Working Papers
Structural Change with Endogenous Input-Output Linkages (Job Market Paper)
This paper studies the externalities of input-output linkages in structural change. In contrast to the consumer side mechanism in the literature, I find a novel producer side mechanism through intersector outsourcing of intermediate inputs (I.I). Qualitatively, as relative outsourcing cost declines over time, the sector with larger growth of I.I demand relies on outsourcing hence shedding out labor and capital. The sector with larger growth of I.I supply absorbs labor and capital to produce external outsourcing. As a result, structural change arises from the sector with larger growth of I.I demand to the sector with larger growth of I.I supply. To quantify the I.I supply and I.I demand channels, I build a multi-sector general equilibrium model, featuring endogenous input-output linkages, nonhomothetic CES preferences and technology, and heterogeneous firms. The model is structurally calibrated with panel data of 35 major economies during 1995-2007. I find that (i) I.I supply effects are at least comparable to price and income effects in the literature; (ii) I.I demand effects are less critical since the I.I demand intensity is stable over time; (iii) about half of the structural change attributes to rising comparative advantage of supplying I.I by services relative to by manufacturing.
paper slides
Input-Output Linkages and Productivity Measurement
This paper studies the role of input-output (IO) linkages in sectoral and aggregate productivity measurement for 39 countries during 1995-2007. The main finding is that without taking IO linkage into account, development accounting (DA) method biasedly measure the sectoral TFP and sectoral labor productivity for all sample countries. For aggregate TFP, the quantitative analysis result suggests that compared with IO based measurement, DA method overestimates annual aggregate TFP growth rate by about 15 percent for the US and underestimates that by about 9 percent for China. For the 39 sample countries on average, DA underestimates the annual aggregate TFP growth rate by about a half percent. Moreover, the counterfactual study implies that previous studies are either insufficient to taking IO linkage into account, or totally ignore IO linkage which result in biased productivity measurement and implication.
paper_1.pdf
Work in Progress
Structural Change in Global Value Chains
In this paper, I extend my job market paper to a global context. Given global sourcing of intermediate input (I.I), I examine to what extent the I.I supply mechanism accounts for domestic structural change. I study this question both empirically and quantitatively. For the empirical study, the regression results suggest that the I.I supply mechanism holds. For the same and recent World Input-Output Database 2016, the positive association between structural change and I.I supply mechanism holds in all of the 29 two-digit sectors. It also holds in 40 out of 44 major economies. To quantify the I.I supply mechanism, I develop a multi-stage, multi-country and multi-sector general equilibrium model. This model is developed by integration of two strands of literature: multi-sector and one-stage Eaton-Kortum trade (Lorenzo and Parro 2015); one-sector and multi-stage Eaton-Kortum trade (Antras and de Gortari 2017). The model features horizontal integration through sourcing finished I.I in a given stage; also features vertical integration through sourcing unfinished I.I from precedent stages.
paper_2.pdf
Secular Structural Change, Marketization, and Rising Long-Run US Skill Premium
This paper studies the rising skill premium in the US since 1950. During this period, both the relative supply and demand of high-skill to low-skill workers arise dramatically. I seek to understand the rise of quantity and price premium through two mechanisms: structural change and service marketization. I argue that since 1950, the US economy experiences strong structural change from agriculture and manufacturing to services. On the one hand, rising relative price and growing income make consumer willing to consume more services than goods. On the other hand, manufacturing and agriculture companies increase domestic outsourcing of non-core service activities to specialized service contractors. Both the consumer side channel and the producer side channel favor services, rather than goods. Moreover, due to technological development and declining outsourcing cost, the opportunity cost of producing at home shifts up drastically. Henceforth consumer has incentive to outsource more and more home production to specialized service companies in the market. As a result, within services, market service arises relative to home production. Since market service is the most intensive use sector of high-skill workers, these two mechanisms push up relative demand of high-skill to low-skill workers. This raises wage premium. Rising wage premium gives the incentive to supply relatively more high-skill to low-skill workers in the labor market. I construct a multi-sector general equilibrium model with interaction of skill heterogeneity, home production, endogenous consumer service choice and upward-sloping skill supply. This unified theory can explain these salient long run features in the US.
paper_3.pdf
Structural Change with Endogenous Input-Output Linkages (Job Market Paper)
This paper studies the externalities of input-output linkages in structural change. In contrast to the consumer side mechanism in the literature, I find a novel producer side mechanism through intersector outsourcing of intermediate inputs (I.I). Qualitatively, as relative outsourcing cost declines over time, the sector with larger growth of I.I demand relies on outsourcing hence shedding out labor and capital. The sector with larger growth of I.I supply absorbs labor and capital to produce external outsourcing. As a result, structural change arises from the sector with larger growth of I.I demand to the sector with larger growth of I.I supply. To quantify the I.I supply and I.I demand channels, I build a multi-sector general equilibrium model, featuring endogenous input-output linkages, nonhomothetic CES preferences and technology, and heterogeneous firms. The model is structurally calibrated with panel data of 35 major economies during 1995-2007. I find that (i) I.I supply effects are at least comparable to price and income effects in the literature; (ii) I.I demand effects are less critical since the I.I demand intensity is stable over time; (iii) about half of the structural change attributes to rising comparative advantage of supplying I.I by services relative to by manufacturing.
paper slides
Input-Output Linkages and Productivity Measurement
This paper studies the role of input-output (IO) linkages in sectoral and aggregate productivity measurement for 39 countries during 1995-2007. The main finding is that without taking IO linkage into account, development accounting (DA) method biasedly measure the sectoral TFP and sectoral labor productivity for all sample countries. For aggregate TFP, the quantitative analysis result suggests that compared with IO based measurement, DA method overestimates annual aggregate TFP growth rate by about 15 percent for the US and underestimates that by about 9 percent for China. For the 39 sample countries on average, DA underestimates the annual aggregate TFP growth rate by about a half percent. Moreover, the counterfactual study implies that previous studies are either insufficient to taking IO linkage into account, or totally ignore IO linkage which result in biased productivity measurement and implication.
paper_1.pdf
Work in Progress
Structural Change in Global Value Chains
In this paper, I extend my job market paper to a global context. Given global sourcing of intermediate input (I.I), I examine to what extent the I.I supply mechanism accounts for domestic structural change. I study this question both empirically and quantitatively. For the empirical study, the regression results suggest that the I.I supply mechanism holds. For the same and recent World Input-Output Database 2016, the positive association between structural change and I.I supply mechanism holds in all of the 29 two-digit sectors. It also holds in 40 out of 44 major economies. To quantify the I.I supply mechanism, I develop a multi-stage, multi-country and multi-sector general equilibrium model. This model is developed by integration of two strands of literature: multi-sector and one-stage Eaton-Kortum trade (Lorenzo and Parro 2015); one-sector and multi-stage Eaton-Kortum trade (Antras and de Gortari 2017). The model features horizontal integration through sourcing finished I.I in a given stage; also features vertical integration through sourcing unfinished I.I from precedent stages.
paper_2.pdf
Secular Structural Change, Marketization, and Rising Long-Run US Skill Premium
This paper studies the rising skill premium in the US since 1950. During this period, both the relative supply and demand of high-skill to low-skill workers arise dramatically. I seek to understand the rise of quantity and price premium through two mechanisms: structural change and service marketization. I argue that since 1950, the US economy experiences strong structural change from agriculture and manufacturing to services. On the one hand, rising relative price and growing income make consumer willing to consume more services than goods. On the other hand, manufacturing and agriculture companies increase domestic outsourcing of non-core service activities to specialized service contractors. Both the consumer side channel and the producer side channel favor services, rather than goods. Moreover, due to technological development and declining outsourcing cost, the opportunity cost of producing at home shifts up drastically. Henceforth consumer has incentive to outsource more and more home production to specialized service companies in the market. As a result, within services, market service arises relative to home production. Since market service is the most intensive use sector of high-skill workers, these two mechanisms push up relative demand of high-skill to low-skill workers. This raises wage premium. Rising wage premium gives the incentive to supply relatively more high-skill to low-skill workers in the labor market. I construct a multi-sector general equilibrium model with interaction of skill heterogeneity, home production, endogenous consumer service choice and upward-sloping skill supply. This unified theory can explain these salient long run features in the US.
paper_3.pdf