Forecasting of armed conflicts is a critical area of research with the potential to save lives and mitigate suffering. While existing forecasting models offer valuable point predictions, they often lack individual-level uncertainty estimates, limiting their usefulness for decision-making. Several approaches exist to estimate uncertainty, such as parametric and Bayesian prediction intervals, bootstrapping, quantile regression, but these methods often rely on restrictive assumptions, struggle to provide well-calibrated intervals across the full range of outcomes, or are computationally intensive. Conformal prediction offers a model-agnostic alternative that guarantees a user-specified level of coverage but typically provides only marginal coverage, potentially resulting in non-uniform coverage across different regions of the outcome space. In this article, we introduce a novel extension called bin-conditional conformal prediction (BCCP), which enhances standard conformal prediction (SCP) by ensuring consistent coverage rates across user-defined subsets (bins) of the outcome variable. We apply BCCP to simulated data as well as the forecasting of fatalities from armed conflicts, and demonstrate that it provides well-calibrated uncertainty estimates across various ranges of the outcome. Compared to SCP, BCCP offers improved local coverage, though this comes at the cost of slightly wider prediction intervals.

It has occasionally been asserted that regulators typically overestimate the costs of the regulations they impose. A number of arguments have been proposed for why this might be the case. The most widely credited is that regulators fail sufficiently to appreciate the effects of innovation in reducing regulatory compliance costs. Most existing studies have found that regulators are more likely to over- than to underestimate costs. While it is difficult to develop summary statistics to aggregate the results of different studies of disparate industries, one such measure is the average of the ratio of ex ante estimates of compliance costs to ex post estimates of the same costs. This ratio is generally greater than one. In this paper I argue that neither the greater frequency of overestimates nor the fact that the average ratio of ex ante to ex post cost estimates is greater than one necessarily demonstrates that ex ante estimates are biased. There are several reasons to suppose that the distribution of compliance costs could be skewed, so that the median of the distribution would lie below the mean. It is not surprising, then, that most estimates would prove to be too high. Moreover, Jensen’s inequality implies that the expected ratio of ex ante to ex post compliance costs would be greater than one. I propose a regression-based test of the bias of ex ante compliance cost estimates, and cannot reject the hypothesis that estimates are unbiased. Failure to reject a hypothesis with limited and noisy data should not, of course, be interpreted as a strong argument to accept the hypothesis. Rather, this paper argues for the generation of more and better information. Despite the existence of a number of papers reporting ex ante and ex post compliance cost estimates, it is surprisingly difficult to get a large sample with which to make such comparisons.