Trade liberalisation and taxation

Rabah Arezki, Alou Adesse Dama, Grégoire Rota-Graziosi 17 August 2021



Conceptually, trade liberalisation is expected to raise economic growth and welfare through more competition reducing domestic rents, raising the availability of capital and increased productivity. The empirical evidence on the relationship between trade orientation and growth in developing countries has been hotly contested for over half a century (Baldwin 2003, Rodriguez and Rodrik 2000). There are several possible reasons for why that is so. 

Trade liberalisation has not done away with pervasive distortions to trade, especially in developing countries because of, among other things, limited enforcement and implementation of trade policies (Atkin and Khandelwal 2020). In turn, these distortions might limit the expansion of trade and its effect on welfare and growth. Trade liberalisation also has important distributional consequences by raising income inequality (Goldberg and Pavcnik 2004). Trade liberalisation may have an ambiguous effect on (aggregate) informality (Ulyssea 2018). 

One important related question is how trade liberalization affect public finances. The public finance channel of trade liberalisation has received less attention, yet it may have important economic and distributional implications. In a recent paper (Arezki et al. 2021), we systematically study the dynamic effects of trade liberalisation on taxation.


There are several mechanisms through which trade liberalisation can affect taxation. Trade liberalisation can lower economic growth for some developing countries and hence limit the expansion of domestic tax revenues. Notwithstanding the contention over the effect of trade liberalisation on growth, tariff revenue losses following trade liberalisation can be hard to replace with domestic sources. In other words, the ‘tax transition’ – that is, the rebalancing away from tariff revenues (raised on transactions related to international trade) towards domestic revenues (levied on the domestic base) – is often incomplete. Indeed, structural factors such as natural resources dependence, informality and weak state capacity are particularly pervasive in developing countries, in turn limiting domestic revenue mobilisation (Besley and Persson 2014). 

Moreover, the decrease in the cost of international capital mobility has fuelled intense tax competition, in turn offering multiple opportunities for multinationals to shift profits to tax-accommodating countries known as tax havens. The tax structure may also be impacted by trade liberalisation. Tax coordination and enforcement at the international level are difficult, and coalitions in support of tax coordination appear theoretically unstable (Rota-Graziosi 2019). Developing countries are particularly vulnerable to trade mis-invoicing or profit shifting by multinational corporations following trade liberalisation because of relatively weak state capacity. 

Dynamic effects 

In our research, we specifically focus on the dynamic effect of trade liberalisation on non-resource tax revenue using a combination of tax revenue data from Government Revenue Dataset (GRD) of the ICTD/UNIWIDER and the extended version of the Sachs and Warner indicator of liberalisation by Wacziarg and Welch (2008) [hereafter ‘SW/WW’]. 

Unlike the existing cross-country studies of trade orientation and taxation, we use a policy-based measure of trade orientation – allowing us to pin down the exact timing of change in trade policy – instead of outcome measures of trade such as trade-to-GDP. Our results point to a lack of robustness in the positive association between outcome-based measure of liberalisation and tax revenues. We document a significant and robust negative effect of trade liberalisation on tax revenues using a policy-based measure of liberalisation. 

Our research also distinguishes between tax revenue from natural resources versus tax collected outside the resource sector. The distinction between the two sources of revenue is important because of large fluctuations in international prices in natural resources – in turn increasing total tax revenue without reflecting any improvement in domestic tax policy. Indeed, when using non-resource tax revenues as dependent variable instead of aggregate tax revenues, the positive association between the outcome-based measure of trade liberalization and tax revenues is no longer robust.

Importantly, our research contributes to the literature by exploring more systematically the dynamic effect of trade liberalisation and the evolution of tax structures. Results point to a significant negative effect of liberalisation on (non-resource) tax revenues in the short term and to no significant effect in the medium term (see Figure 1). 

Figure 1 Dynamic effects of trade liberalisation on non-resource tax revenues

A) Without controls    

B) With controls

Notes: The graph shows impulse response functions. We regress non resource tax revenues as share of GDP on SW/WW liberalization indicator as well as several controls unless otherwise indicated. 95%/90% error bands are displayed. Panel A presents impulse responses with no control variable. Panel B presents impulse response with control variables including real GDP growth, agriculture value added, inflation and net development aid and assistance.

Our research also shows that trade liberalisation alters the tax structure, tilting revenues towards indirect taxes and away from direct ones. Such shift in the structure reduces the redistributive capacity of national tax systems. Additional results point to robust negative and statistically significant effect of trade liberalisation on corporate income tax (CIT) revenues. Trade liberalisation increases competition and welfare, which reduces rents and taxable income of domestic corporations. Trade liberalisation also allows multinational companies to engage in more aggressive tax planning, which affects CIT revenues while being welfare worsening for the source country. The increase in capital mobility and the advent of tax havens make the technology of tax avoidance more pervasive. 

Trade liberalisation and VAT

Following its introduction in 1954, VAT quickly became a major tax. VAT has been adopted by more than 160 countries. Keen and Lockwood (2010) estimated that VAT constitutes more than 20% of world’s tax revenue. In most countries over 50% of the VAT revenues are collected at the border, hence its importance in the analysis of trade liberalisation. The implementation of VAT also offers an opportunity to modernise tax administration. This is because with VAT in place, tax and customs administrations must cooperate, and even end up merging into a single revenue agency. VAT paid at the border to the customs administration is a definitive tax revenue when it is collected on non-VAT-liable firms (encompassing informal firms), otherwise it is deductible from VAT collected on the domestic market by the tax administration. VAT improves the effective implementation of tax laws because of the necessary and active gathering information about business-to-business interactions.

In our research, we explore whether VAT mediates the relationship between trade liberalisation and tax revenues. To do so, we augment our main specification with an interaction term between trade liberalization episodes and VAT adoption as an additional independent variable (see Table 10 in Arezki et al. 2021). The interaction allows us to test specifically whether countries which have implemented VAT prior to liberalisation have fared better in terms of tax revenue mobilization.

VAT appears to play an important role in mediating the effect of liberalisation on tax revenues. The regression results presented in Table 10 of our paper adjust tax revenues for natural resource rents. Across all columns the interaction terms associated between trade liberalisation and VAT implementation are positive and significance pointing to important role of VAT in mitigating the negative effect of liberalisation on tax revenues. 

Quantitatively, the interaction effect is large. Take two countries experiencing liberalisation and differing only in whether they have implemented VAT prior to liberalisation. According to the estimated coefficient presented in Column 1 in Table 10 of our paper, the country without VAT would face a loss in tax revenues of -1.952% of GDP, while the country with VAT will face a loss of (only) -0.371% of GDP (= -1.952   +   1.581). These results are in line with Ebrill et al. (2001) and Keen and Lockwood (2010), who suggest a revenue gain associated with the adoption of VAT. 

Our results here more specifically suggest that having implementing VAT prior to liberalisation allows countries to perform better in terms of tax revenue mobilisation everything else being equal. 

Countries which did not have VAT before openness are the ones that suffer the most in term of revenue lost. In other words, setting up domestic tax sources may ease the revenue loss associated with liberalisation. VAT implementation helps the domestic tax system ‘cope’ with trade liberalisation. 

Concluding remarks

Our research points to significant negative effects of liberalisation on (non-resource) tax revenues in the short term and no significant effect in the medium term. Liberalisation also alters the tax structure, tilting revenues towards indirect taxes away from direct ones. Economies which have implemented VAT prior to liberalisation have mitigated its negative effects on tax revenues. The evidence is supportive of the complementary role of state capacity to reap the benefits of liberalisation. From an operational standpoint, implementing VAT first and tariff cutting second can help move us away from the old ‘cut-or-not’ debate.

Further research could look at the issue of tax spillovers stemming from membership of a customs or monetary union. Indeed, many countries around the world belong to such unions, which entails deeper economic integration and in turn leads to important spillovers from tax policy. A network analysis of tax spillovers in these unions could allow the identification of weakest links – that is, countries that expose the other union members to risks of indirect or direct tax base erosion. This analysis would rely on the existence of tax treaties that link countries inside unions to tax havens, with smaller countries more likely to engage in aggressive tax competition. 


Arezki, R, A A Dama and G Rota-Graziosi (2021), "Revisiting the Relationship between Trade Liberalization and Taxation", FERDI Working Paper P293, June.

Atkin, D, A K and Khandelwal (2020), “How distortions alter the impacts of international trade in developing countries”, Annual Review of Economics 12: 213-238.

Baldwin, R E (2003), “Openness and Growth: What's the Empirical Relationship?”, NBER Working Paper 9578. 

Besley, T and T Persson (2014), “Why do developing countries tax so little?”, Journal of Economic Perspectives 28(4): 99-120.

Ebrill, M, M Keen and V Perry (2001), The modern VAT, International Monetary Fund.

Goldberg, P and N Pavcnik (2004), “Trade, Inequality, and Poverty: What Do We Know? Evidence from Recent Trade Liberalization Episodes in Developing Countries”, NBER Working Paper 10593.

ICTD/UNI-WIDER – International Center for Tax and Development/The United Nations University World Institute for Development Economics Research (2017), “Government Revenue Dataset”. 

Keen, M and B Lockwood (2010), “The value added tax: Its causes and consequences”, Journal of Development Economics 92(2): 138-151.

Rodriguez, F and D Rodrik (2000), “Trade policy and economic growth: A skeptic's guide to the cross-national evidence”, NBER Macroeconomics Annual 15: 261-325.

Rota-Graziosi, G (2019), “The supermodularity of the tax competition game”, Journal of Mathematical Economics 83: 25-35.

Ulyssea, G (2018), “Firms, informality, and development: Theory and evidence from Brazil”, American Economic Review 108(8): 2015-47.

Wacziarg, R and K H Welch (2008), “Trade liberalization and growth: New evidence”, The World Bank Economic Review 22(2): 187-231.



Topics:  Development International trade Taxation

Tags:  trade liberalisation, VAT, tax revenues, developing countries

Chief Economist & Vice President, African Development Bank & Senior Fellow, Harvard’s Kennedy School of Government

Ph.D Candidate, CERDI

Professor of Economics and Director, CERDI-CNRS-UCA


CEPR Policy Research