:19:08. > :19:17.And at the end, you have an assessment about how risky these
:19:18. > :19:24.forecasts are. If it is high risk, highly uncertain, making thd tax
:19:25. > :19:34.digital point, what level of uncertainty does that involve? We
:19:35. > :19:48.haven't tried, tempting as ht would be, to quantify it. Hi meditm is not
:19:49. > :19:57.scientific. It is difficult enough. It is based on past errors. Some of
:19:58. > :20:05.the uncertainties are in ardas where the data is not good, it is not one
:20:06. > :20:13.source of uncertainty you c`n calibrate, often a combinathon of
:20:14. > :20:17.very different things. We leave our live coverage from the
:20:18. > :20:23.Select Committee rooms to go over to the House of Commons. You c`nnot
:20:24. > :20:43.watch this committee live on our website. -- you can.