G. Badri Narayanan
Buongiorno and Zhu (2016) (BZ henceforth) is an insightful and comprehensive study of the potential effects of one of the most widely discussed trade agreements in recent times, namely the Trans-Pacific Partnership (TPP), on forestry products. The details on forestry products may be well-captured in the Global Forest Products Model (GFPM) employed in this model. As far as we can learn from BZ, the GFPM model has two exogenous levers to simulate TPP; the first one is ad valorem tariffs that are reduced to zero and the second one is macroeconomic effects of TPP, which is mainly GDP. A few methodological concerns arise from these assumptions.
Firstly, TPP is much more comprehensive than mere tariff elimination. There are several other policy aspects in TPP, including reducing non-tariff barriers, enforcing standards in labour and environment, increasing stringent intellectual property regulations, etc. Studies such as Petri et al. (2011) cited by BZ and Narayanan et al. (2016), extensively consider these aspects. Narayanan and Sharma (2014) (NS henceforth) is a notable exception, as they only consider tariff eliminations, much like BZ do; however NS clearly characterize their study as one focusing on tariff reductions in TPP, unlike BZ.
The final version of NS is published as Narayanan and Sharma (2016). Therefore, it would have been better to mention upfront that all non-tariff aspects of TPP are beyond the scope of BZ, which is an important aspect, given that non-tariff aspects of TPP have much bigger effects than the tariff reduction.
Secondly, BZ employs NS to calibrate the macro-economic effects. NS employs a comparative static version of the Global Trade Analysis Project
(GTAP) model, while BZ is a recursive dynamic model; perhaps BZ could have employed results from Narayanan et al. (2016) instead since the latter uses a recursive dynamic computable general equilibrium model. More specifically, GFPM has a specific end year in the analysis – 2020, while NS does not have any such end year, given that it is a comparative static model.
Thirdly, since NS already includes forestry and wood products in its analysis, in aggregated sectors. Thus, these macro-economic results from NS model do have a view on forestry products implicitly. Table 2 in NS shows that the forestry sector (GTAP sector ‘frs’) is contained within the Extraction sector and wood products (GTAP sector ‘lum’) are contained within the Light Manufacturing sector. The NS study obtains its results by eliminating tariffs among TPP countries, much like what BZ does. Therefore, it may be argued that the two levers used in the GFPM model, namely tariff and macro-economic change, may contain overlaps between each other. To that extent, the results of BZ may overemphasize the effects of tariff elimination implied by TPP.
One solution to this issue, potentially, could be to use the results for these forestry-related sectors from NS, to net them out from other macro-economic effects. Since NS do not report these results for countries other than India, this is challenging. Therefore, employing studies such as Narayanan et al. (2016) or Petri et al. (2011) would have helped address this issue as well. If the idea is to capture how macro-economic changes affect forestry-related sectors, one could directly take specific results of these sectors and feed them into the GFPM model. In any case, it is important to avoid ‘double-counting’, so to say, that has been done in BZ currently.