Development Economics
Industrial Organization
Applied Econometrics

Development Economics
Industrial Organization
Labor Economics

(Oral) Development Economics (with distinction), Industrial Organization, 2001
(Written) Microeconomic and Macroeconomic Theory, 2000

Firm Dynamics in the Presence of Financing Constraints: Ghanaian Manufacturing

Professor Christopher Udry
Professor Christopher Timmins
Professor Steven Berry

May 2004

M.Phil., Economics, Yale University, 2001
M.A., Economics, Yale University, 2000
Staatsexamen, Mathematics and Geography, University of Cologne, Germany, 1996
Vordiplom, Economics, University of Cologne, Germany, 1996

Social Science Research Council (SSRC)-Program in Applied Economics Fellowship, 2002-2003
John Perry Miller Fund Award, 2002
Ryoichi Sasakawa Fellowship, 2001
Yale University Fellowship, 1999-2003
Dr. Prill-Prize for undergraduate thesis (University of Cologne), 1996

Teaching assistant: Yale University, Department of Economics, Spring 2002
     Development Economics (International and Development Economics Master’s Program)
Teaching assistant: University of Cologne, Germany, Department of Economics, Summer 1999
     Macroeconomics/growth theory and introductory macroeconomics for undergraduates
Teacher: Realschule Geldern, Germany, Spring-Summer 1998
     Taught mathematics and geography at a high school

Consultant: The World Bank, Development Economics Research Group (DECRG)
     Estimated disaggregated poverty and inequality measures combining census and household survey data
        ("poverty map") for South Africa, Summer 2001
     Prepared a forecast of poverty and inequality for 1999-2020 for Madagascar, Fall 2000
     Built a household level data base to examine changes in inequality in Africa, Summer 2000
Research assistant: Professor C. Timmins, Yale University, 2002-2003
Research assistant: Professor P. Funk, University of Cologne, Economics, Winter 1998/99
Research assistant: Professors Radtke and Wiese, University of Cologne, Geography, 1995-1997

Ghana: Dissertation related work, July 2003
     Informal interviews with entrepreneurs at manufacturing firms, industry representatives and bank managers
Kenya: Consultant, International Livestock Research Institute (ILRI), Nairobi, and World Bank, October 2001
     Worked on estimation of disaggregated poverty and inequality measures for Kenya
     Collaboration in an ILRI project into the relation of poverty and land-use patterns in Kenya
South Africa: Consultant, World Bank, Development Economics Research Group (DECRG), June-August 2001
     Estimated disaggregated poverty and inequality measures for South Africa (see also above)
     Trained government personnel from Tanzania and Kenya in a weeklong workshop on "poverty map"
        methodology
South Africa: Intern, Department of Education Mpumalanga Province / LIB (German Development Organization), October 1997 - January 1998
     Evaluated results from a preliminary survey on vocational training needs and prepared a revised
        questionnaire

Modeling Firm Dynamics to Identify the Cost of Financing Constraints in Ghanaian Manufacturing. [Job market paper]

Inequality, Ethnic Heterogeneity, and Public Goods - Theory and Evidence from Kampala, Uganda. Manuscript, 2003.

Are Immigrants More Mobile than Natives? Evidence from Germany. Manuscript, 2003.

Precautionary Savings and Self-selection - Evidence from the German Reunification ‘Experiment’ (with Nicola Fuchs-Schündeln). Manuscript, 2003.

The Savings Behavior of East and West Germans - Theoretical Predictions and Empirical Evidence (with Nicola Fuchs-Schündeln), Journal of Applied Social Science Studies (Schmöllers Jahrbuch) 123(1), 2003, 209-220.

Estimating a Multistage Production Process with Sequential Labor Decisions Under Uncertainty. Manuscript, 2001.

Understanding the Economic Geography of Poverty: Methodological Approaches and Challenges Illustrated by a Case Study of Kenya (with P. Kristjanson, J. Mistiaen, P. Thornton, F. Place) (in progress). 2002.

Tor Vergata University, Rome: Villa Mondragone Workshop in Economic Theory and Econometrics, 7/2003
SSRC Fellows Conference, Santa Cruz, California, 5/2003
WIDER Conference on Poverty, Inequality, International Migration and Asylum, Helsinki (Poster), 9/2002

German: native; English: fluent; French: intermediate; Spanish, Dutch: basic reading and speaking abilities

Economic development requires the growth of productive firms. Theoretical considerations and existing empirical evidence however point to the existence of financing constraints that limit firms’ investment abilities. Despite the potentially large welfare implications, relatively little is known about the cost of financing constraints to individual firms or about their aggregate implications. This motivates the research question that is at the core of the dissertation: what is the cost of financing constraints to firms and what are the aggregate implications? The dissertation quantifies the upper bound of the effect of potential policies that attempt to close the wedge between the cost of internal and external finance. Specifically, I study the behavior of manufacturing firms in Ghana. Since growth is a main goal of economic policy, the results of this work have immediate policy relevance. Further, given the multitude of activities and the amount of money spent on support for enterprise development in developing economies -- especially to support firms’ ability to borrow to finance investment -- the question is particularly relevant for, but not confined to, developing economies.

Chapter one of the dissertation states an apparent puzzle. First, this chapter demonstrates that the marginal returns to capital in Ghanaian manufacturing firms are high for a majority of firms. Using panel data, production functions are estimated to calculate the returns to capital. In the second part of this chapter investment behavior is analyzed. While the first part suggests that the average firm has profitable investment opportunities, the analysis of investment behavior shows that investment levels are low and generally uncorrelated with the estimated returns to capital, except for the largest firms. Hence, more productive firms do not grow faster than other firms.

The second chapter of the dissertation quantifies the role of financing constraints as an explanation for this puzzle. This chapter starts out by providing evidence for a significant role of financing constraints in explaining firm investment behavior. I show direct survey evidence for the existence of financing constraints as well as econometric evidence that uses a version of a standard test for financing constraints. While several other papers have shown that financing constraints matter for manufacturing firms, both in developed and developing economies, the effects of these constraints for firm dynamics have not been estimated.

The central goal of the second chapter is to quantify the effects of financing constraints along several dimensions, in order to help better understand their importance. To this end I develop and estimate a structural dynamic model of firm-level investment. The model explicitly models the firms’ real activities, i.e. the production process and investment decisions, together with the financial side, i.e. the decision of how to use profits (keep them to build up internal funds vs. dividend payments) and how to finance investments (internal vs. external funds). I explicitly model the latent interest rate that firms of different types face. The model also allows for other potential explanations of the puzzle observed in chapter one, in particular adjustment costs and uncertainty. Finally, the model allows for exit of firms.

By modeling and estimating both the real side and the financial side of the firms’ activities simultaneously, the model allows me to deal with the main identification problem faced by tests for and quantifications of financing constraints, namely to identify the investment opportunities and the constraints of a firm separately. The identification of the production function and unobserved productivity relies on the assumption that labor is chosen as the solution to a static optimization problem. Evidence from the survey is presented to support this assumption. Given an estimate of the production function, the investment opportunities can be determined as the solution to a dynamic optimization problem. These can be used to identify the constraints that a firm faces. Intuitively, the separate identification of constraints that are common to all firms, e.g., adjustment costs, vs. financing constraints that only apply to firms with low levels of internal funds, is achieved by the fact that some firms are able to invest optimally out of internal funds, while others are not.

The model is solved using dynamic programming methods and is estimated via simulation methods using the same firm level panel data from Ghana that was used in chapter one. The estimation results indicate that the per-unit cost of credit decreases in the capital stock of a firm and increases in the amount that a firm borrows. These results are consistent with conventional models of imperfect credit markets. Counterfactual analyses are then carried out to answer the main research questions, i.e., to quantify the importance of financing constraints. These counterfactuals indicate that removing the constraints would imply economically significant increases in investment that are associated with higher levels of consumption.

Chapter three [work in progress] explores some methodological implications of financing constraints for production function estimation. I review recent advances in the estimation of production functions and show that these methods rely on the ability of the firm to finance the optimal investment and/or intermediate inputs. I demonstrate that this assumption is violated if firms are financially constrained, which suggests controlling for financial factors to deal with this methodological problem.  Simulations using the parameter estimates of the model in chapter two demonstrate the potential size of the bias.