The digitalisation of trade documents and processes

Annex A. Estimating impacts of going paperless

The gravity model of trade expresses trade flows as a function of the (economic) size of the trading countries and trade costs. A generic sector-specific structural gravity equation can be expressed as:

${\left(X_{ij}\right)}^k=\frac{Y_i^k{E}_j^k}{Y^k}{\left(\frac{T_{ij}^k}{{{\mathrm{\Pi }}_i^{k}P}_j^k}\right)}^{-{Ѳ}^k}$ (1)

where trade flows from country i to country j in sector k, $X_{ij}^k$, are a function of the supply of sector k-goods from country i, $Y_i^k$, and expenditure for sector k-goods in country j, $E_j^k$. $T_{ij}^k>1$are trade costs when sector k-goods are shipped from exporter-country i to importer-country j. ${k }^{}$is the sector-specific trade elasticity, and ${\mathrm{\Pi }}_i^k$and $P_j^k$are the price indices representing outward and inward multilateral resistance terms, respectively. The size term is captured by $\frac{Y_i^k{E}_j^k}{Y^k}$ and shows the hypothetical level of frictionless trade between two countries, which is proportional to their overall share of global economic activity. The trade cost term, $\frac{T_{ij}^k}{{{\mathrm{\Pi }}_i^{k}P}_j^k}$, is a scaling factor that takes into account trade frictions.

Trade flows specifications

The analysis looks at the impact on trade of the different policy elements identified in Section 3. This involves estimating trade flows in the following structural gravity model framework, where $X_{ijt}^k$ are the exports from country i to country j in sector k (including both cross-border trade and domestic trade flows):

$X_{ijt}^k=\exp \left({{{\beta }_{1}Paperless\_policies}_{ijt} }_{}+{{\beta }_{2}RTA\_depth}_{ijt}{+\sum _t^{}{b}_t+\mathrm{ }\mathrm{\eta }}_{it}^k+{\mathrm{\mu }}_{jt}^k+{\nu }_{ij}+s^k\right)*{\mathrm{\epsilon }}_{ijt}^k$ (2)

Paperless policies (${Paperless\_policies}_{ijt}$) reflect the elements covered by section 3:

• TFIs – indicators on automation (TFI G); indicators on automation and documents (average of TFI G and TFI F); indicators on automation, documents, procedures and border agency co-operation (average of TFIs G, F, H, I and J). A bilateral index is constructed based on the minimum between the exporter and importer performance.

When employed in the regression alternatively, the TFIs are introduced as natural logarithm (TFIs range between 0 and 2).

• DSTRI – includes the selected areas of infrastructure and connectivity; electronic transaction frameworks; and e-payments (their value ranges between 0 and 1). A bilateral index is constructed based on the maximum between the exporter and importer scores.

• INDIGO-t – includes the bilateral INDIGO to control for trade-related discussions relevant to digital trade (the bilateral INDIGO-t ranges between 0 and 28).

The variables are interacted with a border dummy to account for their impact on cross-border trade (equal to 1 when the trade flow is international, as opposed to domestic).

The specifications control for the presence and depth of an RTA (${RTA\_depth}_{ijt}$). This represents the depth of an RTA between countries i and j in year t and can take values from 0 (no trade agreement in force) to 52 (where all possible broad policy areas are included in the trade agreement). Information on the depth of RTAs is obtained from the World Bank Deep Integration Dataset (Mattoo, Rocha and Ruta, 2020[32]).

$\sum _t^{}{b}_t$ are international border-year fixed effects, which take a value of 1 in year t if the trade flow is international. These terms are meant to control for heterogeneity in unobservable bilateral international trade costs. Including these controls should help better capture changes in the costs of international trade relative to domestic trade.

${\mathrm{\eta }}_{it}^k$ represent exporter-sector-year fixed effects; ${\mathrm{\mu }}_{jt}^k$ represent importer-sector-year fixed effects; ${\nu }_{ij}$ represent exporter-year fixed effects; and $s^k$ are sector fixed effects.

Trade flows specifications are run using PPML with high dimensional fixed effects. PPML (Poisson Pseudo Maximum Likelihood) allows to account for hetereoscedasticity and for zero trade flows.

Trade data

USITC ITPD-E database

The United States International Trade Commission (USITC) International Trade and Production Database for Estimation (ITPD-E) contains data on international and domestic trade for 243 jurisdictions, 170 sectors, and 17 years (Borchert et al., 2022[22]).

This allows to cover in the analysis here 19 low-income economies, 46 lower-middle income economies, 41 upper-middle income economies, and 55 high-income and OECD economies (Table A B.1). The sectors available in ITPD-E are matched to 18 sectors in 2-digit ISIC Rev. 4 level classification (Table A A.1).

Table A A.1. Sector coverage based on the ITPD-E database

ISIC sector code	ISIC sector name
D01T02	Agriculture, hunting and forestry
D03	Fishing and aquaculture
D05T09	Mining and quarrying, energy production products
D04	Mining and quarrying, non-energy production products
D10T12	Food products, beverages and tobacco
D13T15	Textiles, wearing apparel, leather and related products
D16	Wood and products of wood and cork
D17T18	Paper products and printing
D19	Coke and refined petroleum products
D20	Chemical and chemical products
D21	Pharmaceuticals
D22	Rubber and plastics products
D23	Other non-metallic mineral products
D24	Basic metals
D25	Fabricated metal products
D26	Computer, electronic and optical products
D27	Electrical equipment
D28	Machinery and equipment n.e.c.
D29	Motor vehicles, trailers, and semi-trailers
D30	Other transport equipment
D31T33	Manufacturing nec; repair and installation of machinery and equipment