Date of Completion


Embargo Period



Efficiency, Bad Outputs, DEA, Industry Organization, Merger, Indian Banking, Indian Power Sector

Major Advisor

Subhash C Ray

Associate Advisor

Kathleen Segerson

Associate Advisor

Kankana Mukherjee

Field of Study



Doctor of Philosophy

Open Access

Campus Access


This dissertation focuses on optimal reorganization of production activities in different industries for a more efficient utilization of scarce resources. I primarily use mathematical programming and some resampling techniques on data sets from different industries in India to provide rigorous empirical evidence for the theoretical concepts and methodologies introduced in my thesis. In each of my three core chapters, I seek to answer important research questions. In my first core chapter, I derive conditions under which mergers induce a more cost efficient production of the target output bundle, especially when merging banks operate in the short-run. I show that potential cost economies from a merger can be attributed to three factors: convexity of the technology, sub-additivity of the ray short-run total cost curve, and a trade-off between reduction in the variable cost and an increase in fixed cost arising from an aggregation of the fixed inputs of the merging units. I use this proposed analytical framework within the non-parametric Data Envelopment Analysis to retrospectively evaluate potential gains from various recent bank mergers in India. The results are further corroborated using a smoothed multi-variate bootstrap procedure. The second core chapter provides a framework for intra-regionally reallocating resources and production targets among firms that produce a bad output alongside the good output. The objective is to increase the production of the good output in each region while minimizing use of polluting inputs and the level of the bad output, by increasing efficiency through reallocation. I use the model for a detailed audit of potential reduction in carbon emissions and increase in power generation in different states of India under alternative structural assumptions. Using the same data set as in the first chapter, my third core chapter focuses on efficiency in the Indian banking industry. The technology is modelled to specifically incorporate the effect the of credit quality on bank production. The research agenda in this chapter is to non-parametrically model the relationship between the volume and quality of credit.