Short Abstract:

Using novel long-term district-level panel data (1961-2011) in India, this study investigates the effect of climate change and changing patterns of cultivated land use on household size through the channels of fertility and migration. Results show household size has negative relation with climatic alterations and positive relation with cultivated land share. Several robustness checks are applied.

Long Abstract:

Household is the basic unit of analysis for most of the socio-economic studies. In the demographic context household is the unit for study it three main components i.e., fertility, mortality and migration. On the other hand, changing patterns of climate in the last 50 to 70 years raised concerns not only in the field of physical sciences but social sciences as well. Major climate change and demographic studies to date focused on mortality and morbidity, and some studies on migration. Studies of climate change and fertility are also limited. This study is an attempt to capture all these three components together within an umbrella of household size. Household size can increase through low mortality and high fertility with low levels of outmigration, and it decreases with declining fertility and increasing outmigration. In the Indian context, this study examines the role of climate change and cultivated land use on changing average household size through the channels of fertility and migration.

The conceptual framework of the study states alterations in temperature and precipitation could increase migration due to degradation in agriculture and increasing morbidity. Having a higher share of agricultural or cultivated land could increase household size by engaging people in agriculture and higher labour demand in farms. A decrease in agricultural land use and the introduction of modern farm machinery can reduce labour demand and induce out-migration while lowering fertility, which eventually splits households and decreases the average size. Thus, with climate change there is a negative relation of household size and with agricultural land use there is a positive relation.

The study uses novel district-level panel data spanning from 1961 to 2011 in India. Major sources of data are the Census of India and Ministry of Agriculture and Farmers Welfare publications. Climatic Data has been taken from TerraClimate. The base year of study is 1961 and districts created after 1961 merged with their parent districts to create an unchanged boundary throughout the study period. I have used the Fixed Effect Panel Data Regression model to understand the within-district changes in household size through climate change and agricultural change. For the robustness of the results, I have used Panel Instrumental Regression, Spatial Panel Analysis and Path Analysis using Structural Equation Modeling.

India’s average household size has increased from 1961 (with a steady decline in mortality since the 1940s) to 1981 and started declining since then. Temperature has a steady increase throughout the period but precipitation alters. The most important note on precipitation is that the amount of precipitation increased but the number of days decreased. The cultivated land in India also increased till the 1990s and then started a slow decline. Results show that an increase in temperature and precipitation leads to a decrease in household size after controlling for all the socio-economic factors along with fertility and migration. On the other hand, cultivated land has a positive impact on household size. It can be said that the average household size transitions in tandem with the cultivated land share in India. In the pre-1980s era, an expansion of cultivated land due to population pressure and the absence of modern farm technology positively influences household size by stimulating fertility through increased labour demand and restraining migration by engaging the population in agriculture. However, since the 1980s, agricultural modernization has led to reduced labour demand and cultivated land share, resulting in decreased fertility and induced out-migration, leading to the fragmentation of households. Though with climate the relationship of household size is found negative throughout the study period. It is because alternations in climate especially precipitation hinder agriculture in seasons which causes farmers to migrate for jobs and food. Extreme conditions like frequent floods or drought years also increase the mobility of agricultural households. Migration of all household members is not possible and practicable. Migration of working-age household members especially males leads to the bifurcation of households and decreases the average size. Increased population pressure on land also plays an important role in reducing household size, though agriculture was found the major factor determined by climatic conditions, cultivated land use in district and agricultural modernization.