Econometrics and the Science of Climate Change 30 th June 2011 Timothy Curtin

Econometrics has a long history as the technique of choice for testing the merits of alternative hypotheses across most of the social sciences as well as many of the natural and materials sciences, not to mention pharmaceutical science, where it is widely used to evaluate the efficacy of alternative medications, including the use of placebos as counterfactuals. However its greatest value is in the social sciences where laboratory experiments are not feasible, but least squares linear regression can be used to assess the relative significance of alternative independent variables as explanatory factors. The founding texts of climate science, John Tyndall (1861) and Svante Arrhenius (1896), discovered and estimated the radiation absorption effects of what they called aqueous vapour and carbonic acid (now known as water vapour and carbon dioxide), unlike the oxygen and nitrogen that comprise the bulk of our atmosphere. Tyndall‘s experiments showed that the most powerful radiative effect was that of water vapour. Arrhenius also included water vapour in his more theoretical analysis. Now however most climate scientists‘ models assume that anthropogenic addition to the atmospheric concentration of carbon dioxide (hereafter denoted [CO2]) and of certain other greenhouse gases like ozone and methane (in aggregate denoted as [CO2e]) is the major determinant of climate change, and have relegated Tyndall‘s primary role for atmospheric water vapour (hereafter [H2O]) to having only a secondary, or ―feedback‖, effect arising from the higher temperatures supposed to result primarily from increasing [CO2].