History provides examples of large changes in the relative economic size of nations. Often these economic realignments coincide with heightened geopolitical tension. A historical example is the relative decline of the UK in the face of the emerging giants of Germany and the US in the late-19th century (Kennedy 1988). More recent examples include the relative decline of the US in the wake of rapid Japanese growth after WWII and the emergence of China into global markets at the end of the 20th century (Kleinman et al. 2024). In our recent research (Kleinman et al. 2025), we provide empirical evidence on the impact of rapid economic growth in emerging economies on the income and welfare of other nations. We examine the mechanisms through which these effects occur, and assess the relative importance of foreign productivity growth and trade integration for a country’s income and welfare compared to its own productivity growth.
We use the class of quantitative trade models with a constant trade elasticity (as considered in Arkolakis et al. 2012) to provide empirical evidence on these questions. We combine this class of models with data on bilateral foreign trade, GDP, and population over the period from 1960-2020. We treat population and the ratio of expenditure to income (and hence trade deficits) as exogenous. We treat GDP (and hence wages) and bilateral trade as endogenous. We use the equilibrium conditions of this class of models to recover unique values for domestic trade (a country’s expenditure on its own goods), productivity, and bilateral trade frictions for which the observed data are an equilibrium in this class of models (up to normalisations).
In Figure 1, we show the unweighted average of our inverse measure of bilateral trade frictions (bilateral trade openness) over time. We find a marked increase in trade integration from 1960-2020, which is particularly rapid during the era of so-called ‘hyper globalisation’ from the late-1980s to the Global Financial Crisis in 2008. From that date onwards, we find a slowdown (‘slowbalisation’) and a reverse (‘deglobalisation’) of trade integration, which is reinforced by the onset of the US-China trade war in 2018. We find that our measure of trade openness, which allows for asymmetric trade costs, differs substantially from the conventional Head-Ries Index (Head and Ries 2001), which assumes symmetric trade costs.
Figure 1 Global trade openness over time
Note: Unweighted averages of bilateral trade openness across all exporter-importer pairs in each year (excluding each country’s openness with itself which is normalized to one); bilateral trade openness recovered from observed shares of expenditure of each importer on each exporter and PPML gravity equation estimation; figure shows levels (and not logs); higher values of both openness measures correspond to lower foreign trade frictions relative to domestic trade frictions; vertical lines show 2008 (Global Financial Crisis) and 2018 (beginning of US-China Trade War).
In Figure 2, we show our measures of country productivities for China, Japan and the US, where we normalise the geometric mean of all countries’ productivities to one in each year. The US has by far the highest initial level of productivity at the beginning of our sample period. Productivity in Japan converges rapidly towards that in the US in the opening decades of our sample, before briefly overtaking US levels in the early 1990s, and then falling below US levels by the end of our sample period. China’s productivity begins substantially below levels in both countries and is only on a downward trend during the 1960s and 1970s. Following China’s domestic reforms of 1978, we find a marked acceleration in its rate of productivity growth, although its level of productivity remains substantially below that in the US at the end of our sample period.
Figure 2 Country productivities over time
Note: log country productivities recovered from the general equilibrium equality between country income and expenditure on the goods produced by that country, using our measures of bilateral trade frictions from PPML gravity estimation; domestic trade frictions normalised to one; geometric mean of country productivities normalised to one in each year.
Using our measures of bilateral trade frictions and productivities from this class of models, we undertake counterfactuals to assess the contributions of domestic productivity growth, foreign productivity growth, and trade integration to changes in countries’ shares of global GDP and welfare. Our model features four exogenous primitives: productivities, bilateral trade frictions, populations, and the ratios of expenditure to income in each country. We isolate the contributions of individual model primitives by solving for the general equilibrium of the model allowing that primitive to change over time, but holding all other model primitives constant at their values at the end of our sample period. If we allow all model primitives to change over time, we exactly reproduce the observed data; if we allow only some model primitives to change over time, the model’s counterfactual predictions in general differ from the observed data.
Higher foreign productivity has three main effects on the domestic GDP share in this class of models. First, there is a negative composition effect from an increase in the size of foreign GDP in the denominator of the domestic GDP share. Second, there is a positive market size effect from an increase in foreign income, which raises the demand for domestic goods, and hence increases the domestic wage. Third, there is a negative cross-substitution effect from lower prices of foreign goods in markets around the world, which leads consumers in those markets to substitute towards foreign goods, thereby reducing the demand for domestic goods and hence decreasing the domestic wage. Higher foreign productivity also directly reduces domestic consumer prices, thereby decreasing the domestic cost of living and raising domestic welfare. Lower trade frictions have similar effects on domestic GDP shares and welfare as higher foreign productivity, but there is now bilateral variation in the incidence of these lower trade frictions across pairs of exporters and importers. Again, reductions in trade frictions affect domestic GDP shares and welfare through composition effects, market size effects, cross-substitution, and cost of living effects.
In Figure 3, we show the results of our counterfactuals for the impact of reductions in bilateral trade frictions on the relative income and welfare (real income) of the US. We find that globalisation reduced the share of the US in global GDP (left panel), through negative terms of trade effects. Other countries experienced larger reductions in their bilateral trade frictions to markets around the world than the US, which reduced their prices relative to those of the US. In response, consumers substituted away from US goods towards those of other countries, thereby bidding down wages in the US relative to those in other countries. Nevertheless, we find that the US gained in terms of welfare (as measured by real income) from these reductions in bilateral trade frictions (right panel), highlighting that changes in countries’ relative incomes can be misleading for changes in their welfare.
Figure 3 Globalisation and the US GDP share and welfare
Note: Left panel shows US share of global GDP; right panel shows US welfare (real income); black solid line shows actual values; grey solid line with cross markers shows counterfactual predictions in which all exogenous country characteristics (population shares, proportional deficits, bilateral trade frictions and productivity) are held constant at their 2020 values, except for bilateral trade costs, which are set equal to their values from our model inversion in each year; bilateral trade frictions recovered from PPML gravity equation estimation; domestic trade frictions normalised to one.
In Figure 4, we show the results of our counterfactuals for the impact of China’s productivity growth on the relative income and welfare (real income) of the US. We find that rapid growth in Chinese productivity contributed to the observed decline in the US share of global GDP over our sample period. While this decline is partly explained by cross-substitution away from US goods in each market around world, as the increase in China’s productivity reduces the relative price of its goods, it mostly reflects the mechanical effect of China becoming bigger, as its productivity growth raises its own GDP, thereby increasing the denominator in the US GDP share. Despite this negative impact on the relative economic size of the US, we find that Chinese productivity growth raised US welfare through the resulting reduction in consumer prices and the cost of living. We find a similar pattern for the impact of Japan’s rapid productivity growth in the opening decades of our sample period on the US GDP share and welfare. This pattern of results is consistent with Paul Krugman’s (1990) aphorism that “[p]roductivity isn’t everything, but in the long run it is almost everything”. We find that the impact of foreign productivity growth on domestic welfare is relatively modest compared to that of domestic productivity growth, especially for a large country such as the US for which international trade is a relatively small share of expenditure.
Figure 4 Impact of Chinese productivity growth on US share of global GDP and welfare
Note: Left panel shows U.S. share of global GDP; right panel shows the relative change in US welfare (real income) compared to the final year of our sample; black line corresponds to data; grey line with crosses shows a model counterfactual in which all variables are held constant at their 2020 values, except for Chinese productivity which in each year equals its value from our model inversion; global GDP is the sum of the GDP of all countries included in our sample.
While our focus on the class of constant elasticity trade models implies that we could miss some of the mechanisms through which globalisation and productivity growth can affect income and welfare, this standard workhorse class of models provides an important benchmark for assessing these effects. Overall, our findings highlight that changes in countries’ relative incomes are potentially quite misleading for their welfare, and that while domestic productivity is far from everything, it plays a large role in determining domestic living standards. Although the quantitative magnitudes of the effects of globalisation and productivity growth could differ in models featuring alternative mechanisms, these two insights are of much broader applicability.
References
Arkolakis, C, A Costinot, and A Rodriguez-Clare (2012), “New Trade Models, Same Old Gains”, American Economic Review 102: 94-130.
Head, K and J Ries (2001), “Increasing Returns versus National Product Differentiation as an Explanation for the Pattern of U.S.-Canada Trade,” American Economic Review 91: 858-876.
Kennedy, P (1988), The Rise and Fall of the Great Powers, Unwin-Hyman.
Kleinman, B, S J Redding and E Liu (2024), “International Friends and Enemies,” American Economic Journal: Macroeconomics 16(4): 350-85.
Kleinman, B, E Liu, S J Redding and Z Huang (2025), “Accommodating Emerging Giants in the Global Economy,” NBER Working Paper 34530.
Krugman, P R (1990), The Age of Diminished Expectations, Washington Post Company.
