External shocks and labor market reforms in autocracies and democracies: Evidence from oil price windfalls

Brückner, Markus; Ciminelli, Gabriele; Loayza, Norman Working Paper External shocks and labor market reforms in autocracies and democracies: Evidence from oil price windfalls ADB Economics Working Paper Series, No. 752 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Brückner, Markus; Ciminelli, Gabriele; Loayza, Norman (2024) : External shocks and labor market reforms in autocracies and democracies: Evidence from oil price windfalls, ADB Economics Working Paper Series, No. 752, Asian Development Bank (ADB), Manila, https://doi.org/10.22617/WPS240528-2 This Version is available at: https://hdl.handle.net/10419/310364 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org ADB ECONOMICS WORKING PAPER SERIES NO. 752 November 2024 External Shocks and Labor Market Reforms in Autocracies and Democracies Evidence from Oil Price Windfalls This paper explores how oil price windfalls impact labor market regulation across 83 countries from 1970 to 2014. Findings reveal that windfall gains lead to deregulation in autocracies but not democracies, while windfall losses trigger deregulation in democracies. The paper examines how democracies use windfall gains to boost spending, whereas autocracies may focus on rent extraction and increasing economic efficiency. Deregulation in democracies during losses aligns with the crisis-induced reform hypothesis, linked to twin deficits and heightened banking crisis risks. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 69 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. EXTERNAL SHOCKS AND LABOR MARKET REFORMS IN AUTOCRACIES AND DEMOCRACIES EVIDENCE FROM OIL PRICE WINDFALLS Markus Brueckner, Gabriele Ciminelli, and Norman Loayza ASIAN DEVELOPMENT BANK The ADB Economics Working Paper Series presents research in progress to elicit comments and encourage debate on development issues in Asia and the Pacific. The views expressed are those of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent. ADB Economics Working Paper Series Markus Brueckner, Gabriele Ciminelli, and Norman Loayza No. 752 | November 2024 Markus Brueckner ([email protected]) is a professor and the head of the Economics Program in the Research School of Economics, Australian National University. Gabriele Ciminelli ([email protected]) is an economist at the Economic Research and Development Impact Department, Asian Development Bank. Norman Loayza ([email protected]) is the director of the Global Indicators Group of the World Bank. External Shocks and Labor Market Reforms in Autocracies and Democracies: Evidence from Oil Price Windfalls Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) © 2024 Asian Development Bank 6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines Tel +63 2 8632 4444; Fax +63 2 8636 2444 www.adb.org Some rights reserved. Published in 2024. ISSN 2313-6537 (print), 2313-6545 (PDF) Publication Stock No. WPS240528-2 DOI: http://dx.doi.org/10.22617/WPS240528-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies ofthe Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area inthis document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This publication is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound bytheterms of this license. For attribution, translations, adaptations, and permissions, please read the provisions andterms of use at https://www.adb.org/terms-use#openaccess. This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed toanother source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims that arise as a result of your use of the material. Please contact [email protected] if you have questions or comments with respect to content, or if you wish toobtain copyright permission for your intended use that does not fall within these terms, or for permission to use theADB logo. Corrigenda to ADB publications may be found at http://www.adb.org/publications/corrigenda. ABSTRACT We examine the relationship between oil price windfalls and labor market regulation empirically through panel regressions in a sample of 83 countries spanning 1970–2014. We find that oil price windfall gains lead to a deregulation of the labor market in autocracies but have no effects in democracies. Windfall losses instead cause a sizeable deregulation in democracies but have limited effects in autocracies. We then consider possible transmission channels. Democracies appear to redistribute the rents stemming from a positive windfall by increasing government expenditure. Rent extraction and economic efficiency considerations are instead both plausible deregulation drivers following windfall gains in autocracies, as expenditures are not raised, while gross domestic product and employment gradually increase after positive windfalls. Finally, the deregulation following windfall losses in democracies is consistent with the crisis-inducedreform hypothesis, as windfall losses deteriorate the current account and budget balances and increase the probability of a systemic banking crisis. Keywords: oil price, windfalls, labor market, deregulation, political institutions JEL codes: F16, J41, O13, P11, P16, Q02 __________________________ This paper presents the findings and conclusions of the authors and does not necessarily represent the views of the organizations to which they are affiliated. We are grateful for the comments and suggestions received at the IMF-RESDM and the ADB ERDI Seminar Series. I. Introduction The effects of natural resources and external shocks on economic growth and institutions is an important topic in development economics. In this paper, we contribute to the literature by examining specifically the effects of oil price windfalls on labor market institutions. To the best of our knowledge, we are the first to do so. A major hurdle that prevented progress in the past on answering the question of how natural resources affect labor market institutions is the availability of data on the latter for a large set of countries and years. To this end, we use a novel dataset on employment protection legislation from Ciminelli and Furceri (forthcoming). This dataset measures the extent of the restrictions faced by employers when they want to terminate an indefinite employment contract. Our measure of oil price windfalls is an oil export price index, which we construct following conventional practice as the interaction between a country’s average gross domestic product (GDP) share of petroleum exports and the annual growth rate of the international oil price (e.g., Arezki and Brueckner 2012). Excluding from the sample major oil exporters such as the Russian Federation and Saudi Arabia, we are able to estimate plausibly causal effects of oil price windfalls on employment protection regulation. We start considering reduced form estimates. Our main finding is that the effect of oil price windfalls on employment protection legislation depends on political institutions. Oil price windfall gains lead to a significant deregulation of the labor market in autocracies while they do not have any significant effect in democracies. On the other hand, oil price windfall losses lead to a sizeable deregulation of the labor market in democracies but have only limited effects in autocracies. These empirical results are persistent over the medium term and are robust to a variety of model specifications and estimation techniques that are standard in the literature. Moreover, the difference between the effects of positive and negative oil price windfalls is statistically significant. Quantitatively, the effects of oil price windfalls on labor market institutions are important, particularly in democracies. A 1-standard-deviation positive windfall induces a deregulation reform worth about a sixth of the median reform in autocracies, while a 1-standard-deviation negative windfall induces a deregulation reform roughly equal to the median reform in democracies. 2 In the second step, we explore potential transmission channels that might drive our results. Oil is a capital-intensive commodity. According to standard international trade theory (i.e., the Stolper-Samuelson effect), when markets are perfectly competitive an increase in the international oil price should increase the rental rate relative to wages, and more so, the larger the GDP share of oil exports. However, when labor markets are not perfectly competitive, e.g., due to labor market regulation, the predictions from this theory are not straightforward. One possible explanation of our results is that, in autocracies, well-connected capitalist insiders lobby the regime following oil price windfalls to liberalize the labor market to reduce workers’ bargaining power, thereby keeping wage pressures in check and appropriating a larger share of the rent. At the same time, one of the main goals of deregulating the labor market is to give employers more flexibility in deciding the optimal allocation of labor, which should increase efficiency and may attract higher investments. Hence, an alternative interpretation is that efficiency gains materialize when there is a deregulation of the labor market. Such an interpretation would be in line with the view that competitive markets achieve efficient allocation. However, deregulating the labor market can cause employment losses in the short term, as employers can take advantage of the new rules to dismiss the most unproductive workers (Cacciatore and Fiori 2016). Oil price windfalls may then offer a good window of opportunity for governments to implement a labor market deregulation, as implementing such reforms during a boom might reduce its short-run costs. Deregulating the labor market can also lead to a lower bargained wage (Ciminelli, Duval, and Furceri 2022). Hence, some governments, particularly those in countries in which democratic institutions are stronger, may be reluctant to implement a deregulation reform, even after periods of oil price windfalls, to not alienate voters. Deregulation reforms are then delayed until a crisis hits, when they are perceived to be strictly necessary. According to this crisis-induced-reform hypothesis, democratic countries implement reforms that improve efficiency for the society overall but that may have (high) short-term costs during crises, when voters understand that there may be few alternatives to reform (Tommasi and Velasco 1996, Drazen and Grilli 1993). Insofar as oil price windfall losses induce a crisis, they may be conducive to a deregulation reform. 3 We test these potential transmission channels by expanding the analysis on the effects of oil price windfalls to different dependent variables. These additional estimations suggest that political economy and efficiency considerations might both be important in explaining the result that autocratic regimes deregulate the labor market following positive windfalls. We find that positive windfalls generate important rents, and that the magnitude of the labor market deregulation is larger in more autocratic regimes. Taken together, these results reinforce the hypothesis that well-connected capitalists lobby for deregulation to appropriate a larger share of the rents generated by the windfall (or at least prevent that workers successfully bargain for a higher wage). This political economy interpretation is broadly in line with view that there are synergies between economic and political institutions (e.g., Acemoglu and Robinson 2012); and that there is significant rent extraction by capitalists in autocracies but not in democracies (Brueckner 2017). At the same time, we also find that tax revenues and the primary balance increase on impact. GDP and the employment rate increase gradually after a positive windfall in autocracies while they do not increase in democracies. This gradual increase of employment and GDP in autocracies might be the result of the labor market deregulation implemented in response to the windfall. Given these dynamics, we cannot rule out that governments in autocratic regimes understand that positive windfalls provide a good opportunity to liberalize the labor market, thus stimulating employment and improving overall economic efficiency in the medium term, while minimizing short-term costs. While governments in democratic governments might also be aware that passing a structural reform in good times can minimize its short-run costs, they might still be unwilling to pass such reform if it is ostracized by a majority of voters. Indeed, political economy considerations appear to be very important in explaining the response of the labor market regulation index to oil price windfalls in democracies. As in autocracies, in democracies, both oil rents and tax revenues increase on impact following positive windfalls. But, differently from autocracies, government expenditures gradually increase after positive windfalls, suggesting that democratic governments use the higher tax revenues stemming from the windfall to redistribute parts of the oil rents to the wider population. Turning to negative windfalls, we find that these sharply deteriorate the current account and budget balances. The fact that democratic countries implement 4 labor market deregulation after an oil price windfall loss that significantly deteriorates public finances, as well as the country’s external position supports the crisis-inducedreform hypothesis, according to which the political economy of structural reforms in democratic countries is such that reforms are delayed until a crisis hits. We further test the validity of the crisis-induced-reform hypothesis by analyzing whether oil price windfalls affect the probability of experiencing a crisis. We find that negative windfalls substantially increase the probability of a systemic banking crisis in democracies, while there are no effects in other cases. Our paper mainly contributes to the literature on the macroeconomic effects of natural resources and to that examining the drivers of economic reforms. In the 1990s, the mainstream view was that natural resources, in particular oil wealth, is not beneficial for a country’s economic development. Early empirical studies, mostly based on crosscountry regressions and case studies, uncovered a negative relationship between natural resource dependence and economic growth (e.g., Sachs and Warner 1995 and 1999). There were also arguments made and empirical evidence provided, mostly in the form of cross-country regressions and case studies, that natural resource dependence is positively correlated with corruption, weak state capacity, and the likelihood that countries are ruled by autocrats (e.g., Gelb 1990, Karl 1997, Ross 2001). In the 2000s and 2010s the mainstream view that natural resources are a curse changed somewhat. Alexeev and Conrad (2009) showed that when the dependent variable is the level of GDP per capita, as a measure of long-run living standards, as opposed to transitional growth, there is no evidence of a resource curse. Brunnschweiler and Bulte (2008) showed that natural resource abundance, as opposed to natural resource dependence, is not significantly associated with a higher risk of civil conflict and slower economic growth. Brueckner and Ciccone (2010) documented that commodity price windfalls are, on average, associated with faster GDP growth in sub-Saharan Africa. A number of subsequent empirical studies, using vector autoregressions (VARs) and dynamic panel regressions, uncovered that in the short run, commodity price windfalls have positive effects on GDP growth (e.g., Brueckner, Tesei, and Ciccone 2012; Collier and Goderis 2012; Araujo et al. 2014; Fernandez, Schmidt-Grohe, and Uribe 2020) 11 B. Local Projections The specifications in Equations (1) and (2) focus on the short-term impact of oil price windfalls on EPL reforms. To shed more light on the dynamics and persistence of the effects, we employ the local projection method. This method was pioneered by Jordà (2005) and has been widely used thereafter (Auerbach and Gorodnichenko 2012, Jordà and Taylor 2016, Ramey and Zubairy 2018, among many others). Plagborg-Møller and Wolf (2021) show that local projections and VARs estimate the same impulse responses. Olea and Plagborg-Møller (2021) show that lag-augmented local projections yield standard errors that are asymptotically valid. In practice, the local projection method entails estimating the response of the dependent variable at period t+k to the shock at time t directly. We consider a 5-year period, including the year of the shock and the four following ones. Therefore, for each 𝑘𝑘= 0, … ,4, we estimate the following specification: 𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,𝑡𝑡+𝑘𝑘 −𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,𝑡𝑡−1 =𝛾𝛾𝑖𝑖+𝜏𝜏𝑡𝑡+𝛽𝛽𝑘𝑘 𝐴𝐴∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴+𝛽𝛽𝑘𝑘 𝐷𝐷∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷+��𝛿𝛿𝑘𝑘,𝑗𝑗 𝐴𝐴∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐴𝐴+𝛿𝛿𝑘𝑘,𝑗𝑗 𝐷𝐷∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐷𝐷� 𝑘𝑘 𝑗𝑗=1 +∑�𝜌𝜌𝑘𝑘,𝑙𝑙 𝐴𝐴∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐴𝐴+𝜌𝜌𝑘𝑘,𝑙𝑙 𝐷𝐷∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐷𝐷� 2 𝑙𝑙=1 +ϑ𝑘𝑘𝑌𝑌𝑡𝑡−1 +𝜀𝜀𝑖𝑖,𝑡𝑡 (3) where ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴 (∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷) denote positive oil price windfalls in autocracies (democracies), constructed as ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡∗𝐴𝐴𝑖𝑖,𝑡𝑡 (∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡∗(1 −𝐴𝐴𝑖𝑖,𝑡𝑡)) with 𝐴𝐴𝑖𝑖,𝑡𝑡 being a dummy variable taking value equal to 1 when the polity2 score is below 0; 𝑌𝑌𝑡𝑡 is the log of per capita GDP in PPP, included to control for the stage of development, and the rest of the notation is as in Equation (2). The variables in the summation term are forward shocks á la Teulings and Zubanov (2014), included to take into account of oil price windfalls that occur within the 𝑡𝑡+𝑘𝑘 horizon, but that are not captured by the main explanatory variable (∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡), and whose omission may bias the results. The equation is made dynamically complete by 12 including two lags of the windfall variables. The model includes country and time fixed effects, and the estimation is carried out through OLS.3 We estimate the model on the full sample of autocracies and democracies, allowing for different effects of oil price windfalls in these two groups of regimes, rather than splitting the sample in two. This is to maximize the number of observations in each country. The results would be very similar if we were to split the sample in two. The coefficients of interest are the 𝛽𝛽𝑘𝑘 𝐴𝐴s and 𝛽𝛽𝑘𝑘 𝐷𝐷s. They capture the effect of oil price windfalls at time 𝑡𝑡 on the cumulative change in EPL during the 𝑡𝑡+𝑘𝑘 horizon in, respectively, autocracies and democracies. The results are presented in the form of impulse response functions (IRFs), i.e., we plot the estimated coefficients 𝛽𝛽𝑘𝑘 𝐴𝐴 and 𝛽𝛽𝑘𝑘 𝐷𝐷 and their respective 90% confidence bands. In the next step, we leverage on the flexibility of the local projection method to analyze the effects of positive and negative oil price windfalls separately, given that these may have different effects on EPL. We estimate the following regression specification: 𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,𝑡𝑡+𝑘𝑘 −𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,𝑡𝑡−1 =𝛾𝛾𝑖𝑖+𝜏𝜏𝑡𝑡+𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴,𝑃𝑃+𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴,𝑁𝑁+𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷,𝑃𝑃+𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷,𝑁𝑁 +��𝛿𝛿𝑘𝑘,𝑗𝑗 𝐴𝐴,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐴𝐴,𝑃𝑃+𝛿𝛿𝑘𝑘,𝑗𝑗 𝐴𝐴,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐴𝐴,𝑁𝑁+𝛿𝛿𝑘𝑘,𝑗𝑗 𝐷𝐷,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐷𝐷,𝑃𝑃+𝛿𝛿𝑘𝑘,𝑗𝑗 𝐷𝐷,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡+𝑗𝑗 𝐷𝐷,𝑁𝑁� 𝑘𝑘 𝑗𝑗=1 +∑�𝜌𝜌𝑘𝑘,𝑙𝑙 𝐴𝐴,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐴𝐴,𝑃𝑃+𝜌𝜌𝑘𝑘,𝑙𝑙 𝐴𝐴,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐴𝐴,𝑁𝑁+𝜌𝜌𝑘𝑘,𝑙𝑙 𝐷𝐷,𝑃𝑃∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐷𝐷,𝑃𝑃+𝜌𝜌𝑘𝑘,𝑙𝑙 𝐷𝐷,𝑁𝑁∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡−𝑙𝑙 𝐷𝐷,𝑁𝑁� 2 𝑙𝑙=1 +ϑ𝑘𝑘𝑌𝑌𝑡𝑡−1 +𝜀𝜀𝑖𝑖,𝑡𝑡 (4) where ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴,𝑃𝑃 and ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷,𝑃𝑃denote positive windfalls (i.e., gains) in autocracies and democracies, respectively, constructed by interacting the ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴 and ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷 variables from Equation (3) with a dummy variable taking value equal to 1 for positive changes in the oil price index. ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴,𝑁𝑁 and ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷,𝑁𝑁denote negative windfalls (i.e., losses) in autocracies and democracies, respectively, constructed by interacting the ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐴𝐴 and ∆𝑂𝑂𝑂𝑂𝑂𝑂𝑖𝑖,𝑡𝑡 𝐷𝐷 variables from Equation (3) with a dummy variable taking value equal to 0 for negative changes in the oil price index. The coefficients 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃 and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃 measure the 3 We also considered a random effects local projection model through GLS. The results using the two approaches are very similar. 13 cumulative effect of positive oil price windfalls (i.e. gains) at time 𝑡𝑡 on EPL reforms over horizon 𝑡𝑡+𝑘𝑘 in autocracies and democracies, respectively. The coefficients 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁 and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁 measure the cumulative effect of negative oil price windfalls (i.e. losses) at time 𝑡𝑡 on EPL reforms over horizon 𝑡𝑡+𝑘𝑘 in autocracies and democracies, respectively. IV. Baseline Results A. Short-run Effects Table 3 presents the baseline results, obtained estimating Equations (1) and (2). The results are reported in Panels A and B, respectively for the sample of autocracies and democracies. Columns (1) and (2) show results from a specification including only the contemporaneous value of the oil price windfall variable. The estimates in columns (1) and (2) are obtained by estimating random and fixed effects models, respectively. Columns (3) and (4) report results when we model oil price windfalls to affect EPL with a 1-year lag. Columns (5) and (6) show estimates when including both the contemporaneous and the lagged value of the oil price windfall variable in the model. The reported coefficients are normalized to show the effect of a 1-standard-deviation oil price windfall (calculated on the unrestricted sample of autocracies and democracies, equal to about 1.5). Equations (1) and (2) assume that positive and negative oil price windfalls have symmetric effects. For simplicity, we discuss the results referring to positive windfalls. Further below, we allow for asymmetric effects depending on the sign of the windfall. We estimate negative coefficients for oil price windfalls in autocracies, while the coefficients estimated for democracies are not statistically significant. The coefficients estimated for the contemporaneous oil price windfall variable and its 1-year lag in autocracies are about -0.2 and -0.1, respectively. Both these coefficients are statistically significant at either the 95% or 99% confidence level and are broadly consistent across the different specifications and regression models considered. The fact that the two coefficients have the same sign and are both significant suggests that either the cumulative effect of oil price windfalls on labor market regulation increases over time and/or that some of the reform action happens with a lag. 14 Quantitatively, the estimated coefficient of -0.2 on year t oil price windfalls can be interpreted as follows: a 1-standard-deviation oil price windfall leads to a reduction in the EPL index of around 0.2 units, which is equivalent to around 0.1 standard deviations of the one-period change of the EPL index. B. Medium-run Effects In the baseline analysis, we uncovered a negative, short-run, effect of oil price windfalls on the EPL index in autocracies. Next, to analyze the persistence and dynamics of this effect, we use the local projection method. Through the local projection method, we trace out the response of EPL reforms to oil price windfalls over a 5-year horizon (including the year of the shock plus the four following ones). Figure 3 depicts the cumulative impulse response function (IRF) of EPL reforms to a 1-standard-deviation increase in the oil price index. The IRF is derived by estimating Equation (3). Panel A reports results for autocracies (𝛽𝛽𝑘𝑘 𝐴𝐴 coefficients), while Panel B shows results for democracies (𝛽𝛽𝑘𝑘 𝐷𝐷 coefficients). Blue solid lines are point estimates, while dashed red lines denote 90% confidence bands. The IRFs in Figure 3 indicate that the effects of oil price windfalls on EPL reforms are persistent, and even increase, over time. In autocracies, the effects are concentrated at impact and 1 year following the shock. The cumulated effect 1 year after the shock is about -0.3 (significantly different from zero at the 99% confidence level). The effect slightly decreases for the rest of the horizon and becomes statistically insignificant in the medium term. Turning to democracies, we confirm that oil price windfalls do not have any effects on EPL in the short run, but we instead estimate a positive coefficient in the long run. The coefficient that we estimate for the 5-year horizon is equal to about 0.4 and statistically significant at the 90% confidence level. We draw two conclusions from this analysis. First, the effects of oil price windfalls on labor market regulation take time to fully materialize. Second, the cumulative effects over 5 years have the opposite sign in autocracies and democracies, suggesting that different channels might be at play. 15 C. Effects of Positive and Negative Windfalls The analysis carried out so far assumes that positive and negative oil price changes (windfall gains and losses) have symmetric effects on the EPL index. In what follows, we estimate Equation (4) to explore whether positive and negative changes in the oil price index have differential effects. Figure 4 shows the new IRFs. Panels A1 and A2 respectively report the effects of positive and negative windfalls in autocracies (𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃 and 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁 coefficients), while Panels B1 and B2 refer to democracies (𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃 and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁 coefficients). To ease interpretation, the IRFs reported in Panels A2 and B2 are constructed using the negative value of the 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁 and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁 coefficients. The new results indicate that the effects of oil price windfalls estimated from the restricted model in Equation (3) are entirely driven by positive changes in the oil price index in autocracies. That is, oil price windfall gains cause autocratic regimes to implement employment protection deregulation reforms (Panel A1). This effect of positive windfalls is large, statistically significant, and increases throughout the horizon considered. A 1-standard-deviation positive shock induces a reduction of the EPL index worth about 0.6 over a five-year horizon (statistically significant at the 99% confidence level). To put this in context, among autocracies, the median easing reform over a 5-year period is equal to 3.6. A reduction in oil revenues due to a negative price growth also has negative effects on the EPL index—that is, it induces a deregulation—but this effect is not statistically significant (Panel A2). The picture is different for democracies. Positive windfalls do not have any effects, while negative ones lead to an easing of labor market regulation. This reforming action materializes only gradually. A 1-standard-deviation negative windfall is estimated to induce a reduction of the EPL index of about 0.2 on impact. This effect gradually increases over all the horizons considered, to reach about 1.3 4 years after the shock (statistically significant at the 95% confidence level). This amounts to an important reform in the context of democratic countries, just below the median easing reform implemented over a 5-year period. This analysis suggests that positive and negative windfalls have different effects on EPL and that these differential effects differ among autocracies and democracies. We 16 corroborate this asymmetry result by plotting in Figure 5 the difference between the effect of a positive windfall and that of a negative windfall over each year of the horizon considered, using the coefficients estimated from Equation (4). We find that, except at impact, the coefficients estimated for positive and negative windfalls are statistically different from each other at least at the 90% confidence level for both autocracies and democracies. In the next section, we explore potential channels that could explain the differential effects that we estimate. Before that, however, we perform some additional estimations to verify the robustness of our results. Robustness Checks and Alternative Specifications We assess the sensitivity of our results to several different specifications. First, we check that our results are not driven by individual countries and estimate the medium-run effect obtained from the local projection specification allowing for asymmetric effects between positive and negative windfalls (Equation [4], with k=4) dropping one country at a time. Figure 6 shows a scatterplot depicting the new estimates, with the yand x-axis respectively reporting point estimates and p-values. These are close to the full sample baseline and confirm that our results are not driven by individual countries. Most Organization of the Petroleum Exporting Countries (OPEC) members are excluded from the analysis because they are not covered by the EPL database (Angola, Republic of the Congo, Equatorial Guinea, Iran, Iraq, Lybia, Qatar, Saudi Arabia, United Arab Emirates). We further verify that the results are robust to excluding all remaining OPEC members in Figure 7. As another exercise, we check how the estimates are affected under different lag specifications. Specifically, we estimate a model including two lags (t-1 and t-2) of the first difference of the EPL index as well as a model including the first lag (t-1) of the level of the EPL index. The first model accounts for potential serial correlation in the data—if the dependent variable is correlated over time, omitting its lags can lead to serially correlated regression errors and biased and inefficient estimates. The second model introduces an autoregressive component to labor market regulation to capture the possibility that changes in labor market regulation might depend on the past level of regulation. 17 Estimates from these additional specifications are presented in Figure 8. Those obtained controlling for past changes in EPL are virtually identical to the baseline estimates, while those obtained controlling for the initial level of regulation are qualitatively similar but display some quantitative differences. The medium-term effect of positive windfalls in autocracies and of negative windfalls in democracies are slightly attenuated, but the newly estimated coefficients are still significantly different from zero at the 10% interval. The effect of negative windfalls in autocracies becomes larger and not statistically different from zero, a sign that negative windfalls may be an important driver of labor market regulation also in autocracies. We also check that our results are robust to (i) using a random effects rather than a fixed effects model, and (ii) estimating the effects for autocracies and democracies by splitting the sample into two groups, autocracies and democracies. The results from these alternative specifications are very close to our baseline (available upon request). Next, we test whether we obtain similar results if we used a variable measuring the extent of oil discoveries, developed by Cotet and Tsui (2013), rather than the oil price windfall variable as an explanatory variable. The results for oil discoveries are fairly similar to those that we obtain for positive oil price windfalls (Figure 9). Oil discoveries induce a large deregulation of EPL in autocracies whereas they do not have any effects in democracies. Somewhat differently from positive oil price windfalls, the effect of oil discoveries in autocracies is immediate and tends to fade out over time. We also test whether dividing the sample between (i) advanced countries and (ii) emerging markets and developing countries, rather than between autocracies and democracies, give similar results to our baseline, given that the group of democracies contains almost all advanced countries and therefore there is some overlap between the two. We estimate a negative effect of both positive and negative windfalls in emerging and developing countries (Figure 10). However, these effects are rather small and not statistically significant in the case of negative windfalls. For the sample of advanced countries, the response to negative windfalls is similar to the case of democracies, but it is less precisely estimated and weaker. These results suggest that political institutions matter more than economic development in determining the effect of oil price windfalls on labor market regulation. 18 We next explore whether the effects of oil price windfalls change discretely across autocracies and democracies or whether there is a more linear relationship between the degree of autocracy/democracy and windfalls. We are able to do so by leveraging on the polity2 variable. The polity2 variable varies between –10 and 10. More negative (positive) polity2 scores denote stronger autocratic (democratic) institutions. Figure 11 reports medium-run estimates of the asymmetric effects of oil price windfalls in the sample of autocratic and democratic countries as a function of the polity2 variable. These estimates are obtained by estimating an alternative specification of Equation (4), for k=4, which includes on the right-hand side of the estimating equation: (i) the positive and negative oil price windfall variables, (ii) the lagged polity2 score, and (iii) the interaction between the oil price windfall variables and the lagged polity2 score. Forward and lagged values of oil price windfalls are also interacted with the polity2 score. The main result is that the negative effect of positive oil price windfalls in autocracies are larger the more the polity2 variable takes negative values, while the negative response to negative windfalls in democracies is larger the more the polity2 variable takes positive values. In so-called hereditary monarchies (polity2=10), positive windfalls lead to a significant easing of EPL, worth about -1 of the EPL index. The effect for a polity2 score close to 0 is less than half as large and not statistically significant. In regimes that are democratic just on the surface (polity2 score close to 0) the response to negative windfalls is not statistically significant, while in the strongest democratic regimes (polity2=10) the response to a 1-standard-deviation negative windfall is about -1.4 index points 4 years after the windfall and highly statistically significant. We also find that, among autocratic regimes, the less the regime is autocratic the more it responds to a negative windfall by deregulating the labor market (similar to the response of democratic countries). However, the estimated responses to negative windfalls for autocratic countries are not statistically significant for any polity2 score. These results suggest that political economy considerations may be important in driving the responses to oil price windfalls, in both autocracies and democracies. The more a regime is autocratic, the more the capitalist elite is likely to be closely associated with the government and lobby to liberalize the labor market following positive oil price 19 windfalls to appropriate the oil rents. At the same time, the more a government is democratic, the more it will find it politically unappealing to deregulate the labor markets in normal times, thus postponing deregulation in times of crisis. We explore political economy considerations as potential drivers of our results more thoroughly in Section 5. We close this analysis by considering the response of employment protection legislation to other types of windfalls. In particular, we focus on agricultural commodities, which typically generate less rents. Using the dataset assembled by Bazzi and Blattman (2014), who collect price data for 65 commodities and export data for major countries in Africa, the Middle East, Asia, and Latin America, we construct two variables, measuring windfalls for the production of annual and perennial agricultural commodities.4 Next, we estimate alternative specifications of Equation (4), in which we replace our baseline oil price windfall variables with these alternative variables. Figure 10 presents new estimates for the effect of annual agricultural commodity windfalls on employment protection legislation. The effects are null. Effects of perennial agricultural windfalls are also null (available upon request). V. Channels The analysis conducted so far has shown that autocratic regimes reduce the level of labor market regulation following oil price windfall gains, while democratic governments reduce labor regulation following oil price windfall losses. We find negative but not significant effects from windfall losses in autocracies and no effects from windfall gains in democracies. In this section, we explore some potential transmission channels driving these results. We first discuss the potential transmission channels, then explore how windfalls affect other macroeconomic variables and the probability of experiencing a systemic banking crisis, and conclude with a brief case study analysis. 4 These variables exclude price makers. 20 A. Conceptual Framework We focus on two sets of non-mutually exclusive explanations, related to political economy and efficiency considerations. One of the main goals of deregulating the labor market is to give employers more flexibility in deciding the optimal allocation of labor. In this sense, deregulation improves labor market efficiency and may attract higher investments, including from foreign investors. However, there are some caveats. First, deregulating the labor market can cause employment losses in the short term, as employers can take advantage of the new rules to dismiss the most unproductive workers (Cacciatore and Fiori 2016). Second, by reducing workers’ bargaining power, labor market deregulation reduces wages (Ciminelli, Duval, and Furceri 2022).5 Third, and possibly in part due to the effects just described, deregulating the labor market is unpopular (Alesina et al. 2023). Because of these reasons, some governments, particularly those in countries in which democratic institutions are stronger, may be reluctant to implement deregulation reforms and delay them until when they are perceived to be strictly necessary. Analyzing the effect of oil price windfalls on labor market reforms adds another layer of complexity. If oil price windfall gains induce an expansion of economic activity, they may offer a good window of opportunity to implement a labor market reform, as such reforms tend to have less short-run costs if done during good economic times.6 On top of this, political economy considerations may also be relevant. Workers in the oil and related 5 There exist several mechanisms through which a reduction in EPL decreases the surplus of the workers. The earlier macroeconomics literature assumes EPL to have a direct effect on bargaining power (Blanchard et al. 1997, Blanchard and Giavazzi 2003). Hence, easing EPL lowers workers’ bargaining power and results in workers capturing a lower share of the match surplus. Alternatively, EPL can be thought as a wasteful firing cost in a standard Diamond-Mortensen-Pissarides setting (Mortensen and Pissarides 1994). When wages are re(negotiated), the firm gives workers a wage premium for saving on firing costs today but penalizes them for having to pay firing costs in the future (instance.g., Cacciatore and Fiori 2016). That penalty is the discounted value of future firing costs, which is less than the current value due to time preferences. Hence, EPL increases the worker’s share of the match surplus and, in turn, a reform easing EPL decreases it. EPL is also well-known to reduce job turnover (Bentolila and Bertola 1990, Pissarides 2000), which in turn weakens the ability of workers to extract the surplus from a match (Cahuc, PostelVinay, and Robin 2006). Ciminelli et al. (2022) analyze the effect of EPL reforms on the wage level and the labor share and find that EPL deregulation unambiguously lowers the wage and lowers the labor share when labor and capital are relative complement. 6 A labor market deregulation reform implemented during good economic times may lead employers to expand the labor input knowing that they will have more flexibility to reduce it in the future shall business needs change (Duval et al. 2020). 27 The 1999 labor code reform allowed for more flexible employment contracts, easier dismissal procedures, and short-term labor arrangements. Pressure from domestic elites and foreign oil companies played a significant role in shaping the reforms. Oil majors demanded labor flexibility to manage their operations efficiently, particularly in capitalintensive projects like the Tengiz oil field. Their influence led to a labor market model that prioritized cost efficiency, flexibility in hiring and firing, and the widespread use of manpower agencies (Sorbello 2023). The centralization of political power meant that the government could implement labor deregulation with minimal opposition from workers or unions, which were often weakened by top-down control. The EPL indicator of Ciminelli and Furceri (forthcoming) captures the 1999 reform as a decrease of 27 index points (out of a 0–100 scale). This is a large reform, which falls within the upper decile of all deregulation reforms in our sample. The Crisis-Induced 2002 Deregulation Reform in Colombia The case of the labor market deregulation reform decided by the Government of Colombia in December 2002 is very different from that of Kazakhstan just reviewed, as it was implemented in by a democratic government in response to a major economic crisis that greatly increased the unemployment rate. In the years leading up to the reform, the oil price crashed by over 60% over 1997–1998. It later slowly recovered throughout 1999– 2002, reaching its end-1996 level in December 2002, the month of the reform. Given the delay with which reforms are undertaken in democracies in response to negative windfalls (we estimate the peak effect to be 4 years after the windfall), we see the 1997–1998 oil price crash to be the driver of the liberalizing reform. Colombia has been producing oil since the early 20th century and at the end of the 1990s was one of the leading oil producers in Latin America, reaching production of around 700,000 barrels per day of oil in 1999 (Perry and Olivera 2009). A significant portion of Colombia’s oil production was exported, making oil one of the country’s main export commodities and contributing to a substantial part of Colombia’s foreign exchange earnings. As a result of the Asian financial crisis of 1997 and of the Russian debt default of 1998, global demand for oil and other commodities decreased. From its end-1996 peak to its end-1998 trough, the oil price crashed by over 60%. In 1999, Colombia experienced 28 one of its worst-ever economic crises, with the GDP contracting by over 4% and the unemployment rate rising to over 20%, from just over 12% in 1997. The crisis also led to troubles in the financial sector, with local banks facing significant instability and solvency problems. The economic crisis precipitated a sense of urgency to implement economic reforms. By 2002, the unemployment rate was still around 16%, twice as much the level in the mid-1990s. The 2002 labor market reform in Colombia, embodied in Law 789, was motivated by the need to address the country’s high unemployment and labor market informality. Given the narrow focus of the indicator of Ciminelli and Furceri (forthcoming) on employment protection legislation and the fact that only severance payments were reformed in this area, the reform is worth just a 2-index point decrease index. But the reform is considerably more significant if other areas of labor market regulation are considered, as it introduced greater labor market flexibility in hiring and firing by making it easier for employers to lay off workers in temporary contracts, reduced labor costs by modifying overtime pay rules, and incentivized hiring through apprenticeship programs. The success of the reform is debated (Gaviria 2005), but Colombia did see a gradual labor market recovery over the subsequent years, with the unemployment rate slowing to 11% by 2007. VI. Conclusion The role of natural resources in shaping institutions is an important topic in development economics. In this paper, we focused on oil and took a first stab at analyzing the effects of oil price windfalls on labor market regulation. Basic international trade theory suggests that for an oil-exporting country with perfectly competitive factor markets, an increase in the international oil price leads to an increase in the rental rate relative to wages. However, if markets are not competitive, in particular, if there is significant employment protection legislation, workers have the bargaining power to bid up their wages. Combining a novel database of employment protection legislation with information on plausibly exogenous oil price windfalls, we found that the effects of oil price windfalls on labor market institutions are different between autocracies and democracies. Among 29 the former, oil price windfall gains lead to a weakening of employment protection. This effect is persistent over the medium term. On the other hand, oil price windfall gains have no effect on labor market regulation in democratic countries. At the same time, oil price windfall losses lead to substantial job protection deregulation in democracies, while they have limited effects in autocracies. We considered two sets of explanations, one based on political economy and the other on efficiency considerations, as potential drivers of our results. One explanation is that governments in autocratic regimes recognize that there are efficiency gains that materialize when there is a deregulation of the labor market. Such an interpretation would be in line with the view that competitive markets achieve efficient allocation. Given that deregulating the labor market can cause employment losses in the short term, oil price windfalls may offer a good window of opportunity for governments to implement labor market deregulation, as implementing such reforms during good economic times might reduce its short-run costs. The other set of explanations is based on political economy considerations. On the one hand, after an oil windfall, capital owners may lobby the government to loosen up labor market regulation to weaken the bargaining power of workers and appropriate a larger share of the rents stemming from the higher oil price. Given that in autocracies elections either do not take place or do not carry a real threat to the ruling elite, such lobbying activities may be successful. In democratic countries, instead, governing parties face reelection and therefore need to consider how much a weakening of labor market institutions would dampen their reelection prospects. Reforms are then delayed until they are perceived as a last-resort option, which is when a crisis hits. To shed light on these potential transmission channels, we went beyond employment protection and explored the effect of oil price windfalls on a range of other outcome variables. Positive oil price windfalls generate important rents. In democracies, tax revenues go up after the windfall and governments increase expenditures, possibly to redistribute parts of these rents to the wider population. In autocracies, instead, there is no rent redistribution through higher expenditures, but we find that GDP and the employment rate both increase in the medium term after a positive windfall, which is not the case in democracies. These results suggest that, while the response of democracies 30 to a positive windfall seems to be rent redistribution, efficiency, and rent extraction can both explain the deregulation of the labor market following positive windfalls in autocracies. On the other hand, the crisis-induced-reform hypothesis appears to be a valid explanation for the deregulation of the labor market following windfall losses in democracies, as we find negative oil price windfalls to sharply deteriorate the current account and the budget balance and substantially increase the probability of a systemic banking crisis in these countries. Employment protection legislation is just one aspect of labor market institutions. We focused on it for the sake of identification purposes, but the point is more general. Future research should go beyond employment protection and assess the role of natural resources on other institutions affecting the distribution of income between labor and capital. 31 FIGURES AND TABLES Figure 1: Evolution of Employment Protection Legislation over Time Notes: Median and interquartile range of the employment protection legislation index of Ciminelli and Furceri (forthcoming) for the samples of autocratic and democratic countries over the time sample considered in this paper. The index ranges from 0 to 1, with higher values indicating higher employment protection. Source: Authors’ calculations based on data from Ciminelli and Furceri (forthcoming). 32 Figure 2: Frequency of Reforms over Time (as % of total observations) Notes: Frequency (in %) of negative changes (easing), no changes (status quo) and positive changes (tightening) of the employment protection legislation index of Ciminelli and Furceri (forthcoming) for the sample of autocratic and democratic countries and for each year of the analysis. Higher values of the index indicate higher employment protection. Source: Authors’ calculations based on data from Ciminelli and Furceri (forthcoming). 33 Figure 3: Dynamic Effects of Oil Price Windfalls on the Employment Protection Legislation Index Notes: Effect of a 1-standard-deviation oil price windfall on the employment protection legislation index over a 5-year horizon in autocracies (Panel A) and democracies (Panel B). X-axes denote the horizon of the effect, with 0 indicating the year of the windfall, while Y-axes denote the magnitude of the effect. Blue solid lines report the point estimates, while red dashed lines are 90% confidence bands. Estimates are obtained plotting the 𝛽𝛽𝑘𝑘 𝐴𝐴 (Panel A) and 𝛽𝛽𝑘𝑘 𝐷𝐷 (Panel B) coefficients and their standard errors estimated from Equation (3). Source: Authors’ calculations. 34 Figure 4: Asymmetric Effects of Oil Price Windfalls on the Employment Protection Legislation Index Notes: Effect of a 1-standard-deviation negative and positive oil price windfalls on the employment protection legislation index over a 5-year horizon in autocracies and democracies. Estimates are obtained plotting the 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃 and 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁 (Panel A) and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁 (Panel B) coefficients and their standard errors estimated from Equation (4). See notes in Figure 3. Source: Authors’ calculations. 35 Figure 5: Differential Effects of Positive and Negative Oil Price Windfalls Notes: Difference between the coefficients estimated for the effect of positive and negative windfalls on the employment protection legislation index over a 5-year horizon in autocracies and democracies. Estimates are obtained by plotting the 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃-𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁 (Panel A) and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃-𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁 (Panel B) combinations of the coefficients estimated from Equation (4). See notes in Figure 3. Source: Authors’ calculations. 36 Figure 6: Robustness Check on Country Sample Notes: Point estimates (y-axes) and p-values (x-axes) of the medium-term effect of positive and negative oil price windfalls on the employment protection legislation index obtained from a robustness check exercise in which each country is excluded from the sample one at a time. The estimating equation is Equation (4) with k=4. Blue (red) dots report estimates of the 𝛽𝛽𝑘𝑘 𝐴𝐴,𝑃𝑃 (𝛽𝛽𝑘𝑘 𝐴𝐴,𝑁𝑁) (Panel A) and 𝛽𝛽𝑘𝑘 𝐷𝐷,𝑃𝑃(𝛽𝛽𝑘𝑘 𝐷𝐷,𝑁𝑁) (Panel B) coefficients. Source: Authors’ calculations. 43 Table 1: Descriptive Statistics of Employment Protection Legislation Index A. AUTOCRACIES B. DEMOCRACIES Obs. Media n S.d. Min Max Obs. Media n S.d. Min Max Index 974 21.82 10.70 2.17 58.17 2327 23.25 11.48 2.00 58.15 Tightening 46 3.97 7.80 0.01 30.61 171 2.04 6.19 0.01 39.13 Easing 19 -3.70 6.73 -22.31 -0.20 65 -1.48 3.84 -23.33 -0.02 Notes: The table reports descriptive statistics of the employment protection legislation index of Ciminelli and Furceri (forthcoming) for the two samples of autocracies (polity2 score below 0, Panel A) and democracies (polity2 score above 0, Panel B). The row denoted by Index shows statistics for the level of the index. Rows denoted by easing and tightening report statistics for, respectively, negative and positive changes of the index. The index ranges on a scale from 0 to 100, with higher values indicating more employment protection. Source: Authors’ calculations based on data from Ciminelli and Furceri (forthcoming). Table 2: Descriptive Statistics of Oil Price Windfalls A. AUTOCRACIES B. DEMOCRACIES Obs. Median S.d. Min Max Obs. Median S.d. Min Max Positive 594 0.12 1.41 0.00 13.74 1378 0.10 0.83 0.00 12.77 Negative 410 -0.06 1.00 -6.99 0.00 952 0.04 0.55 -7.62 0.00 Notes: The table reports descriptive statistics of oil price windfalls for the two samples of autocracies (polity2 score below 0, Panel A) and democracies (polity2 score above 0, Panel B). Source: Authors’ calculations. 44 Table 3: Effect of Oil Price Windfalls on Changes in Employment Protection Legislation A. AUTOCRACIES (1) (2) (3) (4) (5) (6) Oil windfall -0.20*** -0.23** -0.17** -0.20*** (0.08) (0.09) (0.07) (0.07) Oil windfall, -0.10** -0.10*** -0.10*** -0.11*** lag 1 (0.04) (0.04) (0.04) (0.04) Observations 915 915 884 884 878 878 R-squared 0.08 0.08 0.06 0.06 0.07 0.08 Model RE FE RE FE RE FE B. DEMOCRACIES (1) (2) (3) (4) (5) (6) Oil windfall 0.08 0.05 0.08 0.04 (0.06) (0.06) (0.06) (0.07) Oil windfall, -0.01 -0.04 -0.01 -0.04 lag 1 (0.04) (0.04) (0.04) (0.04) Observations 2,278 2,278 2,243 2,243 2,243 2,243 R-squared 0.05 0.05 0.05 0.05 0.05 0.05 Model RE FE RE FE RE FE Notes: Estimates of the contemporaneous and lagged effects of oil price windfalls on changes of the employment protection legislation index, obtained estimating a random effects (Equation 1) and a fixed effects model (Equation 2) for the sample of autocratic countries (polity2 score below 0, Panel A) and democratic countries (polity2 score above 0, Panel B). All specifications include time fixed effects. *, **, *** denote statistical significance at the, respectively, 90%, 95%, and 99% confidence level. Standard errors (shown in parentheses) are Huber robust and clustered at the country level. Source: Authors’ calculations. 45 Table 4: Effects of Oil Price Windfalls on Various Dependent Variables (1) (2) (3) (4) (5) (6) (7) (8) EPL Rent CA GDP Employment Revenues Expenditure Budget Impact Responses Aut + -0.19** 0.79** 0.62** 0.01 0.09 0.37*** -0.16 0.55** (0.09) (0.31) (0.30) (0.23) (0.09) (0.12) (0.20) (0.22) Aut - -0.04 -0.42** -0.49** -0.42* 0.05 -0.52*** 0.43*** -0.94*** (0.11) (0.17) (0.22) (0.22) (0.07) (0.13) (0.11) (0.18) Dem + 0.02 0.74*** 0.46*** -0.05 0.00 0.34** 0.16 0.21* (0.17) (0.26) (0.16) (0.13) (0.11) (0.14) (0.12) (0.11) Dem - -0.24 -0.87*** -1.51*** -0.09 0.07 -0.58** 0.43** -1.01*** (0.18) (0.14) (0.51) (0.38) (0.12) (0.26) (0.21) (0.25) Obs 2,886 2,724 2,912 2,989 1,755 2,361 2,319 2,319 Adj. R2 0.025 0.366 0.086 0.136 0.052 0.066 0.071 0.149 3-year responses Aut + -0.47*** 0.34** 0.32 0.63 0.33*** 0.45** 0.19 0.32** (0.15) (0.13) (0.35) (0.53) (0.12) (0.21) (0.26) (0.13) Aut - -0.30 0.03 -0.08 -0.27 0.18 0.97 1.24** -0.26 (0.40) (0.28) (0.21) (0.34) (0.17) (0.70) (0.60) (0.33) Dem + -0.01 0.42*** 0.08 0.36 0.09 0.41*** 0.48** -0.06 (0.27) (0.14) (0.19) (0.33) (0.11) (0.14) (0.22) (0.14) Dem - -0.65** -0.63*** -1.22*** 0.02 0.14 -0.34 0.58 -0.93*** (0.32) (0.17) (0.33) (0.53) (0.20) (0.33) (0.37) (0.25) Obs. 2,804 2,645 2,829 2,906 1,674 2,285 2,240 2,240 Adj. R2 0.044 0.350 0.089 0.173 0.055 0.045 0.106 0.207 5-year responses Aut + -0.60*** 0.36** -0.05 1.38** 0.54*** 0.03 -0.12 0.16 (0.22) (0.16) (0.39) (0.66) (0.14) (0.33) (0.29) (0.40) Aut - -0.53 0.43 -0.77** -0.21 -0.51* 0.85 0.95** -0.12 (0.48) (0.29) (0.35) (0.76) (0.28) (0.60) (0.45) (0.43) Dem + 0.11 0.45*** -0.34 0.72 0.23 0.36* 0.56*** -0.18 (0.21) (0.12) (0.38) (0.67) (0.25) (0.19) (0.21) (0.23) Dem - -1.27** -0.03 -1.44** 0.41 -0.08 -0.24 0.61 -0.89 (0.50) (0.16) (0.58) (0.65) (0.33) (0.38) (0.73) (0.62) Obs. 2,640 2,487 2,665 2,740 1,512 2,135 2,092 2,092 Adj. R2 0.056 0.378 0.082 0.218 0.058 0.032 0.099 0.201 CA = current account balance, EPL = employment protection legislation, GDP = gross domestic product. Notes: Effects of positive (+) and negative (-) oil price windfalls on a range of dependent variables in autocracies (Aut) and democracies (Dem), at impact as well as two and four years after the windfalls. Estimates are obtained estimating Equation (4) for k=0,2,4, replacing 𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,𝑡𝑡 with (i) oil rents as a share of GDP (Column 2), (ii) the current account balance as a share of GDP (Colum 3), (iii) 100 times the log of real per capita GDP in local currency (Column 4), (iv) the employment-to-population ratio (Column 5), (v) tax revenues as a share of GDP (Column 6), (vi) government expenditures as a share of GDP (Column 7), and (vii) the primary balance as a share of GDP (Column 8). *, **, *** denote statistical significance at the, respectively, 90%, 95%, and 99% confidence level. Standard errors (shown in parentheses) are Huber robust and clustered at the country level. Source: Authors’ calculations. 46 Table 5: Effects of Positive and Negative Oil Price Windfalls on a Systemic Banking Crisis Probability (1) (2) (3) (4) (5) (6) (7) (8) (9) Prob Prob Prob Log Log Log Con log Con log Con log Aut + -0.00 -0.02 -0.00 0.02 -0.02 -0.02 0.02 -0.02 -0.02 (0.04) (0.07) (0.09) (0.07) (0.14) (0.20) (0.17) (0.18) (0.23) Aut - 0.04 -0.02 -0.06 0.09 0.02 -0.08 0.09 0.02 -0.07 (0.13) (0.14) (0.19) (0.29) (0.30) (0.37) (0.27) (0.29) (0.35) Dem + -0.19 -0.21 -0.37 -0.34 -0.40 -0.87 -0.34 -0.39 -0.84* (0.14) (0.14) (0.28) (0.26) (0.28) (0.60) (0.30) (0.32) (0.50) Dem - 0.33*** 0.31** 0.38** 0.65*** 0.66** 0.71** 0.65** 0.64** 0.67** (0.12) (0.14) (0.16) (0.23) (0.26) (0.30) (0.27) (0.29) (0.32) Obs. 2,882 2,620 2,108 2,882 2,620 2,108 2,882 2,620 2,620 Pseudo R 2 0.03 0.06 0.17 0.03 0.06 0.17 Time FE NO NO YES NO NO YES NO NO YES Country FE NO YES YES NO YES YES NO YES YES GDP = gross domestic product, USD = United States dollar. Notes: Effects of positive (+) and negative (-) oil price windfalls on the probability of a systemic banking crisis occurring in autocracies (Aut) and democracies (Dem). Estimates are obtained by estimating probit (Prob), logit (Log), and conditional logit (Con log) models. The dependent variable takes value 1 if a systemic banking crisis occurs either at year t or t+1. Explanatory variables include windfalls at year t, as well as forward windfall variables (year t+1), lagged windfall variables (year t-1 and t-2), and lagged GDP per capita in USD (year t-1). Time and country fixed effects are included depending on the specification. Coefficients report the effect of windfall variables at year t. *, **, *** denote statistical significance at the, respectively, 90%, 95%, and 99% confidence level. Standard errors (shown in parentheses) are clustered at the country level. Source: Authors’ calculations. 47 Table 6: Marginal Effects of Positive and Negative Oil Price Windfalls on a Systemic Banking Crisis Probability (1) (2) (3) (4) (5) (6) Prob Prob Prob Log Log Log Aut + -0.00 -0.00 -0.00 0.00 -0.00 -0.00 (0.00) (0.01) (0.01) (0.00) (0.01) (0.01) Aut - 0.00 -0.00 -0.01 0.00 0.00 -0.00 (0.01) (0.02) (0.02) (0.02) (0.02) (0.02) Dem + -0.02 -0.02 -0.04 -0.02 -0.02 -0.04 (0.02) (0.02) (0.03) (0.01) (0.01) (0.03) Dem - 0.04*** 0.04** 0.04** 0.04*** 0.04*** 0.03** (0.01) (0.02) (0.02) (0.01) (0.01) (0.01) Obs. 2,882 2,620 2,108 2,882 2,620 2,108 Pseudo R 2 0.03 0.06 0.17 0.03 0.06 0.17 Time FE NO NO YES NO NO YES Country FE NO YES YES NO YES YES Notes: Marginal effects of positive (+) and negative (-) oil price windfalls on the probability of a systemic banking crisis occurring in autocracies (Aut) and democracies (Dem). Marginal effects are calculated at the means of the dependent variables. See notes in Table 5. Source: Authors’ calculations. 48 REFERENCES Abiad, Abdul, and Ashoka Mody. 2005. “Financial Reform: What Shakes It? What Shapes It?” American Economic Review, 95 (1): 66–88. Acemoglu Daron, and James A. Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Agnello, Luca, Vitor Castro, João Tovar Jalles, and Ricardo M. Sousa. 2015. “What Determines the Likelihood of Structural Reforms?” European Journal of Political Economy, 37: 129–45. Alesina, Alberto, Davide Furceri, Jonathan D. Ostry, Chris Papageorgiou, and Dennis P. Quinn. 2023. “Structural Reforms and Elections: Evidence from a World-Wide New Dataset.” Journal of the European Economic Association, 22 (4): 1936–80 Alexeev, Michael, and Robert Conrad. 2009. “The Elusive Curse of Oil.” Review of Economics and Statistics 91 (3): 586–98. Araujo, Jorge A. De Thompson R., Ekaterina Vostroknutova, Markus Brueckner, Mateo Clavijo Munioz, and Konstantin M. Wacker. 2014. Beyond Commodities: The growth challenge of Latin America and the Caribbean. World Bank Report No. 91015. Arezki, Rabah, and Markus Brueckner. 2012. “Commodity Windfalls, Democracy and External Debt.” The Economic Journal 122, 561 (1): 848–66. Auerbach, Alan J., and Yuriy Gorodnichenko. 2012. “Fiscal Multipliers in Recession and Expansion.” In Alberto Alesina and Francesco Giavazzi, eds. Fiscal Policy After the Financial Crisis. pp. 63–98. National Bureau of Economic Research. Bazzi, Samuel, and Christopher Blattman. 2014. “Economic Shocks and Conflict: Evidence from Commodity Price Shocks.” American Economic Journal: Macro 6: 1– 38. Bentolila, Samuel, and Giuseppe Bertola. 1990. “Firing Costs and Labour Demand: How Bad Is Eurosclerosis?” The Review of Economic Studies, 57 (3): 381–402. Blanchard, Olivier, and Francesco Giavazzi. 2003. “Macroeconomic Effects of Regulation and Deregulation in Goods and Labor Markets.” The Quarterly Journal of Economics, 118 (3): 879–907. Blanchard, Olivier J., William D. Nordhaus, and Edmund S. Phelps. 1997. “The Medium Run.” Brookings Papers on Economic Activity, 1997 (2): 89–158. 49 Brueckner, Markus. 2017. “Rent Extraction by Capitalists.” European Journal of Political Economy, 50 (C), 157–70. Brueckner, Markus, and Antonio Ciccone. 2010. “International Commodity Prices, Growth, and the Outbreak of Civil War in Sub-Saharan Africa.” Economic Journal, 120 (544): 519–34. Brueckner, Markus, Andrea Tesei, and Antonio Ciccone. 2012. “Oil Price Shocks, Income, and Democracy.” Review of Economics and Statistics 94 (2): 389–99. Brunnschweiler, C. and Erwin Bulte. 2008. “Linking Natural Resources to Slow Growth and More Conflict.” Science, 320 (5876): 616–7. Cacciatore, Matteo, and Giuseppe Fiori. 2016. “The Macroeconomic Effects of Goods and Labor Markets Deregulation.” Review of Economic Dynamics, 20: 1–24. Cahuc, Pierre, Fabien Postel-Vinay, and Jean-Marc Robin. 2006. “Wage Bargaining with On-the-Job Search: Theory and Evidence.” Econometrica, 74 (2): 323–64. Ciminelli, Gabriele, Romain Duval, and Davide Furceri. 2022. “Employment Protection Deregulation and Labor Shares in Advanced Economies.” The Review of Economics and Statistics, 104 (6): 1174–90. Ciminelli, Gabriele, and Davide Furceri. Forthcoming. “A New Dataset of Employment Protection Legislation.” Collier, Paul, and Benedikt Goderis. 2012. “Commodity Prices and Growth: An Empirical Investigation.” European Economic Review, 56 (6): 1241–60. Cotet, Anca M., and Kevin K. Tsui. 2013. “Oil Discoveries and Conflict: What Does the Cross-Country Evidence Really Show?” American Economic Journal: Macroeconomics, 5 (1): 49–80. Drazen, Allan, and William Easterly. 2001. “Do Crises Induce Reform? Simple Empirical Tests of Conventional Wisdom.” Economics and Politics, 13 (2): 129–57. Duval, Romain, Davide Furceri, and Joao Jalles. 2020. “Job Protection Deregulation in Good and Bad Times.” Oxford Economic Papers, 72 (2): 370–90. Duval, Romain, Davide Furceri, and Jakob Miethe. 2021. “Robust Political Economy Correlates of Major Product and Labor Market Reforms in Advanced Economies: Evidence from BAMLE for Logit Models.” Journal of Applied Econometrics, 36 (1): 98–124. 50 Esanov, Akram. 2006. “The Growth-Poverty Nexus: Evidence from Kazakhstan.” ADB Institute Discussion Paper No. 51. Asian Development Bank Institute (ADBI). Fernandez, Andrés, Stephanie Schmidt-Grohe, and Martín Uribe. 2020. “Does the Commodity Super Cycle Matter?” NBER Working Paper No. 27589. National Bureau of Economic Research (NBER) Gaviria, Alejandro U. 2005. La Reforma Laboral de 2002: ¿Funcionó o no? Coyuntura Económica, 35 (1): 73–100. Universidad de los Andes. Gelb, Alan. 1990. Oil Windfalls: Blessing or Curse? A Comparative Study of Six Developing Exporters. Oxford University Press, 1990. Grindle, Merilee S., and John W. Thomas. 1991. Public Choices and Policy Change: The Political Economy of Reform in Developing Countries. Johns Hopkins University Press. Haggard, Stephan, and Robert R. Kaufman, eds. 1992. The Politics of Economic Adjustment: International Constraints, Distributive Conflicts, and the State. Princeton University Press, Princeton. Haggard, Stephan, and Steven B. Webb. 1994. Voting for Reform: Democracy, Political Liberalization and Economic Adjustment. Oxford University Press: New York. Jordà, Òscar, and Alan M. Taylor. 2016. “The Time for Austerity: Estimating the Average Treatment Effect of Fiscal Policy.” Economic Journal, 126 (590): 219–55. Jordà, Òscar. 2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review, 95 (1): 161–82. Karl, Terry. 1997. The Paradox of Plenty: Oil Booms and Petro States. University of California Press. Laeven, Luc, and Fabian Valencia. 2018. “Systemic Banking Crises Revisited.” IMF Working Paper No. 18/206. International Monetary Fund (IMF). Lancaster, Tony. 2000. “The Incidental Parameter Problem Since 1948.” Journal of Econometrics, 95 (2): 391–413. Lora, Eduardo A., and Mauricio Olivera. 2004. “What Makes Reforms Likely: Political Economy Determinants of Reforms in Latin America.” Journal of Applied Economics, 7 (1): 99–135. Mehlum, Halvor, Karl Moene, and Ragnar Torvik. 2006. “Institutions and the Resource Curse.” Economic Journal, 116 (508): 1–20. 51 Olea, José Luis Montiel, and Mikkel Plagborg-Møller. 2021. “Local Projection Inference is Simpler and More Robust Than You Think.” Econometrica, 89 (2): 1789–823. Mortensen, Dale T., and Christopher A. Pissarides. 1994. “Job Creation and Job Destruction in the Theory of Unemployment.” The Review of Economic Studies, 61 (3): 397–415. Nelson, Joan M. 1990. Economic Crisis and Policy Choice: The Politics of Adjustment in the Third World. Princeton University Press, Princeton. O’Keefe, John. 1997. “The Texas Banking Crisis: 1980–1989.” Federal Deposit Insurance Corporation. Perry, Guillermo, and Mauricio Olivera. 2009. “El Impacto Del Petróleo Y la Minería en El Desarrollo Regional Y Local en Colombia.” CAF Documento de trabajo N° 2009/06. Caracas, Venezuela: CAF. Pissarides, Christopher A. 2000. “Equilibrium Unemployment Theory.” MIT Press. Plagborg-Møller, Mikkel, and Christian K. Wolf. 2021. “Local Projections and VARs Estimate the Same Impulse Responses.” Econometrica 89 (2): 955–80. Postel-Vinay, Fabien, and Jean--Marc Robin. 2002. “Equilibrium Wage Dispersion with Worker and Employer Heterogeneity.” Econometrica, 70 (6): 2295–350. Rodrik, Dani. 1996. “Understanding Economic Policy Reform.” Journal of Economic Literature, 34 (1): 9–41. Ross, Michael L. 2001. “Does Oil Hinder Democracy?” World Politics, 53 (3): 325–61. Sorbello, Paolo. 2023. “Hiring, Firing, Atomizing: Manpower Agencies and Precarious Labor in Kazakhstan’s Oil Sector.” International Labor and Working-Class 103: 823. Cambridge University Press. Teulings, Coen N., and Nikolay Zubanov. 2014. “Is Economic Recovery a Myth? Robust Estimation of Impulse Responses.” Journal of Applied Econometrics 29 (3): 497– 514. Williamson, John, and Stephan Haggard. 1994. “The Political Conditions for Economic Reform.” In John Williamson, ed. The Political Economy of Policy Reform. Institute for International Economics, pp. 525–96. Wooldridge, Jeffrey M. 2010. Econometric Analysis of Cross Section and Panel Data (2nd ed.). MIT Press. ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org ADB ECONOMICS WORKING PAPER SERIES NO. 752 November 2024 External Shocks and Labor Market Reforms in Autocracies and Democracies Evidence from Oil Price Windfalls This paper explores how oil price windfalls impact labor market regulation across 83 countries from 1970 to 2014. Findings reveal that windfall gains lead to deregulation in autocracies but not democracies, while windfall losses trigger deregulation in democracies. The paper examines how democracies use windfall gains to boost spending, whereas autocracies may focus on rent extraction and increasing economic efficiency. Deregulation in democracies during losses aligns with the crisis-induced reform hypothesis, linked to twin deficits and heightened banking crisis risks. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 69 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. EXTERNAL SHOCKS AND LABOR MARKET REFORMS IN AUTOCRACIES AND DEMOCRACIES EVIDENCE FROM OIL PRICE WINDFALLS Markus Brueckner, Gabriele Ciminelli, and Norman Loayza