自动化技术、结构转型与中国收入分配格局的演化

Automation Technology,Structural Change,and the Evolution of Income Distribution in China

本文建立了一个两部门的内生自动化技术选择模型,定量考察了内生的自动化技术与产业结构转型对我国初次收入分配格局的影响。本文的研究有以下发现:(1)除了产业结构转型外,内生的自动化技术是影响中国劳动收入份额和服务业部门相对制造业部门工资溢价的另一重要机制。(2)中国近二十年来劳动收入份额呈现U型演化趋势,自动化技术和产业结构转型分别主导了劳动收入份额下降和上升阶段。部门工资溢价的持续上升则是自动化技术和产业结构转型共同作用的结果。(3)要素禀赋与劳动力技能禀赋的结构性变化是驱动自动化技术与产业结构转型、进而影响劳动收入份额和部门工资溢价的重要因素。(4)加强对制造型劳动力人力资本投资,缩小其与服务型劳动力之间的技能差距,同时鼓励制造型劳动力向服务型劳动力转移,有助于提高劳动收入份额并降低部门工资溢价。本文的研究为理解中国初次收入分配演化的驱动因素与内在机制提供了新的理论视角,可为进一步改善收入分配格局提供启示。

In the last two decades,the wage premium between China's service and manufacturing sectors has increased;however,the labor share pattern appears to be U-shaped.Concurrently,the processes of automation and tertiarization have accelerated in China.Following the incorporation of more industrial robots into factories and as the Chinese economy transitions from the manufacturing sector to the service sector,this paper explores the driving factors behind these trends and how they will affect factor income distribution.We extend the task-based framework to the context of the manufacturing and service sectors,and then quantitatively investigate the implications of automation and tertiarization for primary income distribution.The core of our model is that the degree of automation is endogenous,firms adopt automation technologies to maximize their profits,and tasks are allocated to capital or labor following these factors'comparative advantage.Moreover,the adoption of automation and capital density are heterogeneous across sectors.We extend the task-based framework to incorporate the manufacturing and service sectors,with a view to examining the quantitative impact of automation and tertiarization on primary income distribution.At the heart of our model is the notion that the extent of automation is endogenous.That is,firms adopt automation technologies to maximize their profits.In addition,the comparative advantages of various factors determine the allocation of tasks to either capital or labor.Finally,the adoption of automation and capital density is heterogeneous across sectors.We use the lens of this model to derive three main findings.First,we observe that in addition to structural transformation,the adoption of endogenous automation technologies serves as a significant mechanism for the evolution of factor income distribution in China.Second,we find that both automation technologies and structural change jointly explain the U-shaped labor share pattern and the increasing sectoral wage premium.Automation technologies dominate the decrease phase in the U-shaped labor share pattern,while structural change dominates the increase phase on the other side.Both automation technologies and structural change play a crucial role in explaining the continuous growth of wage premiums.Finally,we demonstrate that factor and skill endowments are the driving factors behind automation technologies and structural change,which have significant implications for the factor income distribution pattern.In particular,deepening capital reduces the share of labor income and expands the wage premium.An increase in service labor supply increases the labor share and decreases the wage premium.However,manufacturing labor shortages reduce both the labor share and the wage premium.This paper makes three contributions to the literature.First,we build a tractable but general model to investigate automation and structural changes in China,in addition to their implications for factor income distribution.This model can be used to discuss the patterns of both labor share and wage premium using a uniform framework.Hence,we supplement theory on the evolution of China's factor income distribution.Moreover,we complement studies that identify the influential factors in income distribution based on microdata and empirical methods by providing a macro perspective.As mentioned above,our model effectively explains the changes in the aggregate labor income share and the sector wage premium.Second,our model matches the changes in factor endowment,industry structure,employment structure,wage premium,and so on over the last two decades,especially the nonmonotonic pattern of aggregate labor share,which bolsters our confidence that multiple factors must affect the factor income distribution.Our second contribution is to shed light on the driving factors and the intermediate channels behind the pattern of the factor income distribution.Finally,we simulate how factor income distribution varies with factor and skill endowments.We decompose the total effects into the automation and structural change effects.In this respect,our work has both theoretical and practical significance and illuminates future patterns in factor income distribution.In the future,increasing the human capital investment in manufacturing labor,gradually narrowing the skill gap between manufacturing labor and service labor,and encouraging the transfer of manufacturing labor to service labor are possible ways to improve the income distribution pattern.