The role of digital intermediation platforms in opening up international markets

Annex B. Regression analysis: Country structural characteristics

The regression results presented above broadly align with the findings from Costa et al., (2021[5]) for online platforms across the total economy. In both cases, structural factors such as income levels and population size consistently emerge as strong predictors of platform diffusion, with higher GDP per capita and larger populations associated with greater platform traffic across sectors. This supports the idea that demand-side market characteristics play a fundamental role in shaping the uptake of digital platforms. The influence of broadband infrastructure and skills readiness appears more nuanced, showing significance only in certain models and sectors. Moreover, the negative association between domestic and foreign platform traffic in Table 1.B.2 echoes the findings from Costa et al., (2021[5]) on competitive crowding effects, reinforcing the notion that strong local platform presence can constrain foreign platform diffusion, especially when competition is intense.

Table A B.1. DIPs diffusion and structural characteristics: Total traffic

By sector, 2014-2024

		Log of Total Traffic
Sector		Travel & Accommodation Booking	Ride-Hailing & Carpooling	X2C Online Marketplace
DIP type		Pure	Pure	Pure	Pure & Hybrid
Income per capita	Log of GDP per capita (2015 USD PPP)	1.0168*** (0.3112)	0.8672 (0.6229)	0.4683 (0.5751)	1.4190*** (0.3116)
Demographics	Log of total population (millions)	1.1495*** (0.0560)	1.2162*** (0.1329)	1.3114*** (0.1088)	1.4308*** (0.0735)
Level of entrepreneurship	Share of self-employed (% total employment)	-0.0138 (0.0178)	-0.0320 (0.0252)	0.0278 (0.0204)	0.0177 (0.0187)
Infrastructure availability/use	Fixed broadband subscriptions (per 100 inhabitants)	0.0186 (0.0132)	0.0071 (0.0312)	-0.0310 (0.0244)	-0.0120 (0.0210)
Ability to leverage ICT technologies	Skills readiness (Index)	-0.0205 (0.0167)	-0.0537* (0.0301)	0.0491* (0.0247)	-0.0041 (0.0266)
	Time fixed effects (year)	Yes	Yes	Yes	Yes
	Intercept	-12.6737*** (3.1918)	-15.6190** (6.2282)	-12.4786** (6.0600)	-20.9202*** (3.7611)
	Observations	346	346	346	346
	R-squared	0.8873	0.7692	0.8180	0.8840
	Highest VIF value	3.1730	3.1730	3.1730	3.5907

Notes: Dependent variables are the log of total traffic per country year and sector. Log of GDP per capita (OECD ind.) is expressed in 2015 USD PPP; log of population in millions (World Bank ind.); share of self-employed (OECD ind.) is expressed as a percentage of total employment; fixed broadband penetration measures the number of broadband subscriptions per ten inhabitants (OECD ind.); skills readiness comes from the Networked Readiness (WEF ind.), ranging from 1 to 100 (from lower to higher skills). VIF (Variance Inflation Factor) measures multicollinearity among variables. Values below 5 indicate low collinearity; above 10 suggest problematic collinearity. Clustered standard errors at the country level are in parentheses.

Sources: Author’s elaboration based on data from Semrush and Crunchbase.

Table A B.2. DIPs diffusion and structural characteristics: Foreign firms’ traffic

Domestic market competition controlled, by sector, 2014-2024

		Log of Foreign Firms Traffic
Sector		Travel & Accommodation Booking	Ride-Hailing & Carpooling	X2C Online Marketplace
DIP type		Pure	Pure	Pure	Pure & Hybrid
Income per capita	Log of GDP per capita (2015 USD PPP)	1.5597*** (0.5191)	2.0674** (0.9230)	2.1445 (1.3865)	0.9183 (0.6327)
Demographics	Log of total population (millions)	0.9047*** (0.1178)	1.3445*** (0.2757)	1.3709*** (0.3140)	1.4346*** (0.2510)
Level of entrepreneurship	Share of self-employed (% total employment)	0.0138 (0.0250)	-0.0267 (0.0508)	0.0042 (0.0459)	-0.0253 (0.0483)
Infrastructure availability/use	Fixed broadband subscriptions (per 100 inhabitants)	0.0272* (0.0150)	-0.0166 (0.0374)	0.0773 (0.0619)	0.0180 (0.0351)
Ability to leverage ICT technologies	Skills readiness (Index)	-0.0437 (0.0292)	-0.0761 (0.0581)	-0.1464 (0.1039)	-0.0403 (0.0477)
Domestic market competition	Log of domestic firms’ traffic	-0.0822** (0.0337)	-0.1416 (0.0858)	-0.1929 (0.1146)	-0.2434*** (0.0875)
	Time fixed effects (year)	Yes	Yes	Yes	Yes
	Intercept	-12.8612** (5.8253)	-27.6797** (10.4491)	-19.7341 (12.9706)	-10.5433 (7.5182)
	Observations	243	226	231	269
	R-squared	0.7321	0.6964	0.5672	0.6174
	Highest VIF value	3.1014	3.8290	2.6898	3.2014

Notes: Dependent variables are the log of total traffic per country year and sector. Log of GDP per capita (OECD ind.) is expressed in 2015 USD PPP; log of population in millions (World Bank ind.); share of self-employed (OECD ind.) is expressed as a percentage of total employment; fixed broadband penetration measures the number of broadband subscriptions per ten inhabitants (OECD ind.); skills readiness comes from the Networked Readiness (WEF ind.), ranging from 1 to 100 (from lower to higher skills). VIF (Variance Inflation Factor) measures multicollinearity among variables. Values below 5 indicate low collinearity; above 10 suggest problematic collinearity. Clustered standard errors at the country level are in parentheses.

Sources: Author’s elaboration based on data from Semrush and Crunchbase.

Table A B.3. DIPs diffusion and structural characteristics: Foreign firms’ traffic

Domestic market competition uncontrolled, by sector, 2014-2024

		Log of Foreign Firms Traffic
Sector		Travel & Accommodation Booking	Ride-Hailing & Carpooling	X2C Online Marketplace
DIP type		Pure	Pure	Pure	Pure & Hybrid
Income per capita	Log of GDP per capita (2015 USD PPP)	1.4839** (0.5350)	1.9605* (1.0868)	2.1346 (1.3593)	0.8305 (0.6791)
Demographics	Log of total population (millions)	0.7975*** (0.0998)	1.2458*** (0.2951)	1.1144*** (0.2387)	1.0724*** (0.2313)
Level of entrepreneurship	Share of self-employed (% total employment)	0.0127 (0.0257)	-0.0286 (0.0557)	0.0034 (0.0512)	-0.0209 (0.0508)
Infrastructure availability/use	Fixed broadband subscriptions (per 100 inhabitants)	0.0328* (0.0167)	-0.0165 (0.0356)	0.0829 (0.0661)	0.0075 (0.0384)
Ability to leverage ICT technologies	Skills readiness (Index)	-0.0480 (0.0301)	-0.0837 (0.0606)	-0.1609 (0.1121)	-0.0369 (0.0495)
	Time fixed effects (year)	Yes	Yes	Yes	Yes
	Intercept	-11.2876* (5.8709)	-25.6731* (14.2212)	-17.6040 (12.4239)	-7.6815 (8.2158)
	Observations	243	226	231	269
	R-squared	0.7170	0.6728	0.5455	0.5799
	Highest VIF value	3.0954	3.7430	2.6737	3.1865

Notes: Dependent variables are the log of total traffic per country year and sector. Log of GDP per capita (OECD ind.) is expressed in 2015 USD PPP; log of population in millions (World Bank ind.); share of self-employed (OECD ind.) is expressed as a percentage of total employment; fixed broadband penetration measures the number of broadband subscriptions per ten inhabitants (OECD ind.); skills readiness comes from the Networked Readiness (WEF ind.), ranging from 1 to 100 (from lower to higher skills). VIF (Variance Inflation Factor) measures multicollinearity among variables. Values below 5 indicate low collinearity; above 10 suggest problematic collinearity. Clustered standard errors at the country level are in parentheses.

Source: Author’s elaboration based on data from Semrush and Crunchbase.

Table A B.4. DIPs diffusion and structural characteristics: Share of foreign firms’ traffic

By sector, 2014-2024

		Share of Foreign Firms’ Traffic
Sector		Travel & Accommodation Booking	Ride-Hailing & Carpooling	X2C Online Marketplace
DIP type		Pure	Pure	Pure	Pure & Hybrid
Income per capita	Log of GDP per capita (2015 USD PPP)	-0.0054 (0.0154)	-0.0443 (0.1002)	0.0601 (0.0618)	-0.0253 (0.0296)
Demographics	Log of total population (millions)	-0.0130*** (0.0092)	-0.0527 (0.0490)	-0.0046 (0.0119)	-0.0118 (0.0087)
Level of entrepreneurship	Share of self-employed (% total employment)	0.0013 (0.0008)	0.0055 (0.0050)	-0.0026 (0.0026)	-0.0020 (0.0023)
Infrastructure availability/use	Fixed broadband subscriptions (per 100 inhabitants)	0.0011 (0.0007)	0.0030 (0.0049)	0.0034 (0.0026)	0.0005 (0.0018)
Ability to leverage ICT technologies	Skills readiness (Index)	-0.0003 (0.0009)	-0.0025 (0.0034)	-0.0093 (0.0056)	-0.0015 (0.0022)
	Time fixed effects (year)	Yes	Yes	Yes	Yes
	Intercept	1.2069*** (0.1461)	2.2691 (1.6520)	0.9214* (0.5365)	1.5167*** (0.3510)
	Observations	346	346	346	346
	R-squared	0.3382	0.1984	0.2306	0.1276
	Highest VIF value	3.1730	3.1730	3.1730	3.1730

Notes: Dependent variables are the log of total traffic per country year and sector. Log of GDP per capita (OECD ind.) is expressed in 2015 USD PPP; log of population in millions (World Bank ind.); share of self-employed (OECD ind.) is expressed as a percentage of total employment; fixed broadband penetration measures the number of broadband subscriptions per ten inhabitants (OECD ind.); skills readiness comes from the Networked Readiness (WEF ind.), ranging from 1 to 100 (from lower to higher skills). VIF (Variance Inflation Factor) measures multicollinearity among variables. Values below 5 indicate low collinearity; above 10 suggest problematic collinearity. Clustered standard errors at the country level are in parentheses.

Sources: Author’s elaboration based on data from Semrush and Crunchbase.