Impact Evaluation of “Nepal Trade Preference Program”
In 2015, Nepal was hit by a massive earthquake. The earthquake claimed thousands of lives and brought massive destruction to infrastructure creating havoc on the economy. Post Disaster Needs Assessment (PDNA) by Nepal Planning Comissiong estimated that the total value of loss directly or indirectly from the earthquake would be about USD 7 billion.
Immediately following the earthquake, international communities started providing aid and relief to Nepal. Most of the aid came in the form of short term support, like supply of tents, medical equipment, fuels and cooking materials, etc. But some programs also provided long term economic and social support to Nepal. One of such was the Nepal Trade Preference Program (NTPP), provided by the United States of America in the “The Trade Facilitation and Trade Enforcement Act of 2015”. Under this program, 66 different products (on 8 digit HS code) being export to the United States from Nepal would benefit from tariff exemption until 2025. Later the number of products was increased to 77. Handicrafts, pashmina, bags and such products were among the list of 77 products under the NTPP.
The NTPP program was provided to strengthen Nepal’s domestic economy by assisting in export promotion for the products eligible under this program, which should then stimulate economic growth back home and create employment. But with the pandemic disrupting business activities and trade around the globe, Nepal has been losing out on taking full advantage of the benefit for the past two years. Moreover, even before the pandemic, there had been little evidence that the program had been beneficial. The Ministry of Commerce in Nepal has on more than one occasion requested the United States Trade representative to increase the number of eligible products in the NTPP. Thus, this article will take a deep look into whether the program has been beneficial to the export from Nepal or not.
Identifying the indicator of comparison
The NTPP benefit is applicable from the start of 2017. NTPP eligible products would be able to claim for duty-free entry to the United States while being exported by Nepal from the very first day of 2017.
To find the impact of the program we can compare the trade before and after the policy intervention of the eligible products. But we should also consider that the factors apart from the policy, such as the exchange rate, can also influence the export. For example, a depreciation of the dollar may have a positive impact on the export from Nepal, as a result, the impact of the program may be overestimated while making comparisons over time. Another possible way to make a comparison is to compare those products receiving the preference program against those not receiving any preference. But since the products are not random, such comparison can be biased. The products receiving the preference against those not receiving may differ significantly among each other even before the program, so any difference among the group may be explained by factors other than the policy intervention (NTPP), making the comparison biased.
However, both of these methods can be combined to make a Difference-in-Difference (DID) estimation. DID is an econometric model that compares the difference in outcome over time between treated and not treated populations. The population is divided into two groups, one receiving treatment and the others not receiving treatment. The first difference is the difference in outcome within the group, before and after the treatment. This controls for the factor that is constant over time since we are comparing the same group to itself. The difference in outcome between the pre-and post-treatment of the group that did not receive treatment, but was exposed to the same environment is considered the counterfactual. (Paul J. Gertler) This method will make the comparison between the NTPP eligible products before and after the intervention and also between NTPP eligible products against not eligible products.
Using DID to estimate the impact of the program
The DID regression follows the following OLS line –
ln (Yi)=α+ δ(Ti.ti)+βTi+γti+ ϵi
where Yi is the total volume of export, Ti is the treated variable, ti is the time variable and Ti.ti is the DID coefficient.
Note that many products from Nepal are exported to the United States also under the GSP preferential program. Since this study is the comparison between those receiving NTPP treatment against those receiving no treatments, all GSP eligible products apart from those also in NTPP are ignored. (There are 25 products which are both in GSP and NTPP list being exported after 2016)
To run a DID regression, the treated and the control groups must show parallel tendencies over time. International trade is very fluctuating in the short run, a product may be exported over thousands in a month and non in another. However, over time, the trade is expected to be smooth. Thus, to check the parallel tendency, three years moving average of Nepal’s export to the United States is taken. Refer to the figure below-
From the figure, it can be seen that the volume of NTPP eligible exported and volume of export with no benefits have a similar tendency over time. Since the intervention started in 2017, only the line before 2017 is relevant for this case. It can be clearly seen that the GSP eligible product had skyrocketing exports after 2014, but the other two groups have roughly similar trends.
Thus, after finding the parallel tendency in export volume over time, the DID regression gives the following result.
The independent variable ‘NTPP’ in the table is the coefficient of the eligible products of the treatment group. The coefficient of the NTPP suggests that the export of NTPP products on average was 3.384 more than those of the other products (even without the intervention)
The coefficient of the Intervention variable shows the average of how much the log total units of import change if the products were imported after the intervention (i.e after 2016). On average, the overall export has decreased by log 0.10 after 2016 (for all products) but was not statistically significant.
Finally, the NTPP: Intervention is the DID estimate. The coefficient estimate’s impact of the program. The coefficient shows how the products eligible for NTPP were affected after the policy was implemented. This estimate is the answer to the hypothesis of this article.
The DID estimate is 1.034, significant at 0.01 level. This means that the program has on average positively affected the imports of the treated product by the log value of 1.034. A product under the NTPP program was on average log 1.034 times more imported by the United States than the products receiving no benefits at all. This shows that the impact of NTPP has been significantly positive.
Conclusion
The estimator showed that the program has been positive for export promotion. Export of NTPP eligible products was log: 1.034 more exported than products receiving no benefits at all. Even when the raw data showed that the export of NTPP eligible products had been falling since the intervention, the result showed NTPP having a positive impact on the trade. This might seem counter-intuitive, but in reality, other remaining products with no preference also had a loss in export to the United States at the same period. However, because of the preference program, the products eligible for the preference did not fall as much as they would have otherwise. Thus, while the program may not have promoted the export, it ensured that export did not fall as much as it would have if there was no program. The estimator showed that the program had a positive and significant impact on the export of the eligible products.
Find the data and analysis in my Github: https://github.com/AlabhyaMe/Thesis-Nepal-Trade-Preference
