1: \begin{abstract}

2: In order to evaluate the impact of a policy intervention on a group of units over time, 

3: it is important to correctly estimate the average treatment effect (ATE) measure. 

4: Due to lack of robustness of the existing procedures of estimating ATE from panel data, in this paper, 

5: we introduce a robust estimator of the ATE and the subsequent inference procedures using the popular approach of minimum density power divergence inference. 

6: Asymptotic properties of the proposed ATE estimator are derived and used to construct robust test statistics for testing parametric hypotheses related to the ATE.

7: Besides asymptotic analyses of efficiency and powers, extensive simulation studies are conducted to study the finite-sample performances of 

8: our proposed  estimation and testing procedures under both pure and contaminated data. 

9: The robustness of the ATE estimator is further investigated theoretically through the influence functions analyses.

10: %justifying their stability against any data contamination in pre-treatment periods.  

11: %Subsequently, based on these influence analyses, another slightly modified robust ATE estimator is proposed 

12: %to successfully handle data contamination in post-treatment periods, along with empirical illustrations of its robustness.   

13: Finally our proposal is applied to study the long-term economic effects of the 2004 Indian Ocean earthquake and tsunami

14: on the (per-capita) gross domestic products (GDP) of five mostly affected countries, namely Indonesia, Sri Lanka, Thailand, India and Maldives. 

15: \end{abstract}