An Empirical Approach for Evaluating Soft Budget Constraints
Per Pettersson‐Lidbom, Matz Dahlberg
Abstract
Open-access reader
Per Pettersson‐Lidbom, Matz Dahlberg
Abstract
Open-access reader
In this paper, we develop an empirical framework for detecting the existence and estimating the magnitude of the softness of budget constraint. The defining feature of the soft budget constraint is a subordinate organization’s expectations of being bailed out by a superior organization in case of financial trouble. This implies that one has to link the organization’s expectations for being bailed out to its fiscal behavior in order to quantify the extent of soft budget constraints. We postulate that expectations for bailouts are formed rationally and make use of an instrumental variable method to get consistent estimates of the parameters of interest. We argue that past own experience of being bailed out and bailouts of other subordinate organizations can be used to construct credible instruments for the formation of bailout expectations. We apply our empirical approach to a unique panel data set of 286 Swedish local governments where the central government provided a total of 1,697 bailouts between 1974 and 1992. Our results strongly suggest the existence of a soft budget constraint; a local government increases its level of debt by 6-10 percent if it expects to be bailed out with probability one as compared to when the likelihood is zero due to previous experience of being bailed out, while the effect on debt from bailouts of its geographical neighbors is roughly four times as large.
OpenAlex reports 43 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, we develop an empirical framework for detecting the existence and estimating the magnitude of the softness of budget constraint. The defining feature of the soft budget constraint is a subordinate organization’s expectations of being bailed out by a superior organization in case of financial trouble. This implies that one has to link the organization’s expectations for being bailed out to its fiscal behavior in order to quantify the extent of soft budget constraints. We postulate that expectations for bailouts are formed rationally and make use of an instrumental variable method to get consistent estimates of the parameters of interest. We argue that past own experience of being bailed out and bailouts of other subordinate organizations can be used to construct credible instruments for the formation of bailout expectations. We apply our empirical approach to a unique panel data set of 286 Swedish local governments where the central government provided a total of 1,697 bailouts between 1974 and 1992. Our results strongly suggest the existence of a soft budget constraint; a local government increases its level of debt by 6-10 percent if it expects to be bailed out with probability one as compared to when the likelihood is zero due to previous experience of being bailed out, while the effect on debt from bailouts of its geographical neighbors is roughly four times as large.
Key concepts: Bailout, Budget constraint, Constraint (computer-aided design), Debt, Set (abstract data type), Economics, Government (linguistics), Order (exchange)