Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We propose a monotonicity property for methods of estimating multiple structural breaks in linear regressions. A procedure with such a property yields a sequence of monotonically increasing sets of estimated break dates. Due to the uncertainty about the true number of breaks in finite samples, a monotone procedure offers a ranking of breaks from the least uncertain to the most. We propose a new method that imposes monotonicity. Monte Carlo simulations show that the proposed procedure works well in finite samples. We also apply the procedure to a study of the structural changes in the Fed’s monetary policy.