Better Statistics, Better Economic Policies?

More and more economic transactions leave a "digital
footprint". This trend opens unprecedented opportunities for improving economic statistics and underpins demands to give statistical agencies far-reaching access to private-sector data. We analyze the consequences of better economic statistics in a political-agency framework that includes fundamental uncertainty about the impact of potentially welfare-enhancing reforms. We demonstrate that improvements in economic statistics can inhibit - rather than stimulate - reform attempts. With better statistics, the government is less likely to receive the "benefit of the doubt" if the numbers suggest its past reforms are failing. Reforms therefore come with a higher risk of electoral losses, implying that the government has stronger incentives to preserve the status quo. We identify political environments that are particularly vulnerable to this mechanism and contribute to the debate on private-sector data access.