Social Clauses in Free-trade Agreements: An efficient tool to improve Labor Standards? Remi BAZILLIER 1,2 Arslan T. RANA 1,2 1 LEO, CNRS UMR 7322 2 Univ. Orléans 4th Conference on the Regulating Decent Work Network, July 9th 2015
Introduction Labor clauses in trade agreements: an old controversy Strong oppositions between developed and developing countries One explanation of the failure of the WTO conference in Seattle (1999) No labor provisions in WTO rules Proposal is blocked at the multilateral level The 1996 Singapore WTO Ministerial Declaration The International Labour Organization (ILO) is the competent body to set and deal with these standards, and we affirm our support for its work in promoting them. (...) We reject the use of labour standards for protectionist purposes, and agree that the comparative advantage of countries, particularly low-wage developing countries, must in no way be put into question.
The debate Pros. Trade Unions, Developped Countries Unfair competition based on violations of Human Rights at Work Avoid a race-to-the bottom in labor standards Use trade as a way to improve labor standards in the South Cons. Developing countries Hidden protectionism from developed countries Comparative advantage in low wages for countries with low level of technology and capital Violations of labor standards are not mainly observed in the export sector Sanctions are counterproduction as they harm the people they are designated to help
Since the failure of the WTO Doha round: a rapid development of bilateral or regional free trade agreements Increasing number of social clauses within such trade agreements (US, Europe, Canada) Different types of labor clauses EU: chapter on soustainable development US: more binding (DSM in some cases)
Goal of the paper Assess quantitatively the effect of labor clauses on labor standards in low and middle income countries Focus on ILO conventions (especially core conventions) and national labor market institutions Study possible differences of effects between different types of labor clauses Effects of legally enforceable labor clauses Effects of dispute settlement mechanisms (DSM) Empirical strategy Difference-in-difference strategy We take into account the endogeneity of labor clauses by using instrumental variables
Outline 1 Introduction Introduction 2 Literature 3 Data 4 Empirical Strategy Difference-in-difference Endogeneity 5 Results Overall effect of labor provisions Effects of different types of labor clauses 6 Conclusion and future works
Related Literature Analyses of labor clauses: Chan and Ross 2010, Siroen et al 2008, Siroen 2013, ILO 2013, Grandi 2009, Bolle 2014, Ebert and Posthuma 2011, Pills and Fino 2015 Trade Policy and Labor Standards: Theoretical works: Brown, Deardorff and Stern (1996), Maskus (1997), Bagwel and Staiger (1998), Martin and Maskus (2001), Limao (2005), Spagnolo (2001), Zhao (2009) Normative Approach: Srinivasan (1998), Bhagwati (2003), Effects of Generalized Systems of Preferences (GSM): Granger and Siroen (2006), Ebert and Posthuma (2011), Cooper (2006), Zhou and Cuyvers (2012), Krivoi (2008) Effects of PTA Baier and Bergstrand (2007), Bergstrand et al. (2010), Horn, Mavroidis and Sapir (2010), Vicard (2009), Cipollina and Salvatici Effects of PTA on labor: Haberli et al. 2012
Outline 1 Introduction Introduction 2 Literature 3 Data 4 Empirical Strategy Difference-in-difference Endogeneity 5 Results Overall effect of labor provisions Effects of different types of labor clauses 6 Conclusion and future works
Data ILO conventions Core labor standards (8 conventions), Governance Conventions, Social Security Convention, collective bargaining convention, termination of employment convention Labor Market Institutions: Schindler and Aleksynska (2011) Minimum Wage (% of mean wage), Unemployment benefits, employment protection PTA and labor clauses: Global Preferential Trade Agreements Database (GPTAD) Content analysis following the criteria proposed by Horn et al. (2010) Area covered, legally enforceable, dispute settlement mechanisms Controls: GDP, GDP per cap., Gross secondary enrollment, total trade, imports, share of manufacturing and services (WDI)
Empirical Strategy Difference-in-Difference Approach LS i,t = α+β 1 LP AC i,t +β 2 LP LE i,t +β 3 PTA i,t +Controls i,t +F i +γ t +ɛ i,t (1) β 3 : effects of PTA (compared to countries with no PTA) β 1 : effects of labor clauses, compared to countries with PTA but without labor clauses β 2 : additional effects of legally enforceable labor clauses (compared to countries with labor clauses with area covered only) Intensity of the treatment : total number of PTA, LP AC, LP LE Countries and time fixed effects
Endogeneity of labor clauses Different sources of endogeneity: Cofounding factors: correlated with both labor clauses and labor standards Selection: countries with high labor standards more likely to accept social clauses Solutions: Instruments Correlated with labor clauses but with no direct link with labor standards Here: the propension of partner countries to include labor clauses in their trade agreements Ex: the probability to have a social clause in PTA is higher if i sign a PTA with the US (that includes social clauses in almost all their PTA) than with India We use the average share of social clauses for all countries with a PTA with one given country
Outline 1 Introduction Introduction 2 Literature 3 Data 4 Empirical Strategy Difference-in-difference Endogeneity 5 Results Overall effect of labor provisions Effects of different types of labor clauses 6 Conclusion and future works
IV results
IV results
Strong and positive impact of labor clauses on ratifications of ILO conventions One more core convention ratified following a PTA with labor clause In average, low and middle income countries have ratified 3.99 core ILO conventions +27% in average Negative impact of PTA alone But result is not very robust (turns not significant in several estimates) No significant effect on other labor market institutions But this last result should be taken cautiously as the sample of countries and the time span are much more limited
Effects of different types of labor clauses Distinction between labor provisions: Area covered Legally enforceable Dispute settlement mechanism
Still positive impact of labor clauses on ILO conventions But striking results for legally enforceable labor clauses the additional effect (compared to labor clauses in general) is negative the net effect is closed to zero Interpretation If labor clauses are legally enforceable, it may be more costly to ratify ILO convention if the country does not have the capacity to implement it effectively Trade-off between maximization of ratifications and effective implementation Future researches should investigate the effect on de facto level of labor standards (using violations of labor standards reported by the ILO?)
Results still not significant for other labor market institutions But although not significant, interesting result for employment protection Negative sign for labor clauses and positive for legally enforceable clauses We also find no additional effects when DSM are included
Conclusion and future works Main results Positive (and strong) impact of labor clauses on ratifications of ILO conventions, especially fundamental conventions No significant impact on other labor market institutions Additional effect of legally enforceable labor clauses on ratifications of ILO conventions is negative Further invetigations on de facto level of labor standards Additional things to do: Check if the effect is different for PTA with EU and US More work on other labor market institutions More countries, time span (limitation of existing indicators) Additional dimension: social protection
Thanks for your attention! remi.bazillier@univ-orleans.fr