Recipient Organization
MICHIGAN STATE UNIV
(N/A)
EAST LANSING,MI 48824
Performing Department
Agricultural, Food, & Resource Economics
Non Technical Summary
In the past few decades, many developing countries have taken measures to improve the condition and performance of land and labor markets. In China, the dual system of rural and urban divide in its land property rights has created a true divide in rural and urban areas in social and economic aspects and has increasingly become a major obstacle to China's future economic development. To explore innovative ways of addressing these issues, Chinese government has been piloting with various new reforms/policies to reduce the rural-urban divide and increase rural residents' chances to participate in and benefit from the country's economic development. In the end of 2016, the Chinese government announced a major policy change related to China's agricultural land, which is known as "Three Rights Separation". Under this new policy, the ownership of agricultural land is still with the village collectives, but the contract rights and management rights become two separate rights that are potentially given to two separate entities. While villagers who contracted land from village collectives still have the contract rights, the management rights of land go to land users who can be the villagers contracting land or tenants/agricultural companies who leased in the land from the villagers. The main purpose of this new land policy is to protect the management rights of the land users especially in the case when the users are tenants or agricultural companies. In Vietnam, government have taken various measures to improve land tenure security and promote land rental markets over the past several decades. And with the economic growth, land and labor are more closely related than before. In India, land markets are found to perform under its potential and labor markets including migration are found to be not as active as other countries with similar level of development, partly due to the unique social and culture features of India. A recent study also found that female labor market participation has been stagnant or declined. The poor participation and underperformance of land and/or labor markets in India, Vietnam and China are likely to cause the misallocation of resources that have far reaching implications on productivity and economic growth. In the meantime, all three countries are undergoing drastic social and economic transformation. The urbanization has been unprecedented, and the governments invested enormous amount of money to improve infrastructure conditions as their top policy strategies to achieve economic growth and reduce poverty. According to statistics from the Chinese Ministry of Transport, at the end of 2017, the country's total highway mileage reached 4.7735 million kilometers, 5.2 times that at the end of 1984. Among them, the length of expressways reached 136,500 kilometers, ranking first in the world. The rising of wage and loss of labor to completing sectors in combination with the much improved infrastructure conditions have induced some organization and institutional changes (e.g., rapid emergence of agricultural production services, agricultural cooperatives), as well as changes in production technologies (mechanization, scaled production), and market innovations (e-commerce). It is well established in the institutional economic theory that secure property rights provide better incentives for efficient use of resources and long-term investments, and therefore lead to increased productivity and economic growth. Better property rights including women's rights to own and inherit assets are essential to achieve more inclusive development and more harmonious society. Less restrictive labor policy is also essential to ensure a success economic transformation and more inclusive economic development. Meanwhile, the huge improvement in transportation infrastructure and implementation of rural pension program are expected to have significant effects on land and labor markets and consequently on economic development. However, a key challenge is how the impacts of these various policy changes can be scientifically measured and whether and how the causal relationship between the road infrastructure (or implementation of the rural pension program) and farmers' decisions in land transfer, labor employment can be identified. It is in these areas where my research in the next five years will focus on by linking theoretical concepts and literature to rigorous empirical methods and carefully designed survey instruments that are implemented with proper attention to assuring data quality. Under the general goal of explore determinants and impacts of the functioning of the land and labor markets in three important developing countries, we aim to achieve six specific objectives. To meet the overall goal and the specific objectives, we propose to use various rigorous econometrics techniques to analyze household/community level panel data and/or data created from experiments. The research will not only yield high quality journal articles, policy briefs, reports that will be widely disseminated through a variety of dissemination modes, the findings will provide policy guidance to government in their future design and implementation of land and labor policies in developing countries.
Animal Health Component
85%
Research Effort Categories
Basic
10%
Applied
85%
Developmental
5%
Goals / Objectives
The overall goal of this proposed study is to use rigorous econometrics and impact evaluation methods and high quality household and community level survey data to investigate a few important but largely understudied development topics related to land and labor issues, the interrelation between labor and land, the effect of infrastructure and governance on labor allocation and labor participations with a geographical focus in East and South Asia. To achieve this overall goal, there are 6 specific objectives:Objective 1: Quantify the willingness to pay for the "Separation of three rights" from the perspective of different types of land users (e.g., individual farmers versus agricultural companies), as well as the willingness to accept for the policy from the perspective of villagers who contracted land from their village collectives.Objective 2: Examine the effects of the implementation of the "Separation of Three Rights" on a wide range of outcomes, such as the frequency and quality of land actual transfers, contract terms, land use, land investment, agricultural productivity, household income, etc.Objective 3: Analyze the effect of off-farm labor demand on land rental markets in the context of Vietnam.Objective 4: Evaluate the effects of the Chinese rural pension program on senior farmers' behaviors in land transfer, and explore the heterogeneity effects of the pension programs across a number of households and member characteristics.Objective 5: Evaluate the effect of political reservation on female labor force participation and explore the mechanisms through which the effects take place. Are there supply of labor effects? Are there female empowerment effects?Objective 6: Examine the effects of road investment on household's employment and income growth. Unlike in the literature, we will examine the relative importance of different types of roads (high way, provincial road, county road, township road, village road) to better inform government's future infrastructure policies.
Project Methods
To meet the objectives in the previous section, both conceptual and empirical approaches will be used in the proposed research. While the randomized control trial (RCT) is considered to be the "golden rule" in establishing a causal relationship between a policy intervention (e.g., titling program) and an outcome indicator (e.g., investment), but for various practical reasons RCT is not always feasible. A quasi-experimental design is commonly used when RCT is not possible in practice. In the proposed research, we will mainly rely on the combination of quasi-experimental design and regression analysis. Specifically, the following methods and approaches will be used to meet each of the individual objectives:To achieve objectives (1), we will use the choice experiment (CE) to estimate the willingness to pay of different types of land users for the new policy of separation of three rights. Whether the land users is another individual farmer who leased in land from the contractor or an agricultural company is likely to be different. Therefore, it is important to solicit willingness to pay from different types of land users. A separate CE will also be implemented to estimate the willingness to accept for the change of land policy. This study is expected to start either the fall of 2019 or the summer of 2020.To achieve objective (2), we will use the existing panel data sets from multiple sources and panel fixed effect estimation methods to evaluate the impact of the new land policy on a variety of outcomes. The candidate datasets include the longitudinal rural residents survey data collected by NBS, or the fixed point panel data set collected by the Chinese Research Center of Rural Economics (RCRE), or the Chinese Household Income Project Survey (CHIPS) by the Southwestern University of Finance and Economics in collaboration with Zhejiang University. The common feature of these data sets is that they are all long panel that covers the period before the policy was introduced and after the policy was implemented in some areas, which allow us to use standard DID or DID with time varying treatment to identify the effects of the policy change on a variety of outcomes such as the frequency and formality of land transfer, labor allocation, agricultural production, household income, consumption, etc.To achieve objective (3), we will use the combination of household level panel data from panel data from household survey and the enterprise census data over time. The enterprise census data over time allow us to create the key right hand side variable - the total the number of enterprises (with workers more than 10) and the total number of workers in own district or nearby districts. Econometrically, we will use an ordered probit/logit model to estimate household's overall rental participation decision (rent-in, rent-out, or staying autarkic) as a function of labor demand from the enterprises and other household and community control variables (farming skills, land endowment, head's age, education, etc.). To address the potential endogeniety problem of the labor demand variable, we adopt the correlated random effects ordered probit model (CRE-ordered probit) based on the assumption that the unobservables are correlated with the means of the time varying variables.To achieve objective (4), we will first build a simple household model to derive the hypothesis on the effects of pension income on senior farmers' behaviors in renting in or out land. We posit that the effect of pension income on household's rental decisions through the completing effects of pension income on leisure and on labor used in own production. Our main identification strategy is a regression discontinuity design (RDD) using data from the 2013 round of CHARLS, the year when NRPS was extended to the entire country. The RDD approach exploits the fact that the eligibility of NRPS is strictly determined by a cut-off age (eligible if 60 or above, ineligible otherwise). Essentially, the RDD estimates compare household's renting behaviors between households with at least one member just above the cut-off age (e.g., 60-70) and those with the most senior member just below the cut-off ages (e.g., 50-60). To complement the RDD approach, we also conducted difference-in-difference (DID) analysis using the panel data.To estimate how the female empowerment in response to female reservation policy in local election on female labor participation (objective 5), we will use the Rural Economic and Demographic Survey (REDS) data and take advantage of the fact that the implementation of female reservation was randomly implemented across villages. The recent round of the REDS was conducted in 11 major states of India in 2014-15. The 2014-15 data and the data from earlier rounds of survey (1999 and 2006) are formed panel for the analysis. The total sample size is roughly 7500 households across 238 villages in 17 major states. Since village councils or Gram Panchayats (GPs) are randomly reserved in each election period for women in all states of India, we can simply compare the villages that are reserved to those that are not reserved. In an equivalent regression context, the base regression is to regress labor participation on reservation dummy, interaction of reservation dummy and female members, and other controls of individual characteristics. To explore the short-term versus long-term effects, the base model is augmented by including reservation at both the current period and previous period are also included. Additional regressions were run to explore the mechanisms of the effects to disentangle the potential labor supply effect of the reservation policy from the empowerment effect of the reservation policy.Finally, to achieve objective (6), the main empirical approach is the fixed-effect panel regression. In the econometrics model, we regress different outcome variables (employment, agricultural production, income, etc) on a vector of road infrastructure variables (nearest distances from village center to different types of road, total length of each type of road, etc.), while controlling for a large number of community and household characteristics. The advantage of panel fixed-effect estimation is that all the time constant unobservables at both the household level and community levels are automatically controlled. However, time-varying unobservables are not controlled for in a basic fixed effect model (base model in our case). In an augmented model, we also add the interaction terms of village dummy variable and the time dummy variable to the base model to further reduce the potential bias associated with the time-varying unobservables at the village level. However, the potential biases arising from the omission of individual household-specific time-varying unobservables may still be a concern. Our strategy is to include as many control variables as possible and also explore the possibility of using instrumental variables.