Tamberi, N (2026) The elasticity of imports to MFN tariffs, Centre for Inclusive Trade Policy, Working Paper 032
" } ["group_access"]=> array(4) { ["apply_restriction"]=> bool(false) ["role_access"]=> array(2) { [0]=> string(13) "administrator" [1]=> string(9) "developer" } ["hide_links_public"]=> bool(false) ["hide_links_unselected"]=> bool(false) } ["clone_access"]=> array(1) { ["group_access"]=> array(4) { ["apply_restriction"]=> bool(false) ["role_access"]=> array(2) { [0]=> string(13) "administrator" [1]=> string(9) "developer" } ["hide_links_public"]=> bool(false) ["hide_links_unselected"]=> bool(false) } } ["use_contents"]=> bool(false) ["layouts"]=> array(2) { [0]=> array(2) { ["acf_fc_layout"]=> string(11) "layout_text" ["clone"]=> array(11) { [""]=> NULL ["group_bespoke"]=> array(3) { ["style"]=> string(7) "default" ["column_num"]=> string(1) "1" ["width_2_column"]=> string(8) "split-50" } ["group_column_1"]=> array(2) { ["text"]=> string(962) "When the UK left the EU, it introduced a new import tariff schedule, one of the largest MFN tariff reforms in decades of a developed economy. This paper exploits this policy change to estimate the elasticity of imports to MFN tariffs, addressing the endogeneity of trade policy through two strategies: a country-based approach using trade agreement partners whose MFN exposure is plausibly exogenous, and a product-based approach leveraging mechanical tariff changes from schedule simplification. Using a difference-indifferences design with EU imports as a control group, I find that ignoring endogeneity yields an elasticity of 2.5, while IV estimates range from 4.5 to 8.7. Dynamic responses are even larger: three years after the reform, elasticities reach 10-12. These results suggest that widely used tariff-based elasticities are substantially underestimated, with implications for quantitative trade models and welfare analysis.
" ["align"]=> string(7) "align_l" } ["group_column_2"]=> array(2) { ["text"]=> string(0) "" ["align"]=> string(7) "align_l" } ["_copy"]=> NULL ["group_column_3"]=> array(2) { ["text"]=> string(0) "" ["align"]=> string(7) "align_l" } ["group_core"]=> array(8) { ["layout_label"]=> string(8) "Abstract" ["tog_progress"]=> bool(false) ["tog_header"]=> bool(true) ["tog_padding"]=> bool(true) ["tog_background"]=> bool(false) ["tog_restriction"]=> bool(false) ["grp_lifecycle"]=> array(2) { ["datetime_start"]=> string(0) "" ["datetime_end"]=> string(0) "" } ["grp_identity"]=> array(2) { ["id"]=> string(0) "" ["class"]=> string(0) "" } } ["group_header"]=> array(3) { ["title"]=> string(22) "CITP Working Paper 032" ["intro"]=> string(0) "" ["align"]=> string(7) "align_l" } ["group_padding"]=> array(2) { ["top"]=> string(1) "1" ["bottom"]=> string(1) "0" } ["group_background"]=> array(2) { ["group_bg_colour"]=> array(2) { ["bg_colour"]=> string(7) "bg_none" ["bg_colour_pct"]=> int(100) } ["bg_image"]=> string(10) "bgimg_none" } ["group_restriction"]=> array(1) { ["role_access"]=> array(2) { [0]=> string(13) "administrator" [1]=> string(9) "developer" } } } } [1]=> array(2) { ["acf_fc_layout"]=> string(11) "layout_text" ["clone"]=> array(11) { [""]=> NULL ["group_bespoke"]=> array(3) { ["style"]=> string(7) "default" ["column_num"]=> string(1) "1" ["width_2_column"]=> string(8) "split-50" } ["group_column_1"]=> array(2) { ["text"]=> string(3892) "When the UK left the European Union in January 2021, it introduced the UK Global Tariff (UKGT) – the first full overhaul of its independent tariff schedule in nearly fifty years. The reform was unusually large for a developed economy: it reduced Most Favoured Nation (MFN) tariffs on two thirds of products, simplified the tariff structure, and expanded the share of zero tariff lines from 27% to 47%. This major policy shift offers a rare opportunity to understand how imports respond to MFN tariff changes, a question central to the calibration of quantitative trade models and the evaluation of trade policy.
Estimating the elasticity of imports to tariffs is challenging because tariff rates are not set randomly. They often reflect political economy pressures, sectoral competitiveness, and concerns about protecting domestic producers. This endogeneity can bias standard regression estimates toward zero. The UKGT, however, provides two sources of plausibly exogenous variation that help overcome these concerns.
The first strategy is country based. It exploits the behaviour of countries that had Preferential Trade Agreements (PTAs) with the EU before Brexit. These partners often continue to face MFN tariffs because many firms do not fully use their preferential access. Since the UK’s MFN schedule in 2021 was not designed around imports from PTA partners, MFN tariff changes affecting these countries can be treated as exogenous. Moreover, the degree to which each PTA partner is exposed to MFN tariffs varies widely due to differences in preference utilisation. This variation provides rich identifying power.
The second strategy is product based. Many tariff changes under the UKGT resulted mechanically from tariff simplification rules – eliminating “nuisance” tariffs below 2.5% and rounding down rates into standardised bands. These mechanical changes depend only on the initial tariff level and not on product specific economic shocks, making them another source of quasi random variation.
Using a difference in differences design that compares UK import changes to those of EU member states (which did not change their MFN schedule), the paper shows that naïve estimates significantly understate the tariff elasticity. Ignoring endogeneity produces an elasticity of about -2.5. In contrast, instrumental variable estimates exploiting either strategy are much larger: -5.4 to -8.7 in static specifications.
Importantly, the paper documents strong dynamic adjustment. Three years after the reform, import responses intensify, and the implied tariff elasticity reaches 10-12. Robustness tests focusing on smaller PTA flows and non Generalised Scheme of Preferences partners suggest that even these large estimates may represent a conservative lower bound.
These results challenge the widely used benchmark elasticity of around five. They imply that most existing studies – particularly those relying on PTA based tariff variation without accounting for partial preference utilisation or policy endogeneity – substantially underestimate how responsive trade flows are to tariff changes. Taken together, the evidence indicates that the true trade elasticity is considerably larger, and that quantitative trade models calibrated using traditional estimates may understate the impact of tariff reforms on trade and welfare.
The UKGT thus provides compelling new evidence that tariff based elasticities should be revised upward, reinforcing the need for empirical approaches that properly account for multiple tariff regimes, utilisation behaviour, and the dynamics of adjustment.
" ["align"]=> string(7) "align_l" } ["group_column_2"]=> array(2) { ["text"]=> string(0) "" ["align"]=> string(7) "align_l" } ["_copy"]=> NULL ["group_column_3"]=> array(2) { ["text"]=> string(0) "" ["align"]=> string(7) "align_l" } ["group_core"]=> array(8) { ["layout_label"]=> string(21) "Non Technical Summary" ["tog_progress"]=> bool(false) ["tog_header"]=> bool(false) ["tog_padding"]=> bool(false) ["tog_background"]=> bool(false) ["tog_restriction"]=> bool(false) ["grp_lifecycle"]=> array(2) { ["datetime_start"]=> string(0) "" ["datetime_end"]=> string(0) "" } ["grp_identity"]=> array(2) { ["id"]=> string(0) "" ["class"]=> string(0) "" } } ["group_header"]=> array(3) { ["title"]=> string(0) "" ["intro"]=> string(0) "" ["align"]=> string(7) "align_l" } ["group_padding"]=> array(2) { ["top"]=> int(1) ["bottom"]=> int(1) } ["group_background"]=> array(2) { ["group_bg_colour"]=> array(2) { ["bg_colour"]=> string(7) "bg_none" ["bg_colour_pct"]=> int(100) } ["bg_image"]=> string(10) "bgimg_none" } ["group_restriction"]=> array(1) { ["role_access"]=> array(2) { [0]=> string(13) "administrator" [1]=> string(9) "developer" } } } } } }