Essays on misallocation

Guide(s)

Subramanian, Chetan

Department

Economics

Area

Economics

University

Indian Institute of Management Bangalore

Place

Bangalore

Publication Date

3-31-2024

Year Awarded

March 2024

Year Completed

March 2024

Year Registered

June 2018

Abstract

One of the most important questions in economics asks, "Why are some countries so rich and others so poor?". The macro development literature (Klenow and Rodriguez-Clare, 1997; Prescott, 1998; Hall and Jones, 1999) argues that the primary source of differences in income per capita is differences in total factor productivity (TFP) of economies. This poses the all-important question: what generates differences in TFP? Researchers have offered different reasons for low TFP in poor countries: slow adoption of productive technology (Nelson and Phelps, 1966; Parente and Prescott, 1994) o firms unable to operate technologies efficiently (Parente and Prescott, 1999; Schmitz Jr, 2005). These explain why individual firms in poor countries have low efficiency relative to those in rich countries and how it leads to lower aggregate TEP. A new and exciting line of work looks at the misallocation of inputs across heterogeneous production units (Restuccia and Rogerson, 2008; Hsieh and Klenow, 2009; Hopen-hayn, 2011) and how it leads to aggregate TFP reduction. As more productive units get fewer resources while unproductive units get more, both aggregate TFP and output decline in the economy. Such allocation distortions can be due to a variety of reasons, including financial frictions, trade restrictions, and a host of regulations (Restuccia and Rogerson, 2013). In this Ph.D. thesis, I study two distinct types of resource misallocation. The first aspect is related to sub-optimal land allocation to various crops by Indian farmers, which results in i) reduced crop yields, and ii) a higher-than-optimal usage of land in the Indian agricultural sector. As a result, land becomes scarce, adversely affecting industrialization and urbanization. In the first essay of the thesis, I measure the total extent of land misallocation between the agricultural and non-agricultural sectors of the Indian economy. That is the amount of agricultural land that can be released for non-agricultural purposes if a social planner optimally allocates various crops across heterogeneous land. The analysis suggests that the Indian farmers suboptimally allocated crops across land with heterogeneous productivity. In a baseline scenario where the planner uses the most conservative level of human inputs (rainfed water supply and low level of complementary inputs), I find that up to 13 million hectares, or 20 percent of agricultural land, can potentially be released across all districts in India. The quantum of land that can be released rises substantially under alternative scenarios. In the second essay of the thesis. I examine the welfare impact of a land-augmenting productivity increase in the agriculture sector due to the spatial reallocation of crops as measured in the first essay. I use a two-sector model setting where both agriculture and manufacturing sectors use land and labor as production factors. The movement of both factors across sectors involves friction. In such a setting, following an increase in agricultural productivity in the model, I find that i) both land and labor move from agriculture to the manufacturing sector, (ii) land prices decline while wages go up, (iii) both agricultural and manufacturing output goes up. (iv) agricultural prices go down. and (v) real income in the economy increases by roughly 12 percent. The second aspect is connected to the misallocation of financial capital because of imperfections in financial markets. It stems from the inefficiency of credit markets, where creditors do not ex-ante know or accurately predict the efficiency of a firm in an investment project for which it borrows money. Additionally, there is also information asymmetry and enforcement limitations. Therefore, creditors lend based on certain firm-level financial variables rather than the capability or productivity of a firm. Broadly, there are two types of external borrowing raised by a firm: based on the liquidation value of a firm's physical assets, or its operational earnings (EBITDA). The third essay of thesis looks at both the micro- and macro-level implications of different types of borrowing constraints in the US credit markets. Though the majority of non-financial corporate debt is cash-flow based in the US as found by (Lian and Ma, 2021), most macro models assume collateral-based borrowing while quantifying the impact of borrowing frictions like in (Buera et al., 2011; Midrigan and Xu, 2014). At the micro-level, I compare the performance of an earnings-based borrowing constraint (EBC) with a collateral-based one (CBC) in explaining key stylized facts of firms in the US manufacturing sector. When US manufacturing firms are classified using the textual analysis-based index of debt constraints as developed by (Hoberg and Maksimovic, 2015), I find that debt-constrained firms (i) have a higher debt-to-earnings ratio, (ii) have a lower debt-to-asset ratio, (iii) are more productive, (iv) are not necessarily small, and (v) have a lower net worth. In addition, I also find that the correlation between size and the marginal revenue product of capital is close to zero. Using a static input choice model with both EBC and CBC, I find that the EBC does a much better job of capturing the above-stylized facts. At the macro level, using a simple dynamic framework, I link the misallocation of capital to the type of loan contract. The results indicate that the loss in TFP is lower by about 40 percent under EBC when compared to CBC.

Pagination

x, 109p.

Copyright

Indian Institute of Management Bangalore

Document Type

Dissertation

DAC Chairperson

Subramanian, Chetan

DAC Members

Dasgupta, Kunal; Thampy, Ashok; Murali, Srinivasan

Type of Degree

Ph.D.

Relation

DIS-IIMB-FPM-P24-13

This document is currently not available here.

Share

COinS