In his 2019 Nobel Prize acceptance paper, William Nordhaus (2019) highlighted the uncertainty over climate damages by using two completely different damage functions: Nordhaus and Moffat (2017) and Howard and Sterner (2017). Despite their vastly different implications for climate policies, both were estimated using the meta-analysis technique, a method long considered the objective and scientifically rigorous way for combining results from multiple studies to develop a consensus estimate. This paper demonstrates that this disparity stems from differences in both methodological decisions (addressing methodological impacts and heteroskedasticity) and subjective decisions (about data search, selection, and weighting). Combining the two data sets, applying Nordhaus’s quality weights, and applying the best methodological practices, we find damages of approximately 7 to 10 percent of gross domestic product (GDP) for a 3°C increase in global average surface temperature, depending on the inclusion of catastrophic and productivity impacts; the result is relatively robust to alternative data selection, weighting, and methodological assumptions. However, subjective differences between the weighting assumptions of the two earlier studies are still unresolved. To address subjectivity, this paper makes transparent existing weighting rules, develops new weighting rules, and applies a recently developed quality effects estimator from the medical literature (Doi et al. 2015). Operationalizing these rules, we demonstrate that damages are approximately 7 to 16 percent of GDP for a 3°C increase, though the upper end of this range, which includes catastrophic and productivity impacts, is sensitive to the selected model and weight specification.