The second part of the table corresponds to the results presented in Table 6, where the policy dummies other than those of primary interest (feed-in tariffs and renewable energy certificates) are combined into a single variable to avoid potential multicollinearity. The third part of the table corresponds to the results presented in Table 7, where we also drop the Kyoto Protocol dummy variable.
Table 1

Corrected solar energy results for Johnstone et al. (2010)

 

Table 5

Table 6

Table 7

Corrected

Original

Corrected

Original

Corrected

Original

Electricity price

2.299

14.140***

2.890

16.279***

3.272

14.699***

(0.512)

(0.000)

(0.400)

(0.000)

(0.340)

(0.000)

Growth of electricity cons.

\(-\)0.004

0.012

\(-\)0.005

0.017

\(-\)0.007

0.013

(0.773)

(0.452)

(0.662)

(0.329)

(0.602)

(0.459)

TOTAL EPO filings

\(-\)0.008

0.062***

\(-\)0.004

0.091***

0.035**

0.100***

(0.688)

(0.000)

(0.825)

(0.000)

(0.009)

(0.000)

Specific R&D expenditures

1.201**

3.224***

1.147**

4.208***

1.789***

4.219***

(0.007)

(0.000)

(0.004)

(0.000)

(0.000)

(0.000)

Feed-in tariff levels

0.016

0.042***

0.019*

0.038***

0.031***

0.042***

(0.063)

(0.000)

(0.012)

(0.000)

(0.000)

(0.000)

REC targets

0.002

\(-\)0.028

0.002

0.051

0.139

0.122

(0.984)

(0.778)

(0.985)

(0.501)

(0.160)

(0.126)

Kyoto protocol

0.647***

0.381***

0.660***

0.329**

  

(0.000)

(0.001)

(0.000)

(0.003)

  

Investment incentives

0.249*

0.228

    

(0.021)

(0.083)

    

Tax measures

\(-\)0.223

0.140

    

(0.071)

(0.307)

    

Guaranteed price

\(-\)0.050

0.899***

    

(0.838)

(0.000)

    

Voluntary programs

0.072

0.009

    

(0.627)

(0.952)

    

Obligations

0.123

\(-\)0.027

    

(0.315)

(0.841)

    

Other renewable policies 

  

0.226*

0.191

0.312**

0.256*

  

(0.038)

(0.122)

(0.004)

(0.033)

N

418

418

418

418

418

418

Log-likelihood

\(-\)759.85

\(-\)794.67

\(-\)761.93

\(-\)807.08

\(-\)780.64

\(-\)810.90

\(\chi ^2\)

2496.91

1840.04

2376.70

1591.83

2071.55

1702.17

(\(p > \chi ^2\))

0.00

0.00

0.00

0.00

0.00

0.00

Table numbers refer to tables in the original publication. Estimation using negative binomial estimation.

p Values in parentheses, based on robust standard errors

* \(p<\) 0.05; ** \(p<\) 0.01; *** \(p<\) 0.001

Given the focus of the paper on the role of feed-in tariffs and renewable energy certificates on patented invention of different renewable energy sources, it is important to note that the variable capturing feed-in tariff levels is now insignificant at the 5% level for the first model estimated (the p value is now 0.063). However, it remains significant when the other policy dummy variables are combined into a single variable to avoid multicollinearity (Tables 6, 7 in the original paper). The inclusion of the missing dummy variable also results in the loss of statistical significance and a much lower magnitude for the coefficient for electricity price. The change in this result is now consistent with the result for electricity prices for the other technologies reported in the paper, and is in line with results from other studies (e.g., Nesta et al. 2014).

In addition, results for two policies represented only by dummy variables change: Investment incentives now have a statistically significant impact in the first model estimated, whereas guaranteed prices do not. The results remain similar when combining all other policies into a single dummy variable for Tables 6 and 7, with a slight gain in precision. Total EPO patent filings are no longer significant in the specifications corresponding with Tables 5 and 6, which is now in line with the results presented for renewables in general. As this variable is simply a control for cross-country differences in patenting behavior, this change suggests that fully capturing all country fixed effects is sufficient to control for cross-country patenting differences.

Footnotes

  1. 1.

    The authors gratefully acknowledge Stephan Bruns and Martin Kalthaus for bringing this to our attention. The authors have verified that the results for all other technologies reported in the original paper are correct.

References

  1. Bruns SB, Kalthaus M (2017) Flexibility in the selection of patent counts: implications for \(p\)-hacking and policy recommendations. Mimeo. http://www.stephanbbruns.de/fileadmin/user_upload/Bruns_Kalthaus_2017.pdf
  2. Nesta L, Vona F, Nicolli F (2014) Environmental policies, competition and innovation in renewable energy. J Environ Econ Manag 67:396–411CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2017

Authors and Affiliations

  1. 1.Structural Policy DivisionOECD Directorate for Science, Technology and InnovationParis Cedex 16France
  2. 2.Environmental Performance and Information DivisionOECD Environment DirectorateParis Cedex 16France
  3. 3.The Maxwell SchoolSyracuse UniversitySyracuseUSA
  4. 4.NBERCambridgeUSA

Personalised recommendations