Winners and Losers in Industrial Policy Mohamed Ali Marouani & - PowerPoint PPT Presentation

Industrial Upgrading in Tunisia Winners and Losers in Industrial Policy Mohamed Ali Marouani & Michelle 2.0 : An Evaluation of the impacts of the Marshalian Tunisian Industrial Upgrading Program Introduction Data Description & Identification Strategy Mohamed Ali Marouani & Michelle Marshalian Approach & Econometric Specification LEDA-DIAL Findings & Conclusions PSL - Dauphine, Paris 1 and Bibliography UMR D´ eveloppement et soci´ et´ es Universit´ e Paris 1, Panth´ eon-Sorbonne michelle.marshalian@dauphine.eu UN-WIDER Conference in Bangkok, Thailand, September 11-13, 2019

Industrial Overview Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Introduction Data Description & Identification Strategy Approach & Data Description & Identification Strategy Econometric Specification Findings & Conclusions Approach & Econometric Specification Bibliography Findings & Conclusions

Industrial Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Introduction Data Description & Identification Strategy Data Description & Identification Strategy Approach & Econometric Specification Findings & Approach & Econometric Specification Conclusions Bibliography Findings & Conclusions

Industrial Motivation Upgrading in Tunisia Mohamed Ali Marouani & Michelle ◮ Over the past 2-3 decades increasing openness to Marshalian trade and focus on increasing competitiveness to Introduction meet these demands Data Description & Identification ◮ Industrial policies are unpopular : market Strategy Approach & distortions , political capture and its misguided Econometric Specification focus on sectors. Findings & Conclusions ◮ But continued focus on industrial development and Bibliography the success of Asian countries has brought such policies back to the limelight. ◮ Who gains from IPs ? What is it’s impact on jobs and wages? And implicitly, what does this say about its purpose ?

Industrial Literature Review Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian ◮ The literature tells us that the impacts of firm Introduction subsidies on productivity are almost always negative Data Description & or non-significant. Identification Strategy ◮ Negative or no impact on firms (Criscuolo, 2019; Approach & Cerqua, 2014) Econometric ◮ If there are positive impacts they are : Specification ◮ 2-4 yrs after (Bernini, 2017) Findings & Conclusions ◮ only for on small firms (Criscuolo, 2019) Bibliography ◮ But the state also uses IP to guarantee its clients a non-competitive environment (Cammett 2007, Murphy 2006 and Rijkers 2017 in Tunisia; and Rougier 2016 in Egypt).

Description: The Industrial Upgrading Industrial Upgrading in Tunisia Program (PMN) Mohamed Ali Marouani & Michelle ◮ The Industrial Upgrading Program (PMN) was Marshalian implemented after the Free Trade Agreement with Introduction the EU with the following goals: Data Description & ◮ competitiveness, Identification Strategy ◮ exports, Approach & ◮ innovation and Econometric ◮ labor market outcomes. Specification Findings & Conclusions Bibliography Source: Office of the Industrial Upgrading Program

Industrial Allocation of IUP funds Upgrading in Tunisia ◮ More than 5K grants in the last 20 years equivalent Mohamed Ali Marouani & to 1.26 Billion Tunisian Dinars (500 Million US$). Michelle Marshalian ◮ 2/3 of the amount were spent on material purchases Introduction and the rest on immaterial acquisitions. Data Description & ◮ Focused on large firms : 60% of recipient firms had Identification Strategy over 50 workers Approach & Econometric Specification Findings & Conclusions Bibliography

Industrial How were funds allocated? Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian ◮ The COPIL - a board of multi-stakeholders – and the Introduction bureau of the IUP decided on who received benefits. Data Description & Identification Strategy ◮ These were closed door sessions, with low-oversight Approach & Econometric → It quickly became well known that members of the Specification inner circle of the regime benefited from this. Findings & Conclusions ◮ But overall there was support from business and civil Bibliography society. International donors were positive about it. → largely perceived as beneficial for Tunisian firms and employment.

Industrial Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Introduction Data Description & Identification Strategy Data Description & Identification Strategy Approach & Econometric Specification Findings & Approach & Econometric Specification Conclusions Bibliography Findings & Conclusions

Industrial Data Description & Identification Strategy Upgrading in Tunisia Mohamed Ali Marouani & Michelle Data Description Marshalian 1. National firm-level enterprise registry ( R´ epertoire Introduction nationale des entreprises ) from 2000 to 2016. Data Description & Identification Strategy ◮ A sample of firms with at least 6 employees Approach & ◮ Approximately 125,000 obs in an unbalanced panel Econometric Specification of 7,000 firms. ◮ Firm-level data on exports from national export Findings & Conclusions agency from 2005-2010. Bibliography 2. PMN survey by ITCEQ ( Institut tunisien de la comp´ etitivit´ e et des ´ etudes quantitatives ) 3. Treatment data from database online and in consultation with research institute in Tunisia.

Industrial Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Introduction Data Description & Identification Strategy Data Description & Identification Strategy Approach & Econometric Specification Findings & Approach & Econometric Specification Conclusions Bibliography Findings & Conclusions

Industrial Approach Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Data Description & ◮ Approach Identification Strategy ◮ A (double) weighted propensity score matching Approach & method to create control groups, with assignment Econometric Specification based on fuzzy matching technique. ◮ Combined with a re-weighted panel Findings & Conclusions differences-in-differences (Card, 1990; Hirano, Bibliography Imbens, and Ridder, 2003).

Industrial Econometric Specification Upgrading in Tunisia Mohamed Ali Marouani & � n = 3 y i , t = β 0 + β 1 Treated ∗ After i , t + β 2 t + n After i , t + Michelle Marshalian β 3 TreatmentGroup i + β 4 Anticipation i , t − 1 + (1) � n Introduction β 5 t Treated ∗ After ∗ Year i , t + Data Description & β 6 X ′ i , t γ + τ t + λ i + ζ i + ǫ i Identification Strategy Approach & ◮ y i , t : log of employment, log of average wages per Econometric Specification worker and the log of net job creation. Findings & ◮ β 1 : main treatment variable of interest Conclusions Bibliography ◮ β 2 : time-specific treatment effects (1-3 years) ◮ β 3 : treatment group assignment ◮ β 4 : anticipation effect of the program (one year prior) ◮ β 5 : year-specific treatment effect ◮ β 6 : controls (age, age-squared, size, distance to ports and lagged and growth components) ◮ year ( τ t ), regional ( λ i ), and sector ( ζ i ) fixed effects

Industrial Upgrading in Tunisia Mohamed Ali Marouani & Michelle Marshalian Introduction Introduction Data Description & Identification Strategy Data Description & Identification Strategy Approach & Econometric Specification Findings & Approach & Econometric Specification Conclusions Bibliography Findings & Conclusions

Industrial Small but significant increase in wages Upgrading in Tunisia Mohamed Ali Table: Impact of the IUP on Average Wages. Marouani & Michelle Marshalian Introduction OLS Panel Fixed Effects Models Reg. Adj. Models Log of (1) (2) (3) (4) (5) Data Description & Identification Ave. Wages PSM IPW Strategy Treatment -0.003 0.007 0.013** -0.070*** 0.023** Approach & [-0.447] [1.208] [2.081] [-5.134] [2.249] Econometric 1-year after 0.004 0.018*** 0.021*** -0.006 Specification [0.579] [3.621] [3.646] [-0.486] Findings & 2-years after 0.007 0.020*** 0.020*** -0.012 Conclusions [1.118] [3.625] [3.249] [-1.133] Bibliography 3-years after 0.003 0.019*** 0.017*** -0.008 [0.430] [3.126] [2.605] [-0.672] Anticipation 0.030*** 0.011** 0.022*** -0.008 [4.654] [2.052] [3.687] [-0.637] Treat*Year No No Yes No Yes Full Controls No Yes Yes Yes Yes Observations 327,234 195,501 195,501 69,077 69,077 R-squared 0.347 0.458 0.458 0.0004 0.693