Skip to main content


Log in

The world income distribution: the effects of international unbundling of production

  • Published:
Journal of Economic Growth Aims and scope Submit manuscript


Trade in intermediates, also known as unbundling of production, and trade in capital have become increasingly important in the world economy during the last 25 years. To jointly study these two phenomena, we develop a dynamic, factor-proportions model with trade in final goods, intermediates and capital where countries differ in their aggregate productivity levels (TFP). Our central result is to identify a novel channel whereby trade in intermediates generates a reallocation of capital across countries that exacerbates world inequality in both income and welfare. With unbundling, high-productivity countries sort into the production of capital-intensive intermediates. They increase their capital stock (via capital imports and accumulation), and, ultimately, their real wages. This exacerbates initial productivity differences across countries and increases world income inequality. We also show that income inequality rises with unbundling (i) even in the case of ex-ante identical countries (symmetry breaking), (ii) when emerging countries start participating in trade in intermediates and (iii) when a labor-saving technology (computerization) is introduced. For an empirically-motivated model parametrization, middle-income countries experience the largest output decline with unbundling.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Source: Feenstra World Trade Database. To classify goods as intermediates, we use the end-use classification of Feenstra and Jensen (2012). Southern countries are defined as countries with GDP per capita (PPP) lower than 50% of the United States in 2000

Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others


  1. Antràs (2016), Johnson (2014) and Hummels et al. (2001), among others, provide evidence consistent with an increase in trade in intermediate inputs over time.

  2. We can reject the null of identical dispersion at a 5% level using the one-sided F-test for the ratio of variances. We find similar results if we define final goods as bundles of intermediates using direct requirements input-output tables (see Table A.1 in the online appendix).

  3. Table B.1 (in the online appendix) provides evidence consistent with this pattern of specialization in intermediates. Moreover, it shows that it is quantitatively important. We find that, if the productivity of a country moved from the 25th to the 75th percentile, the rise in the value of exports of intermediates in the 75th percentile of capital intensity would be 40% larger than the increase in the 25th percentile. See also, among others, Baxter and Kouparitsas (2003), Hanson (2012) and Schott (2003a, b, 2004).

  4. This production structure has been subsequently used to model trade in intermediates both in the international trade literature, e.g., Yi (2003), and the international macroeconomics literature, e.g., Jin (2012).

  5. Online Appendix C provides an exact microfoundation that delivers this result as an endogenous outcome.

  6. The online supplemental material A (available at the authors’ websites) considers the case in which only a fraction of intermediates can be traded.

  7. Even though output inequality emerges in equilibrium, consumption (and welfare) remains the same across countries. The reason is that the asymmetric allocation of capital does not change factor prices. Thus, total factor payments remain the same in all countries.

  8. Recent working papers by Reyes-Heroles (2017) and Sposi et al. (2017) have analyzed the interaction between Ricardian trade and capital flows driven by changes in trade costs. These models abstract from heterogeneity in capital intensity across intermediates, which is at the heart of our results.

  9. The sequential production process in which intermediates are first produced and then used to assemble each variety pioneered in Dixit and Grossman (1982) is also assumed in, among others, Antràs and Chor (2013), Caliendo and Parro (2012), Deardorff (1998, 2001a) and Kohler (2004).

  10. We note that for the case in which countries have identical productivities, if we eliminated the varieties stage of production, we would have a standard (dynamic) Heckscher-Ohlin model. In this case, we can also derive the unique steady-state world income distribution (using the symmetry breaking equilibrium refinement). This is in contrast with the indeterminacy of equilibria in the case of a finite number of traded goods (e.g., Bajona and Kehoe 2010). We focus on the case of trade in intermediates because they are more heterogeneous in capital intensity.

  11. In terms of techniques, the characterization of the unbundling equilibrium relies in solving for the equilibrium assignment in a similar manner to Matsuyama (2013). Our solution differs from the Ricardo-Roy assignment models as used in Costinot and Vogel (2010) and Grossman and Helpma (2014) because our production functions are not linear and factors of production are homogeneous.

  12. The online appendix is available from the journal’s or authors’ websites.

  13. The microfoundation is based on Jones (1995). There exists an innovation sector that produces varieties using the final good. Innovators sell patents to the producer of varieties and extract all the surplus of the producer, who has monopoly rights on the production of the variety. With this microfoundation, the Armington assumption is not needed as firms endogenously choose to produce different products. We also note that this particular microfoundation for \(\mu _{j}\) plays a simplifying role. However, the mechanism driving our results is independent of this microfoundation. We could carry \(\{ \mu _{j}\}_{j=1}^J\) as additional parameters or simply set \( \mu _j=1\) for all j and obtain the same qualitative results.

  14. This formulation has been used, among others, by Dornbusch et al. (1980) in the context of Heckscher-Ohlin models.

  15. This assumption is made to simplify the exposition of the results. In the next section, we show that it corresponds to the steady-state of the equilibrium without unbundling. Importantly, our results on changes of relative GDP per capita and welfare do not depend on this assumption. See also footnote 18.

  16. If countries had different labor endowments, we would obtain the same qualitative results. In this case, all variables would need to be interpreted in per capita terms. In particular, \(\theta _{j}\) would be the number of varieties per capita (i.e., \(\mu _{j}=\kappa \theta _{j}L_{j})\). Under this assumption, we can compute the labor market clearing condition following the same steps as in the main text and check that the same exacerbation results hold (even though the exact value of the thresholds governing the endogenous selection of intermediates in the unbundling equilibrium would change). For example, we can use an analogous argument to the one below to show that unbundling increases real wage differences.

  17. In Section A in the online supplemental material (available at the authors’ websites), we consider the intermediate case in which only a fraction \(\alpha \) of the intermediates is traded and perform comparative statics on \(\alpha \).

  18. These results are a direct application of the Ramsey model, see Section A in Caselli and Ventura (2000). Note that \(V_{j}=\int _{0}^{\infty }\frac{1}{1-\gamma }c_{j}(t)^{1-\gamma }e^{-\rho t}dt=\int _{0}^{\infty }\frac{1}{1-\gamma }\left[ c_{j}(0)e^{\int _{0}^{t}\frac{r(v)-\rho }{\gamma }dv}\right] ^{1-\gamma }e^{-\rho t}dt\). Thus, \(\frac{V_{j}}{V_{i}}=\left[ \frac{c_{j}(0)}{c_{i}(0)} \right] ^{1-\gamma }\).

  19. If the initial asset distribution were different, the trade balance would also be zero in the steady-state because \( b_{j}^{R}=\pi _{j}\) in the steady-state (but generically would be non-zero along the transition path).

  20. We are implicitly assuming that each intermediate is produced only by one country, which is indeed true almost everywhere in equilibrium.

  21. In the case that \(\theta _{1}=\theta _{2}=\cdots =\theta _{J}\), this is still an equilibrium but there exist other equilibria. See Sect. 3.1.

  22. If we had not made the assumption that each country has an initial stock of capital proportional to its productivity, capital flows would be different than zero in the equilibrium without unbundling. In this case, the results discussed in this paragraph would simply imply that high productivity countries import relatively more capital in the unbundling equilibrium.

  23. Section B in the online supplemental material (available at the authors’ website) shows that this result does not depend on the Cobb-Douglas assumption in the production function for intermediates. We show that it extends to a general CES production function.

  24. Top-bottom inequality is the ratio between the 90th percentile of GDP per capita (PPP) and the 10th percentile. We choose to finish in 2008 to exclude the effects of the Great Recession. If we include all available information (up to 2014), the top-bottom GDP inequality is 27.1 in the post-1990 period. The same qualitative results hold if we define top-bottom inequality as the ratio between the 95th and 5th percentile.

  25. We discuss in more detail this refinement concept in Section F of the online supplemental material (available at the authors’ websites).

  26. The first derivative is proportional to \(j^{-1/2}-(j-1)^{-1/2}\), which is negative for \(j>1\). The second derivative is proportional to \( -j^{-3/2}+(j-1)^{-3/2}\), which is positive for \(j>1\).

  27. World output with and without unbundling is \(Y^{\text {World}}=2\theta \left( BJ\right) ^{ \frac{1}{2}}\).

  28. With two ex-ante identical countries H and F,  the threshold equilibrium is \(z_{1}=1-\sqrt{1/2}\). It is readily verified that this equilibrium implies that the relative world production share of country H becomes\(\left( \frac{s_{H}^{y}}{s_{F}^{y}}\right) ^{with}=\frac{ Z_{H}}{Z_{F}}=\frac{1-z_{1}}{z_{1}}=\frac{1}{\sqrt{2}-1}>1=\left( \frac{ s_{H}^{y}}{s_{F}^{y}}\right) ^{without}\).

  29. There exists an analogous equilibrium that corresponds to the permutation of H and F, \(\theta _{F}=\theta _{H}+\varepsilon \), \(\varepsilon \rightarrow 0^{+}\), which corresponds to point B in the figure.

  30. Section A in the online supplemental material (available at the authors’ websites) shows that the symmetry breaking result also extends to the intermediate case in which only a fraction \(\alpha >0\) of intermediates are traded.

  31. From the equilibrium assignment in the discrete case, we know that more productive countries specialize in capital-intensive (higher index z) intermediates, \(z^{\prime }(j)<0\). Thus, \(\theta ^{\prime }(j)z^{\prime }(j)>0\). Rearranging (16),we find that z(j) is convex, as \( z^{\prime \prime }(j)=(1-z(j))^{-1}(\theta ^{\prime }(j)z^{\prime }(j)/\theta (j)+z^{\prime 2}(j))>0\).

  32. This fit is better than a “power law” where we regress the log of TFP or country GDP per capita on the log of the country ranking. In this case, the log-log regression yields an \(R^{2}\) of 0.86 and 0.69, respectively, see Table I.1 in the online appendix for the estimation results. We can also compute the solution of the differential equation for the power-law case, i.e., \(\theta (j) \propto j^{-\upsilon }\) for some \(\upsilon >0\). In this case, the solution is not more tractable than with an exponential. The choice of 1988 is given by our data source, Hall and Jones (1999), which report TFP data for this year. We choose this source because it roughly coincides with the increase in trade in intermediates documented in Fig. 1a.

  33. We would obtain the same differential equation if we allowed the constant in equation (17) to be different from the exponential decay \(\lambda \), i.e., \(\theta (j)={\tilde{\lambda }} e^{-\lambda j}\). Using \({\tilde{\lambda }} =\lambda \) allows us to economize on notation in the computation of the Lorenz curves below.

  34. To see this, rewrite the Lorenz curves in terms of the assignment j(z), (18), so that \(L(z)^{without}=ze^{1-z}\) and \( L(z)^{with}=z\), and the result follows.

  35. In this case, the ratio \(s^{y,\text {with}}(j)/s^{y,\text {without}}(j)\) is a monotonically decreasing function, positive for j lower than a threshold and negative thereafter. This means that there is a strict ranking in the percentage increase of world output share, being highest for the most productive country and decreasing thereafter.

  36. For the case of a discrete number of countries, denoting by \(\eta _{\omega }\) the amount of varieties produced by southern countries (i.e., \(\eta _{\omega }\)\(=\sum _{\omega ={\underline{j}}}^{J}\mu _{\omega })\), the trade balance of northern countries j becomes

    $$\begin{aligned} \frac{\mu _{\omega }}{N}\left( Y-y_{\omega }\right) +Z_{\omega }\frac{N-\eta _{\omega }-\mu _{\omega }}{N} Y=\frac{N-\mu _{\omega }}{N}y_{\omega }+(1-Z_{\omega })\frac{\mu _{\omega }}{N}Y. \end{aligned}$$

    Rearranging, \(s_{\omega }^{before}=Z_{\omega }^{before}\left( 1-\frac{\eta _{\omega }}{N} \right) \). Taking the limit to a continuum of countries, the expression becomes (20).

  37. This condition has been applied in other economic contexts, see Hopkins and Kornienko (2004) and the references therein. The normal, uniform and exponential distribution among other distributions satisfy this condition.

  38. Moreover, they also show that a Monotone Likelihood Ratio (MLRP) order implies MPR. Thus, MPR is more stringent than first-order stochastic dominance but less stringent than MLRP.

  39. Note that if \(\beta (z)=1,\) we obtain that \(j(z)=\lambda ^{-1}(z-\ln z-1)\) as in the baseline model. Also, note that, for simplicity, we are reporting the case in which the support of intermediates remains [0, 1]. Online Appendix H.2 discusses the case when \(\beta (z)\) takes the form of (21), in which the support changes with computerization.

  40. Note that \(j(z,\chi )\) increases monotonically with an increase in \({\mathcal {I}} (z,\chi )\).

  41. Using that \(\theta (j)\) follows an exponential decay (equation (17)), the threshold \( {\overline{j}}\) is \({\overline{j}}=\frac{\ln \left( \frac{\lambda _{1}}{\lambda _{2}}\right) }{\lambda _{1}-\lambda _{2}}\).

  42. See online appendix H.3 for a derivation and further characterization of the change in the world output distribution.

  43. Online Appendix E derives these aggregate production functions. We show that \(f(\theta _{j})\equiv 2\theta _{j}\), \(\alpha =1/2\), \(g(\theta _{j})\equiv \theta _{j}\left[ \exp \int _{z\in {\mathcal {Z}}_{j}} \ln \left[ \frac{\Delta _{j}^{1-z}}{\left( Z_{j}-\Delta _{j}\right) ^{z}}\right] dz \right] \) and \(\alpha _j\equiv Z_{j}-\Delta _j\). We also show that \(\alpha _j\) is increasing with the productivity of the country, which directly follows from Proposition 1.

  44. This accounting exercise with heterogeneous capital shares has recently been done by Feenstra et al. (2015), who document sizable differences from traditional TFP accounting. For example, the fraction of the variance of GDP explained by the variance of inputs (in 2005) sharply increases when allowing for heterogeneous shares from 0.25 to 0.34.

  45. We obtain similar results if we use long differences of 5-years or if we use the initial TFP level in 1990 as a regressor instead of \(\log (TFP)_{ct}\).

  46. The proof is immediate and follows from the fact that our competitive equilibrium replicates the social planner allocation. If the social planner is allowed a thinner geographical allocation of inputs, aggregate output cannot decrease.

  47. From the steady-state level condition of world capital and \({\dot{b}}_{j}(t)=0\), we can find the amount of assets hold by agents in country j. However, there exists an indeterminacy in the composition of the asset portfolio of agents. This is inconsequential for our purposes given that the interest rate is equalized across countries and, thus, the composition of the asset portfolio does not affect income nor consumption.


  • Acemoglu, D., Gancia, G., & Zilibotti, F. (2015). Offshoring and directed technical change. American Economic Journal: Macroeconomics, 7, 84–122.

    Google Scholar 

  • Acemoglu, D., & Ventura, J. (2002). The world income distribution. The Quarterly Journal of Economics, 117, 659–694.

    Article  Google Scholar 

  • Antràs, P. (2016). Global production: Firms, contracts and trade structure. Princeton: Princeton University Press.

    Book  Google Scholar 

  • Antràs, P., & Caballero, R. J. (2009). Trade and capital flows: A financial frictions perspective. Journal of Political Economy, 117, 701–744.

    Article  Google Scholar 

  • Antràs, P., & Chor, D. (2013). Organizing the global value chain. Econometrica, 81, 2127–2204.

    Article  Google Scholar 

  • Armington, P. S. (1969). A Theory of Demand for Products Distinguished by Place of Production (Une thorie de la demande de produits diffrencis d’aprs leur origine) (Una teora de la demanda de productos distinguindolos segn el lugar de produccin). IMF Staff Papers, 16, 159–178.

    Article  Google Scholar 

  • Bajona, C., & Kehoe, T. (2010). Trade, growth, and convergence in a dynamic Heckscher–Ohlin model. Review of Economic Dynamics, 13, 487–513.

    Article  Google Scholar 

  • Baldwin, R. (2012). Global supply chains: Why they emerged, why they matter, and where they are going. CEPR discussion papers 9103. Discussion papers: C.E.P.R.

  • Baldwin, R. (2016). The great convergence: Information technology and the new globalization. Cambridge: Harvard University Press.

    Google Scholar 

  • Baldwin, R., & Robert-Nicoud, F. (2014). Trade-in-goods and trade-in-tasks: An integrating framework. Journal of International Economics, 92, 51–62.

    Article  Google Scholar 

  • Baldwin, R., & Venables, A. J. (2013). Spiders and snakes: Offshoring and agglomeration in the global economy. Journal of International Economics, 90, 245–254.

    Article  Google Scholar 

  • Basco, S., & Mestieri, M. (2013). Heterogeneous trade costs and wage inequality: A model of two globalizations. Journal of International Economics, 89, 393–406.

    Article  Google Scholar 

  • Baxter, M. (1992). Fiscal policy, specialization, and trade in the two-sector model: The return of Ricardo? Journal of Political Economy, 100, 713–44.

    Article  Google Scholar 

  • Baxter, M., & Kouparitsas, M. A. (2003). Trade structure, industrial structure, and international business cycles. American Economic Review, 93, 51–56.

    Article  Google Scholar 

  • Caliendo, L., & Parro, F. (2012). Estimates of the trade and welfare effects of NAFTA. NBER working papers 18508, National Bureau of Economic Research, Inc.

  • Caselli, F. (2005). Accounting for cross-country income differences. In P. Aghion & S. Durlauf (Eds.), Handbook of economic growth, Chap. 9 (Vol. 1, pp. 679–741). Amsterdam: Elsevier.

    Google Scholar 

  • Caselli, F., & Ventura, J. (2000). A representative consumer theory of distribution. American Economic Review, 90, 909–926.

    Article  Google Scholar 

  • Costinot, A., & Vogel, J. (2010). Matching and inequality in the world economy. Journal of Political Economy, 118, 747–786.

    Article  Google Scholar 

  • Costinot, A., Vogel, J., & Wang, S. (2013). An elementary theory of global supply chains. Review of Economic Studies, 80, 109–144.

    Article  Google Scholar 

  • Cunat, A., & Maffezzoli, M. (2004). Neoclassical growth and commodity trade. Review of Economic Dynamics, 7, 707–736.

    Article  Google Scholar 

  • Deardorff, A. V. (1998). Fragmentation across cones. Papers, Michigan - Center for Research on Economic & Social Theory 98-14, Michigan - Center for Research on Economic & Social Theory.

  • Deardorff, A. V. (2001a). Fragmentation in simple trade models. The North American Journal of Economics and Finance, 12, 121–137.

    Article  Google Scholar 

  • Deardorff, A. V. (2001b). Rich and poor countries in neoclassical trade and growth. Economic Journal, 111, 277–94.

    Article  Google Scholar 

  • Deardorff, A. V. (2013). Growth or decline of comparative advantage. Journal of Macroeconomics, 38, 12–18.

    Article  Google Scholar 

  • Dixit, A. K., & Grossman, G. M. (1982). Trade and protection with multistage production. Review of Economic Studies, 49, 583–594.

    Article  Google Scholar 

  • Dornbusch, R., Fischer, S., & Samuelson, P. (1980). Heckscher–Ohlin trade theory with a continuum of goods. The Quarterly Journal of Economics, 95, 203–224.

    Article  Google Scholar 

  • Eeckhoudt, L., & Gollier, C. (1995). Demand for risky assets and the monotone probability ratio order. Journal of Risk and Uncertainty, 11, 113–22.

    Article  Google Scholar 

  • Feenstra, R. C., & Hanson, G. (1996). Foreign investment, outsourcing and relative wages. In R. Feenstra, G. Grossman, & D. Irwin (Eds.), The political economy of trade policy: Papers in Honor of Jagdish Bagwati. Cambridge: MIT Press.

    Google Scholar 

  • Feenstra, R. C., Inklaar, R., & Timmer, M. (2015). The next generation of the penn world table. American Economic Review, 105, 3150–3182.

    Article  Google Scholar 

  • Feenstra, R. C., & Jensen, J. B. (2012). Evaluating estimates of materials offshoring from US manufacturing. Economics Letters, 117, 170–173.

    Article  Google Scholar 

  • Grossman, G. M., & Helpman, E. (2014). Growth, trade, and inequality. NBER working papers 20502, National Bureau of Economic Research, Inc.

  • Grossman, G. M., & Rossi-Hansberg, E. (2008). Trading tasks: A simple theory of offshoring. American Economic Review, 98, 1978–97.

    Article  Google Scholar 

  • Hall, R. E., & Jones, C. I. (1999). Why do some countries produce so much more output per worker than others? The Quarterly Journal of Economics, 114, 83–116.

    Article  Google Scholar 

  • Hanson, G. (2012). The rise of middle kingdoms: Emerging economies in global trade. Journal of Economic Perspectives, 26, 41–64.

    Article  Google Scholar 

  • Hopkins, E., & Kornienko, T. (2004). Ratio orderings and comparative statics. ESE discussion papers 91. Edinburgh School of Economics: University of Edinburgh.

  • Hummels, D., Ishii, J., & Yi, K.-M. (2001). The nature and growth of vertical specialization in world trade. Journal of International Economics, 54, 75–96.

    Article  Google Scholar 

  • Jin, K. (2012). Industrial structure and capital flows. American Economic Review, 102, 2111–46.

    Article  Google Scholar 

  • Johnson, R. C. (2014). Five facts about value-added exports and implications for macroeconomics and trade research. Journal of Economic Perspectives, 28, 119–42.

    Article  Google Scholar 

  • Jones, C. I. (1995). R&D-based models of economic growth. Journal of Political Economy, 103, 759–84.

    Article  Google Scholar 

  • Jones, R. W. (1967). International capital movements and the theory of tariffs and trade. The Quarterly Journal of Economics, 81, 1–38.

    Article  Google Scholar 

  • Jones, R. W. (2000). Globalization and the theory of input trade. Cambridge: The MIT Press.

    Google Scholar 

  • Kemp, M. C. (1966). The gain from international trade and investment: A Neo-Heckscher–Ohlin approach. The American Economic Review, 56, 788–809.

    Google Scholar 

  • Kohler, W. (2004). International outsourcing and factor prices with multistage production. The Economic Journal, 114, C166–C185.

    Article  Google Scholar 

  • Krugman, P. R., & Venables, A. J. (1995). Globalization and the inequality of nations. The Quarterly Journal of Economics, 110, 857–80.

    Article  Google Scholar 

  • Lane, P. R., & Milesi-Ferretti, G. M. (2007). The external wealth of nations mark II: Revised and extended estimates of foreign assets and liabilities, 1970–2004. Journal of International Economics, 73, 223–250.

    Article  Google Scholar 

  • Li, N., & Jin, K. (2012). International transmission through relative prices. 2012 meeting papers 1185, Society for Economic Dynamics.

  • Matsuyama, K. (2004). Financial market globalization, symmetry-breaking, and endogenous inequality of nations. Econometrica, 72, 853–884.

    Article  Google Scholar 

  • Matsuyama, K. (2005). Credit market imperfections and patterns of international trade and capital flows. Journal of the European Economic Association, 3, 714–723.

    Article  Google Scholar 

  • Matsuyama, K. (2013). Endogenous ranking and equilibrium Lorenz curve Across (ex ante) identical countries. Econometrica, 81, 2009–2031.

    Article  Google Scholar 

  • Reyes-Heroles, R. (2017). The role of trade costs in the surge of trade imbalances. 2017 meeting papers 212, Society for Economic Dynamics.

  • Rodrik, D. (2016). Premature deindustrialization. Journal of Economic Growth, 21, 1–33.

    Article  Google Scholar 

  • Schott, P. K. (2003a). A comparison of Latin American and Asian product exports to the United States, 1972 to 1999. Latin American Journal of Economics, 40, 414–422.

    Google Scholar 

  • Schott, P. K. (2003b). One size fits all? Heckscher–Ohlin specialization in global production. American Economic Review, 93, 686–708.

    Article  Google Scholar 

  • Schott, P. K. (2004). Across-product versus within-product specialization in international trade. The Quarterly Journal of Economics, 119, 646–677.

    Article  Google Scholar 

  • Sposi, M., Santacreu, A. M., & Ravikumar, B. (2017). Capital accumulation and dynamic gains from trade. 2017 meeting papers 915, Society for Economic Dynamics.

  • Stiglitz, J. E. (1970). Factor price equalization in a dynamic economy. Journal of Political Economy, 78, 456–88.

    Article  Google Scholar 

  • Ventura, J. (1997). Growth and interdependence. The Quarterly Journal of Economics, 112, 57–84.

    Article  Google Scholar 

  • Yi, K.-M. (2003). Can vertical specialization explain the growth of world trade? Journal of Political Economy, 111, 52–102.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Martí Mestieri.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

We thank Daron Acemoglu, Pol Antràs and Jaume Ventura for useful comments and suggestions, and Guido Lorenzoni, Kiminori Matsuyama, an associate editor and three anonymous referees for comments on earlier drafts. We have also benefited from comments in our presentations at Barcelona GSE Summer Forum, Bank of Spain, Berkeley, Calgary, CEMFI, CEPR-ESSIM, Chicago Fed Trade Conference, IMF, Nottingham, Stanford, UC3M and Universitat de Barcelona. First draft: July 2014. We thank Shekhar Tomar for his excellent research assistance. Basco acknowledges financial support from Fundación Ramón Areces. Mestieri thanks the ANR for their financial support while at TSE.

Electronic supplementary material


A Data Appendix: Construction of figure 1b

To construct the histogram for final goods we weight the capital share of each 6-digit NAICS by its final-use content. The final-use content is a weighted average of the different end-uses that each 6-digit NAICS has. The weight is computed as a function of the U.S. imports in 1990. Capital shares of each 6-digits NAICS are computed from the NBER CES Manufacturing database definition. The histogram for intermediates is computed in an analogous way. We use the end-use classification of Feenstra and Jensen (2012). Table A.1 in the online appendix reports additional measures of capital intensity dispersion. It also documents the same fact defining final goods as bundles of intermediates.

B Proofs

This section presents the proofs to some of the claims in the paper. The rest can be found in the online appendix.

Derivation of trade balance in steady state The trade balance in the steady-state is \(TB_{j}=r(k_{j}-b_{j})\). From the Euler Equation, \(r=\rho \). Thus, from the demand of capital in country j, \( rk_{j}=\rho k_{j}=\left[ Z_{j}-\Delta _{j}\right] Y\). Lastly, from the evolution of asset holdings, \(b_{j}=\pi _{j}B=2\Delta _{j}B=2\Delta _{j}K=\Delta _{j}\frac{2}{2\rho }Y=\frac{\Delta _{j}}{\rho }Y\). Thus, \( TB_{j}=\left[ Z_{j}-2\Delta _{j}\right] Y\).

Proof of claim on change in relative welfare Note that \(\left[ \frac{c_{j}(0)}{c_{j+1}(0)}\right] ^{with}>\left[ \frac{ c_{j}(0)}{c_{j+1}(0)}\right] ^{without}\) if \(\frac{\Delta _{j}}{\Delta _{j+1} }>\frac{\mu _{j}}{\mu _{j+1}}=\frac{\theta _{j}}{\theta _{j+1}}\). This is true because, from the threshold equilibrium, equation (16), \( \frac{\Delta _{j}}{\Delta _{j+1}}=\left( \frac{\theta _{j}}{\theta _{j+1}} \right) ^{\frac{1}{1-z}}>\frac{\theta _{j}}{\theta _{j+1}}\) given that \(z<1\).

Proof of claim on change in relative GDP To prove this result, we use the equality \(\frac{\Delta _{j}}{\Delta _{j+1}}= \frac{Z_{j}}{Z_{j+1}}\frac{a_{j}}{a_{j+1}}\) where \(a_{j}\equiv 1-\frac{1}{2} (z_{j-1}+z_{j})\). First, notice that \(\frac{a_{j}}{a_{j+1}}<1\) since \(z_{j}\) is decreasing with j. Second, we know from the threshold equilibrium, equation (16), that \(\frac{\Delta _{j}}{\Delta _{j+1}}=\left( \frac{\theta _{j}}{\theta _{j+1}}\right) ^{\frac{1}{1-z}}>1\). These two statements imply that \(\frac{Z_{j}}{Z_{j+1}}>\frac{\Delta _{j}}{\Delta _{j+1} }\). This last inequality means that production inequality increases because \( \frac{Z_{j}}{Z_{j+1}}>\frac{\Delta _{j}}{\Delta _{j+1}}=\left( \frac{\theta _{j}}{\theta _{j+1}}\right) ^{\frac{1}{1-z}}>\frac{\theta _{j}}{\theta _{j+1} },\) where we have used the threshold equilibrium and that \(0<z<1\).

Derivation of production shares as a function of the assignment of intermediates To derive this expression note that \(s^{y,\text {without}}(z)=\lambda e^{-\lambda \left( -\frac{1+\ln \left( ze^{-z}\right) }{\lambda }\right) }=\lambda ze^{1-z}\). To express the output share with unbundling, note that \( s^{y,\text {with}}(j)=-\frac{dz}{dj}\Longleftrightarrow s^{y,\text {with}}(z)=- \frac{1}{\frac{dj}{dz}}\). Using that \(\frac{dj}{dz}=-\frac{1-z}{\lambda z},\) we have that \(s^{y,\text {with}}(z)=\frac{\lambda z}{1-z}\). The change in output share in terms of j is \(\Delta s_{j}^{y}=\frac{\lambda W(-\exp (-1-\lambda j))}{1+W(-\exp (-1-\lambda j))}-\lambda \exp (-\lambda j)\).

Proof of claim that the change in world output share is negative for \(z\in (0,{{\bar{z}}})\)and positive for\(z\in (\bar{z},1]\). Note that \(\Delta s^{y}(z)\) is continuous, increasing for z\(\in (1-3W(1/3),1]\) and decreasing otherwise. Moreover, \(\Delta s^{y}(0)=0\), \( \frac{d\Delta s^{y}}{dz}(0)<0\), \(\frac{d\Delta s^{y}}{dz}(1)=\infty \) and the result follows. We note in passing that \(\Delta s^{y}(z)\) is convex for all z.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Basco, S., Mestieri, M. The world income distribution: the effects of international unbundling of production. J Econ Growth 24, 189–221 (2019).

Download citation

  • Published:

  • Issue Date:

  • DOI:


JEL Classification