Abstract
This paper studies the consequences of predation when firms deploy guard labor as a means of protecting themselves. We build a simple model and combine it with data for 144 countries from the World Bank enterprise surveys, which ask about firm-level experiences with predation and spending on protection. We use the model to estimate the output loss caused by the misallocation of labor across firms and from production to protection. The loss due to protection effort is substantial and patterns of state protection at the micro level can have a profound impact on aggregate output losses. Various extensions are discussed.
Citation
References
References go here
Online appendix
Citation
Supplementary materials
- Code and Data
2018. "Replication data for: Predation, Protection, and Productivity: A Firm-Level Perspective." American Economic Association [publisher] DOI: 10.1257/mac.20160120.data - Data is freely accessible under CC BY-NC 4.0 at 10.1257/mac.20160120.data.
- Data
2015. "Firm-level statistics 1975-2013 [dataset]" Norwegian Data Archive [curator], v2. DOI: 10.7654/nda::7643A::34 - Data restricted-access, under Norwegian Data Access license (has residency requirement, has citizenship requirement), accessible at Norwegian Data Archive in Oslo, Norway
Notes
Licensing
- A taxonomy of frequent/common licenses should be referenced
- For non-standard licenses, a repository should be created and referenced. Licenses from that repository, as well as from expert guidance (Michigan 2015) should be reusable.
- Concerns about intellectual property need to be explicitly addressed here
Access rights
- Access rights should come from a taxonomy, with explanatory statements easily accessible. Simplification here should be fine, augmented with a list of explanatory terms, such as restricted-access, open access.
- Note: proprietary is a licensing term, not an access term.
Access methods
- Methods also need both a taxonomy (download, at location) and explanatory text ("at 10.1257/mac.20160120.data", "at Norwegian Data Archive in Oslo, Norway").
Metadata
The entire information set should be driven (in the display) by XML or JSON schema and an auxiliary file, should also be parseable (by web scrapers) or searchable (Google and friends).- We are developing a schema
- This web page serves as a mock-up, but should ultimately be driven by the schema (for instance using a XSLT or JSON parser)
- The technology should be implementable on simple websites, e.g.
in the header of the website, and a JS library.< link rel="resource" href="https://doi.org/10.1257/mac.20160120.mjsd.xml" >