It is increasingly recognized that well-defined property rights are crucial for realizing the benefits of market exchange and that such rights are not exogenously given but evolve over time in response to economic and political forces. The reduction of expropriation risk and the facilitation of market transactions are the two main categories through which property rights systems affect economic outcomes. However, the mechanisms by which these two categories affect outcomes differ in important ways.
As long as minimum legal recognition is enforced, adjudication processes that are based on local consent can significantly reduce expropriation risk. In contrast, full realization of gains from trade requires that reasonably complete, and current information on the assignment of property rights -which is normally provided by public registries- be available at low cost to agents with sufficiently diverse skills to allow efficiency-enhancing transactions. Credit impacts from land titling or registration would be expected only if such efforts are comprehensive, registries remain up to date over time, and third parties, such as mortgage lenders, can access reliable registry information at low cost on a routine basis. The fact that many previous studies find significant investment but no credit effects from land administration interventions could thus imply that these interventions managed to increase tenure security at the local level but failed to create the pre-conditions for efficiency-enhancing trade.
In a recent paper (Deininger and Goyal 2012), we focus on a program that computerized land registries in the South Indian state of Andhra Pradesh. We argue that, although it will not alter the quality of underlying property rights information, computerization can reduce the cost associated with keeping the property register up to date, eliminate informal side payments that have traditionally been associated with property registration, and improve third party access to registry information. These could potentially increase the volume of registered land transactions, reduce the level of informality, and thus improve registry comprehensiveness. The program we studied also made records and abstracts of past transactions available online to interested third parties including commercial banks, thereby allowing them to ascertain ownership status or existence of pre-existing liens on any property that might be offered as collateral.
While we expect that state-wide computerization of land registries can help improve credit access, institutional factors characteristic to India imply that such an effect will be stronger in urban settings than in rural where it may even be completely absent. In urban areas, for instance, land registries are the only available form of documentary evidence on land ownership. In rural areas, by contrast, land records maintained by revenue department officials at the village level are more accessible than land registries maintained at the taluk (block) level, and thus more frequently used and up to date. Although AP computerized land records about a decade ago, it failed to either maintain or link them to registry information. This limits potential impacts of registry computerization even under ideal circumstances -and acts as a warning against viewing computerization as a miracle cure to the ills of land administration systems.
To test whether registry computerization affected credit access, we used variation in the date of shifting to fully computerized operations across the state’s 387 sub registry offices (SROs), a process that started in February 1999 and was completed in March 2005. The main dependent variable is the volume of credit disbursed by all scheduled commercial banks, based on bank’s mandatory reporting to the Reserve Bank of India, on a quarterly basis for the 1995 to 2007 period. Data on the volume of different types of registered land transactions at the SRO level allow us to independently test the paper’s main hypothesis.
The results from our study point towards a significant impact of computerization on credit in areas that are fully or partially urbanized but not in entirely rural ones, with the point estimate suggesting that, in a completely urban area, access to credit increased by 10.5% on average. The effect appears to grow stronger with time; for a fully urbanized area, we estimate increases of approximately 3% and 5% in the two years immediately following computerization and 15% thereafter.
Data on the volume of registered land transactions support the hypothesis of computerization having led to an expansion of credit access in urban but not in rural SROs; point estimates suggest that, in entirely urban SROs, the number of registered mortgages increased by 18 percentage points in the two years immediately after computerization and by 32% thereafter. The fact that these numbers are larger than the estimated increase in the volume of credit could imply that the intervention helped expand coverage rather than only provide higher volumes of credit to existing borrowers. Computerization had no effect on either the volume of registered land sales or that of non-monetary land transfers such as gifts, inheritances, and leases, supporting our argument that reduced transaction cost of institutional providers of credit in urban areas could act as a main channel for direct effects of computerization to materialize.
The fact that the effects are limited to a narrow segment of urbanized locations without other means of documenting land ownership suggests that the extent to which changes to property rights will affect outcomes is context specific and the magnitude of impact will depend on the characteristics of existing property institutions, and alternative options to document property available to people.