人口结构转型、人口红利演进与出口增长——来自中国城市层面的经验证据
|
摘要
本文基于2000年、2005年和2010年三次全国人口普查数据,构建城市层面包含人口流动信息的人口结构指标,从微观层面研究人口结构变动对出口的影响及其作用机制。研究发现,以劳动人口比或抚养比为表征的城市人口结构变动会显著影响出口,城市劳动人口比的提升或城市抚养比的下降会显著促进出口。对影响机制的实证检验表明,城市人口结构变动对出口的影响会伴随用工成本的上升而逐渐弱化,即存在人口结构变动影响出口的"成本效应"。城市人口结构变动对出口的影响会伴随人力资本水平的提升而逐渐弱化,即存在人口结构变动影响出口的"人力资本效应"。城市人口结构变动对出口的影响主要源于对劳动力投入的依赖性。本文的研究结论意味着,促进人力资本提升,加快实现"人口红利"向"人才红利"的转变,是应对人口老龄化背景下中国出口转型升级的重要思路。
Since China's reform and opening up, demographics have played an important role in the development of China's overseas trade. The huge population provides an abundant and cheap labor force for Chinese firms and also lays a solid foundation for the successful implementation of the export-oriented economic development strategy. Since the start of the 21~(st) century, however, China's demographic structure has been undergoing profound changes. On the one hand, the demographic dividend is gradually disappearing, and the problems of population aging and labor shortage have become more prominent. On the other hand, with the gradual easing of the household registration system(Hukou system), the domestic population flow has become more frequent. Since the 1980 s, more than 340 million people in China have migrated; that is the largest population migration in human history. Amid a shifting demographic structure and the increasing frequency of population migration, it is important to study the impact of structural demographic changes on exports and its mechanism from a micro perspective. Doing so can not only highlight the necessity of focusing on the "talent dividend" rather than the "demographic dividend", but also help clarify how to transform and upgrade China's overseas trade. Using data from the three national censuses in 2000, 2005, and 2010 and highly detailed firm export information provided by the Customs Database, this paper constructs a demographic structure indicator that includes population migration information at the city level and studies the effect of demographic structure changes on firm exports. The results indicate that an increase in the labor-population ratio or a decline in the dependency ratio will significantly promote exports, a result that holds under a multi-dimensional robustness test. The impact of city demographic structure on exports will gradually weaken with the increase of labor costs and human capital level; that is, there are cost-and human capital-related effects of demographic structure changes on exports. Further analysis shows that the impact of city demographic structures on exports is mainly due to the dependence on labor input.To solve the potential endogeneity problem in the empirical estimation, this paper first adopts a more exogenous explanatory variable, namely the demographic structure. In addition, by using highly detailed panel data(firm-product-destination-trade mode-year), potential reverse causality problems are avoided. Third, this paper constructs high-dimensional panel data for firm-product-destination-trade mode-year, which greatly reduces the biases caused by omitted variables through strict fixed effect control. Fourth, multiple demographic indicators, such as the city-level dependency ratio, are constructed to deal with the measurement error problem. Fifth, the data dimensions are scaled and the benchmark results are verified repeatedly across multiple dimensions. Sixth, using historical variables as instrumental variables for regression, the credibility of the benchmark results is further verified.This paper contributes to the literature in several ways. First, it builds a demographic structure index based on population migration information at the city level, which enables micro-level analysis of the relationship between demographic structures and exports. Second, this paper identifies cost and human capital mechanisms in the effect of demographic structures on exports from a micro perspective. Compared to the existing theoretical analysis framework, which is based on the saving rate mechanism, these mechanisms are important supplement. Third, the conclusions of this paper not only provide empirical evidence for the necessity of accelerating from a "demographic dividend" focus to a "talent dividend" focus in China, but also help to explain the current "talent war" between cities. It provides empirical evidence for promoting the reform of the current household registration system and realizing the rational flow of labor.
Since China's reform and opening up, demographics have played an important role in the development of China's overseas trade. The huge population provides an abundant and cheap labor force for Chinese firms and also lays a solid foundation for the successful implementation of the export-oriented economic development strategy. Since the start of the 21~(st) century, however, China's demographic structure has been undergoing profound changes. On the one hand, the demographic dividend is gradually disappearing, and the problems of population aging and labor shortage have become more prominent. On the other hand, with the gradual easing of the household registration system(Hukou system), the domestic population flow has become more frequent. Since the 1980 s, more than 340 million people in China have migrated; that is the largest population migration in human history. Amid a shifting demographic structure and the increasing frequency of population migration, it is important to study the impact of structural demographic changes on exports and its mechanism from a micro perspective. Doing so can not only highlight the necessity of focusing on the "talent dividend" rather than the "demographic dividend", but also help clarify how to transform and upgrade China's overseas trade. Using data from the three national censuses in 2000, 2005, and 2010 and highly detailed firm export information provided by the Customs Database, this paper constructs a demographic structure indicator that includes population migration information at the city level and studies the effect of demographic structure changes on firm exports. The results indicate that an increase in the labor-population ratio or a decline in the dependency ratio will significantly promote exports, a result that holds under a multi-dimensional robustness test. The impact of city demographic structure on exports will gradually weaken with the increase of labor costs and human capital level; that is, there are cost-and human capital-related effects of demographic structure changes on exports. Further analysis shows that the impact of city demographic structures on exports is mainly due to the dependence on labor input.To solve the potential endogeneity problem in the empirical estimation, this paper first adopts a more exogenous explanatory variable, namely the demographic structure. In addition, by using highly detailed panel data(firm-product-destination-trade mode-year), potential reverse causality problems are avoided. Third, this paper constructs high-dimensional panel data for firm-product-destination-trade mode-year, which greatly reduces the biases caused by omitted variables through strict fixed effect control. Fourth, multiple demographic indicators, such as the city-level dependency ratio, are constructed to deal with the measurement error problem. Fifth, the data dimensions are scaled and the benchmark results are verified repeatedly across multiple dimensions. Sixth, using historical variables as instrumental variables for regression, the credibility of the benchmark results is further verified.This paper contributes to the literature in several ways. First, it builds a demographic structure index based on population migration information at the city level, which enables micro-level analysis of the relationship between demographic structures and exports. Second, this paper identifies cost and human capital mechanisms in the effect of demographic structures on exports from a micro perspective. Compared to the existing theoretical analysis framework, which is based on the saving rate mechanism, these mechanisms are important supplement. Third, the conclusions of this paper not only provide empirical evidence for the necessity of accelerating from a "demographic dividend" focus to a "talent dividend" focus in China, but also help to explain the current "talent war" between cities. It provides empirical evidence for promoting the reform of the current household registration system and realizing the rational flow of labor.
引文
蔡昉,2009:《未来的人口红利——中国经济增长源泉的开拓》,《中国人口科学》第1期。
蔡昉,2010:《人口转变、人口红利与刘易斯转折点》,《经济研究》第4期。
陈勇兵、李燕、周世民,2012:《中国企业出口持续时间及其决定因素》,《经济研究》第7期。
陈钊、熊瑞祥,2015:《比较优势与产业政策效果——来自出口加工区准实验的证据》,《管理世界》第8期。
戴翔、刘梦,2018:《人才何以成为红利——源于价值链攀升的证据》,《中国工业经济》第4期。
都阳、曲玥,2009:《劳动报酬、劳动生产率与劳动力成本优势——对2000—2007年中国制造业企业的经验研究》,《中国工业经济》第5期。
郭雄、夏燕月、李效莲、李俊臣,2005:《抗日战争时期国民党正面战场》,四川人民出版社。
胡翠、许召元,2014:《人口老龄化对储蓄率影响的实证研究——来自中国家庭的数据》,《经济学(季刊)》第4期。
李兵、任远,2015:《人口结构是怎样影响经常账户不平衡的?——以第二次世界大战为工具变量的经验证据》,《经济研究》第10期。
陆旸、蔡昉,2014:《人口结构变化对潜在增长率的影响:中国和日本的比较》,《世界经济》第1期。
孙楚仁、田国强、章韬,2013:《最低工资标准与中国企业的出口行为》,《经济研究》第2期。
田巍、姚洋、余淼杰、周羿,2013:《人口结构与国际贸易》,《经济研究》第11期。
铁瑛、刘啟仁,2018:《人民币汇率变动与劳动力技能偏向效应——来自中国微观企业的证据》,《金融研究》第1期。
王德文、蔡昉、高文书,2005:《全球化与中国国内劳动力流动:新趋势与政策含义》,《开放导报》第4期。
谢建国、赵锦春、林小娟,2015:《不对称劳动参与、收入不平等与全球贸易失衡》,《世界经济》第9期。
杨继军、马野青,2011:《中国的高储蓄率与外贸失衡:基于人口因素的视角》,《国际贸易问题》第12期。
姚洋、余淼杰,2009:《劳动力、人口和中国出口导向的增长模式》,《金融研究》第9期。
张同斌,2016:《从数量型“人口红利”到质量型“人力资本红利”——兼论中国经济增长的动力转换机制》,《经济科学》第5期。
朱超、张林杰,2012:《人口结构能解释经常账户平衡吗》,《金融研究》第5期。
Bahar,D.,and H.Rapoport,2018,“Migration,Knowledge Diffusion and the Comparative Advantage of Nations”,Economic Journal,128(612),F273—F305.
Blatter,M.,S.Muehlemann,and S.Schenker,2012,“The Costs of Hiring Skilled Workers”,European Economic Review,56(1),20—35.
Bloom,D.E.,and J.G.Williamson,1998,“Demographic Transitions and Economic Miracles in Emerging Asia”,World Bank Economic Review,12(3),419—455.
Brecher,R.A.,1974,“Minimum Wage Rates and the Pure Theory of International Trade”,Quarterly Journal of Economics,88(1),98—116.
Cai,H.,and Q.Liu,2009,“Competition and Corporate Tax Avoidance:Evidence from Chinese Industrial Firms”,Economic Journal,119(537),764—795.
Cai,J.,and A.Stoyanov,2016,“Population Aging and Comparative Advantage”,Journal of International Economics,102,1—21.
Chan,K.W.,2013,“China:Internal Migration”,The Encyclopedia of Global Human Migration,Blackwell Publishing.
Ding,H.,H.Fan,and S.Lin,2018,“Connect to Trade”,Journal of International Economics,110,50—62.
Egger,H.,P.Egger,and J.Markusen,2012,“International Welfare and Employment Linkages Arising from Minimum Wages”,International Economic Review,53(3),771—789.
Kaiser,B.,and M.Siegenthaler,2016,“The Skill-biased Effects of Exchange Rate Fluctuations”,Economic Journal,126(592),756—780.
Krueger,A.,and L.Summers,1988,“Efficiency Wages and the Inter-industry Wage Structure”,Econometrica,56(2),259—293.
Leland,H.E.,1968,“Saving and Uncertainty:The Precautionary Demand for Saving”,Quarterly Journal of Economics,82(3),465—473.
Lewis,W.,1954,“Economic Development with Unlimited Supplies of Labour”,Manchester School,22(2),139—191.
Ma,L.,and Y.Tang,2018,“Geography,Trade,and Internal Migration in China”,Working Paper.
Manova,K.,S.Wei,and Z.Zhang,2015,“Firm Exports and Multinational Activity under Credit Constraints”,Review of Economics and Statistics,97(3),574—588.
Naito,T.,and L.Zhao,2009,“Aging,Transitional Dynamics,and Gains from Trade”,Journal of Economic Dynamics & Control,33(8),1531—1542.
Shapiro,J.M.,2006,“Smart Cities:Quality of Life,Productivity,and the Growth Effects of Human Capital”,Review of Economics and Statistics,88(2),324—335.
Shapiro,C.,and J.Stiglitz,1984,“Equilibrium Unemployment as a Worker Discipline Device”,American Economic Review,74(3),433—444.
Silvanto,S.,J.Ryan,and Y.Mcnulty,2015,“An Empirical Study of Nation Branding for Attracting Internationally Mobile Skilled Professionals”,Career Development International,20(3),238—258.
Verhoogen,E.,2008,“Trade,Quality Upgrading and Wage Inequality in the Mexican Manufacture Sector”,Quarterly Journal of Economics,123(2),489—530.
Yakita,A.,2012,“Different Demographic Changes and Patterns of Trade in a Heckscher-Ohlin Setting”,Journal of Population Economics,25(3),853—870.
① 国内已有文献对于“人才红利”有着不同的提法,比如蔡昉(2009)提出的“第二次人口红利”,张同斌(2016)提出的“质量型‘人力资本红利’”,戴翔和刘梦(2018)所使用的“人才红利”,这些概念其实质是相同的。
(1)人口结构的变动周期往往被认定为60年以上,很难在短期内发生大的变动。
(2)这一处理并不能完全规避2000—2003年出口的反向作用,其外生性依然受到挑战。虽然可以基于2005年和2010年的面板进行回归,但对面板的个体层面进行固定效应控制存在困难,回归结果面临遗漏变量问题的挑战。
(3)一般而言,男性和女性在退休年龄上有差异,女性往往会比男性早5年退出劳动市场。囿于数据,我们并不能对性别进行分类,因此,本文将15—65岁的人口视为劳动力人口。
(4)本文进一步做了分年份的细致对照表,限于篇幅,备索。
(5)本文同时绘制了人口结构指标在各省的分布图,可以清晰地看出人口结构在城市间的差异性,限于篇幅,备索。
(6)鉴于本文的实证研究直接针对出口企业样本,将企业维度加总至城市维度有助于涵盖企业在出口市场的进入退出行为,从而有助于缓解样本选择所带来的问题。
(7)本文在控制“企业-HS六位码产品-目的国-贸易方式”固定效应的情形下进行回归,样本损失大约90%,估计结果与基准结果保持一致。本文也做了对所有解释变量进行一期滞后的估计,与基准结果保持一致。本文还做了固定效应调整后的稳健性考察,在不断将联合固定效应逐步放松至一维固定效应的过程中,回归结果均保持稳健。限于篇幅,上述结果不再列出,备索。
(8)在历时八年的抗日战争中,正面战场上所爆发的主要战役、持续时间与主要波及城市如下:淞沪会战(1937.8.13—1937.11.12),上海;太原会战(1937.10.13—1937.11.2),太原、忻州、石家庄、阳泉、晋中;南京保卫战(1937.11.12—1937.12.13),苏州、嘉兴、无锡、湖州、宣城、镇江、南京;徐州会战(1938.1—1938.5),蚌埠、枣庄、临沂、菏泽、徐州;武汉会战(1938.3—1938.10),武汉;南昌会战(1939.3—1939.5)、随枣会战(1939.5)、枣宜会战(1940.5—1940.6),南昌、随州、襄阳、宜昌;第一次长沙会战(1939.9.14—1939.10.10)、第二次长沙会战(1941.9—1941.10)、第三次长沙会战(1941.12—1942.1),长沙;豫湘桂战役(1944.4—1944.12),洛阳、郑州、许昌、长沙、衡阳、桂林、柳州等。
(9)相当于一个渐进式的DID设定,控制组为没有受到战争影响的城市,处理组为受到战争影响的城市。
(10)虽然中国在全面抗日战争胜利后又爆发了解放战争,但两次战争的本质是不同的,在保家卫国的抗日战争中,中国军人作战顽强,寸土不让;而在解放战争中,国民党人心尽丧,尤其在解放战争后期,大量国民党军人临阵起义,向解放军投诚。再者,两次战争对城市的影响程度是不同的,抗日战争中日本侵略者的目标是各大城市,城市经常成为大型战役的主战场;但中国共产党一直秉承“农村包围城市”战略,国民党军队在解放战争早期的进攻也主要集中在以农村、山地为主的解放区,中国共产党在解放战争中后期开始解放城市时,不仅通过谈判的手段尽可能争取和平解放,而且也非常注重对大城市基础设施的保护(例如,解放上海的战斗),人心涣散的国民党军队整体上并未进行强有力的反击。
(11)主要包括如下四个方面:(1)仅保留了抗日战争中几乎沦陷的东部沿海省份;(2)剔除了战争中表现活跃的“川军”和“滇军”的来源省份四川省和云南省;(3)仅保留了日本军队主要影响地区,即距离海岸线球面距离1000千米之内的城市;(4)在东部沿海省份样本的基础上补充了中部地区省份。结果均与表3保持高度一致,限于篇幅,具体结果不再列出,备索。
(12)导致变量不显著的原因可能很多,这一结果只能为工具变量的合理性提供一定佐证。
(13)我们使用了三种方案:第一种方案按照洪水实际影响,选定城市包括郑州、开封、许昌、周口、亳州、阜阳、淮南、蚌埠、徐州、淮阴、济南、商丘、淮安、宿州、扬州和镇江在2005年进入对照组;第二种方案涵盖了后续8年发生的黄河改道影响,增补城市济宁、菏泽、临沂和枣庄在2005年进入对照组,同时增补信阳、南阳、驻马店、淮安、盐城和日照在2010年进入对照组;第三种方案考虑到自然灾害可能造成人口流动的扩散,使用河南省、山东省、江苏省和安徽省的所有城市在2005年进入对照组。
(14)之所以没有对(1)式进行估计,是因为对过于大的样本量进行估计费时太久,而表2的基准结果已经说明(1)式和(2)式的估计在定性方面没有差异。
(15)这种方法本质上是一种子样本分析方法,每个分位点的估计都是前一个分位点的子样本估计,因此将参数估计值与前一个分位点的参数估计值进行对比就可以近似地反映该1%子样本中人口结构对出口的影响。具体来说,第x次回归对应Quantile≥x%的样本,从而第(x+1)次回归相当于第x次回归的子样本估计,这样一来,我们虽然无法直接对Quantile=x%的分样本进行估计,但通过比对第x次和第(x+1)次的回归结果,同样可以捕捉到Quantile=x%的分样本信息,即如果第(x+1)次所得到的参数值估计结果(即分布均值)更大,则有助于说明刚刚剔除的Quantile=x%相对于之后的样本,城市人口结构对出口的正向促进作用相对更小,进而与图中的趋势是一致的。类似的做法也可见于铁瑛和刘啟仁(2018)。
(16)当参数估计值无法在10%显著性水平下显著时,我们将其记为0。
(17)行业人力资本水平以本科以上学历雇员占总从业人员的比例来衡量。
(18)“城市-年份”联合固定可以吸收人口结构的水平项,以及来自城市层面的控制变量。
(19)利用2004年经济普查数据,将固定资产净值和从业人员数分别加总至GBT四位码产品(行业)层面并对应到HS六位码产品维度,而后以固定资产净值除以从业人员,可获得资本劳动比数据。
(20)为了进一步确保行业指标的外生性,我们利用2005年和2010年的数据,将原来的“城市-HS六位码产品-目的国-贸易方式”四维联合固定放松为三维联合固定进行估计,结果仍然保持一致。具体结果不再列出,备索。
蔡昉,2010:《人口转变、人口红利与刘易斯转折点》,《经济研究》第4期。
陈勇兵、李燕、周世民,2012:《中国企业出口持续时间及其决定因素》,《经济研究》第7期。
陈钊、熊瑞祥,2015:《比较优势与产业政策效果——来自出口加工区准实验的证据》,《管理世界》第8期。
戴翔、刘梦,2018:《人才何以成为红利——源于价值链攀升的证据》,《中国工业经济》第4期。
都阳、曲玥,2009:《劳动报酬、劳动生产率与劳动力成本优势——对2000—2007年中国制造业企业的经验研究》,《中国工业经济》第5期。
郭雄、夏燕月、李效莲、李俊臣,2005:《抗日战争时期国民党正面战场》,四川人民出版社。
胡翠、许召元,2014:《人口老龄化对储蓄率影响的实证研究——来自中国家庭的数据》,《经济学(季刊)》第4期。
李兵、任远,2015:《人口结构是怎样影响经常账户不平衡的?——以第二次世界大战为工具变量的经验证据》,《经济研究》第10期。
陆旸、蔡昉,2014:《人口结构变化对潜在增长率的影响:中国和日本的比较》,《世界经济》第1期。
孙楚仁、田国强、章韬,2013:《最低工资标准与中国企业的出口行为》,《经济研究》第2期。
田巍、姚洋、余淼杰、周羿,2013:《人口结构与国际贸易》,《经济研究》第11期。
铁瑛、刘啟仁,2018:《人民币汇率变动与劳动力技能偏向效应——来自中国微观企业的证据》,《金融研究》第1期。
王德文、蔡昉、高文书,2005:《全球化与中国国内劳动力流动:新趋势与政策含义》,《开放导报》第4期。
谢建国、赵锦春、林小娟,2015:《不对称劳动参与、收入不平等与全球贸易失衡》,《世界经济》第9期。
杨继军、马野青,2011:《中国的高储蓄率与外贸失衡:基于人口因素的视角》,《国际贸易问题》第12期。
姚洋、余淼杰,2009:《劳动力、人口和中国出口导向的增长模式》,《金融研究》第9期。
张同斌,2016:《从数量型“人口红利”到质量型“人力资本红利”——兼论中国经济增长的动力转换机制》,《经济科学》第5期。
朱超、张林杰,2012:《人口结构能解释经常账户平衡吗》,《金融研究》第5期。
Bahar,D.,and H.Rapoport,2018,“Migration,Knowledge Diffusion and the Comparative Advantage of Nations”,Economic Journal,128(612),F273—F305.
Blatter,M.,S.Muehlemann,and S.Schenker,2012,“The Costs of Hiring Skilled Workers”,European Economic Review,56(1),20—35.
Bloom,D.E.,and J.G.Williamson,1998,“Demographic Transitions and Economic Miracles in Emerging Asia”,World Bank Economic Review,12(3),419—455.
Brecher,R.A.,1974,“Minimum Wage Rates and the Pure Theory of International Trade”,Quarterly Journal of Economics,88(1),98—116.
Cai,H.,and Q.Liu,2009,“Competition and Corporate Tax Avoidance:Evidence from Chinese Industrial Firms”,Economic Journal,119(537),764—795.
Cai,J.,and A.Stoyanov,2016,“Population Aging and Comparative Advantage”,Journal of International Economics,102,1—21.
Chan,K.W.,2013,“China:Internal Migration”,The Encyclopedia of Global Human Migration,Blackwell Publishing.
Ding,H.,H.Fan,and S.Lin,2018,“Connect to Trade”,Journal of International Economics,110,50—62.
Egger,H.,P.Egger,and J.Markusen,2012,“International Welfare and Employment Linkages Arising from Minimum Wages”,International Economic Review,53(3),771—789.
Kaiser,B.,and M.Siegenthaler,2016,“The Skill-biased Effects of Exchange Rate Fluctuations”,Economic Journal,126(592),756—780.
Krueger,A.,and L.Summers,1988,“Efficiency Wages and the Inter-industry Wage Structure”,Econometrica,56(2),259—293.
Leland,H.E.,1968,“Saving and Uncertainty:The Precautionary Demand for Saving”,Quarterly Journal of Economics,82(3),465—473.
Lewis,W.,1954,“Economic Development with Unlimited Supplies of Labour”,Manchester School,22(2),139—191.
Ma,L.,and Y.Tang,2018,“Geography,Trade,and Internal Migration in China”,Working Paper.
Manova,K.,S.Wei,and Z.Zhang,2015,“Firm Exports and Multinational Activity under Credit Constraints”,Review of Economics and Statistics,97(3),574—588.
Naito,T.,and L.Zhao,2009,“Aging,Transitional Dynamics,and Gains from Trade”,Journal of Economic Dynamics & Control,33(8),1531—1542.
Shapiro,J.M.,2006,“Smart Cities:Quality of Life,Productivity,and the Growth Effects of Human Capital”,Review of Economics and Statistics,88(2),324—335.
Shapiro,C.,and J.Stiglitz,1984,“Equilibrium Unemployment as a Worker Discipline Device”,American Economic Review,74(3),433—444.
Silvanto,S.,J.Ryan,and Y.Mcnulty,2015,“An Empirical Study of Nation Branding for Attracting Internationally Mobile Skilled Professionals”,Career Development International,20(3),238—258.
Verhoogen,E.,2008,“Trade,Quality Upgrading and Wage Inequality in the Mexican Manufacture Sector”,Quarterly Journal of Economics,123(2),489—530.
Yakita,A.,2012,“Different Demographic Changes and Patterns of Trade in a Heckscher-Ohlin Setting”,Journal of Population Economics,25(3),853—870.
① 国内已有文献对于“人才红利”有着不同的提法,比如蔡昉(2009)提出的“第二次人口红利”,张同斌(2016)提出的“质量型‘人力资本红利’”,戴翔和刘梦(2018)所使用的“人才红利”,这些概念其实质是相同的。
(1)人口结构的变动周期往往被认定为60年以上,很难在短期内发生大的变动。
(2)这一处理并不能完全规避2000—2003年出口的反向作用,其外生性依然受到挑战。虽然可以基于2005年和2010年的面板进行回归,但对面板的个体层面进行固定效应控制存在困难,回归结果面临遗漏变量问题的挑战。
(3)一般而言,男性和女性在退休年龄上有差异,女性往往会比男性早5年退出劳动市场。囿于数据,我们并不能对性别进行分类,因此,本文将15—65岁的人口视为劳动力人口。
(4)本文进一步做了分年份的细致对照表,限于篇幅,备索。
(5)本文同时绘制了人口结构指标在各省的分布图,可以清晰地看出人口结构在城市间的差异性,限于篇幅,备索。
(6)鉴于本文的实证研究直接针对出口企业样本,将企业维度加总至城市维度有助于涵盖企业在出口市场的进入退出行为,从而有助于缓解样本选择所带来的问题。
(7)本文在控制“企业-HS六位码产品-目的国-贸易方式”固定效应的情形下进行回归,样本损失大约90%,估计结果与基准结果保持一致。本文也做了对所有解释变量进行一期滞后的估计,与基准结果保持一致。本文还做了固定效应调整后的稳健性考察,在不断将联合固定效应逐步放松至一维固定效应的过程中,回归结果均保持稳健。限于篇幅,上述结果不再列出,备索。
(8)在历时八年的抗日战争中,正面战场上所爆发的主要战役、持续时间与主要波及城市如下:淞沪会战(1937.8.13—1937.11.12),上海;太原会战(1937.10.13—1937.11.2),太原、忻州、石家庄、阳泉、晋中;南京保卫战(1937.11.12—1937.12.13),苏州、嘉兴、无锡、湖州、宣城、镇江、南京;徐州会战(1938.1—1938.5),蚌埠、枣庄、临沂、菏泽、徐州;武汉会战(1938.3—1938.10),武汉;南昌会战(1939.3—1939.5)、随枣会战(1939.5)、枣宜会战(1940.5—1940.6),南昌、随州、襄阳、宜昌;第一次长沙会战(1939.9.14—1939.10.10)、第二次长沙会战(1941.9—1941.10)、第三次长沙会战(1941.12—1942.1),长沙;豫湘桂战役(1944.4—1944.12),洛阳、郑州、许昌、长沙、衡阳、桂林、柳州等。
(9)相当于一个渐进式的DID设定,控制组为没有受到战争影响的城市,处理组为受到战争影响的城市。
(10)虽然中国在全面抗日战争胜利后又爆发了解放战争,但两次战争的本质是不同的,在保家卫国的抗日战争中,中国军人作战顽强,寸土不让;而在解放战争中,国民党人心尽丧,尤其在解放战争后期,大量国民党军人临阵起义,向解放军投诚。再者,两次战争对城市的影响程度是不同的,抗日战争中日本侵略者的目标是各大城市,城市经常成为大型战役的主战场;但中国共产党一直秉承“农村包围城市”战略,国民党军队在解放战争早期的进攻也主要集中在以农村、山地为主的解放区,中国共产党在解放战争中后期开始解放城市时,不仅通过谈判的手段尽可能争取和平解放,而且也非常注重对大城市基础设施的保护(例如,解放上海的战斗),人心涣散的国民党军队整体上并未进行强有力的反击。
(11)主要包括如下四个方面:(1)仅保留了抗日战争中几乎沦陷的东部沿海省份;(2)剔除了战争中表现活跃的“川军”和“滇军”的来源省份四川省和云南省;(3)仅保留了日本军队主要影响地区,即距离海岸线球面距离1000千米之内的城市;(4)在东部沿海省份样本的基础上补充了中部地区省份。结果均与表3保持高度一致,限于篇幅,具体结果不再列出,备索。
(12)导致变量不显著的原因可能很多,这一结果只能为工具变量的合理性提供一定佐证。
(13)我们使用了三种方案:第一种方案按照洪水实际影响,选定城市包括郑州、开封、许昌、周口、亳州、阜阳、淮南、蚌埠、徐州、淮阴、济南、商丘、淮安、宿州、扬州和镇江在2005年进入对照组;第二种方案涵盖了后续8年发生的黄河改道影响,增补城市济宁、菏泽、临沂和枣庄在2005年进入对照组,同时增补信阳、南阳、驻马店、淮安、盐城和日照在2010年进入对照组;第三种方案考虑到自然灾害可能造成人口流动的扩散,使用河南省、山东省、江苏省和安徽省的所有城市在2005年进入对照组。
(14)之所以没有对(1)式进行估计,是因为对过于大的样本量进行估计费时太久,而表2的基准结果已经说明(1)式和(2)式的估计在定性方面没有差异。
(15)这种方法本质上是一种子样本分析方法,每个分位点的估计都是前一个分位点的子样本估计,因此将参数估计值与前一个分位点的参数估计值进行对比就可以近似地反映该1%子样本中人口结构对出口的影响。具体来说,第x次回归对应Quantile≥x%的样本,从而第(x+1)次回归相当于第x次回归的子样本估计,这样一来,我们虽然无法直接对Quantile=x%的分样本进行估计,但通过比对第x次和第(x+1)次的回归结果,同样可以捕捉到Quantile=x%的分样本信息,即如果第(x+1)次所得到的参数值估计结果(即分布均值)更大,则有助于说明刚刚剔除的Quantile=x%相对于之后的样本,城市人口结构对出口的正向促进作用相对更小,进而与图中的趋势是一致的。类似的做法也可见于铁瑛和刘啟仁(2018)。
(16)当参数估计值无法在10%显著性水平下显著时,我们将其记为0。
(17)行业人力资本水平以本科以上学历雇员占总从业人员的比例来衡量。
(18)“城市-年份”联合固定可以吸收人口结构的水平项,以及来自城市层面的控制变量。
(19)利用2004年经济普查数据,将固定资产净值和从业人员数分别加总至GBT四位码产品(行业)层面并对应到HS六位码产品维度,而后以固定资产净值除以从业人员,可获得资本劳动比数据。
(20)为了进一步确保行业指标的外生性,我们利用2005年和2010年的数据,将原来的“城市-HS六位码产品-目的国-贸易方式”四维联合固定放松为三维联合固定进行估计,结果仍然保持一致。具体结果不再列出,备索。