Thank you for the thoughtful response! I agree with you. It’s certainly not my view that EAs should focus solely on researching how they could improve the Civil Service and politics. I agree that there are many other legislative, fiscal and other ideas that could be pursued to make government analysis more long term and better. We do point to the ‘three layers’ proposal against x-risk as potentially promising (and indeed the step in that direction is one of the best parts of the Bill in my view, although I think it needs to be improved). We agree that substantial resources should be focused on investigating new lines of approach. I’m afraid we don’t have developed suggestions on what those should be, sorry – and as we say the answers may have to be highly sector-specific – but one of our hopes is that this will encourage more attention on that.
‘Many’ was carefully chosen not to imply more than half—perhaps very substantially less, depending on the metric. I started this analysis by assuming a base rate of less than half, but then updated my view based on granular analysis of the proposed Bill language and what we have been able to learn about the theory of change.
‘Many’ was carefully chosen not to imply more than half—perhaps very substantially less, depending on the metric. I started this analysis by assuming a base rate of less than half, but then updated my view based on granular analysis of the proposed Bill language and what we have been able to learn about the theory of change.
Then it seems to me that it would have been just as accurate to say: ‘many efforts to improve policy in the last 50 years have succeeded’ and conclude the opposite.
Thank you for your comment. I agree we could have said ‘many efforts to improve policy in the last 50 years have succeeded’. However, given our substantive analysis of the Bill, I think we would have ended up with the same concerns about its potential outcomes. In view of the impression that some people who do not work in policy or government seem to have that attempts to improve policy generally or always move things in the intended direction, we thought it helpful to highlight the risk of unintended consequences. The alternative formulation would not have made that point as clearly.
Thank you for the thoughtful response! I agree with you. It’s certainly not my view that EAs should focus solely on researching how they could improve the Civil Service and politics. I agree that there are many other legislative, fiscal and other ideas that could be pursued to make government analysis more long term and better. We do point to the ‘three layers’ proposal against x-risk as potentially promising (and indeed the step in that direction is one of the best parts of the Bill in my view, although I think it needs to be improved). We agree that substantial resources should be focused on investigating new lines of approach. I’m afraid we don’t have developed suggestions on what those should be, sorry – and as we say the answers may have to be highly sector-specific – but one of our hopes is that this will encourage more attention on that.
‘Many’ was carefully chosen not to imply more than half—perhaps very substantially less, depending on the metric. I started this analysis by assuming a base rate of less than half, but then updated my view based on granular analysis of the proposed Bill language and what we have been able to learn about the theory of change.
Then it seems to me that it would have been just as accurate to say: ‘many efforts to improve policy in the last 50 years have succeeded’ and conclude the opposite.
Thank you for your comment. I agree we could have said ‘many efforts to improve policy in the last 50 years have succeeded’. However, given our substantive analysis of the Bill, I think we would have ended up with the same concerns about its potential outcomes. In view of the impression that some people who do not work in policy or government seem to have that attempts to improve policy generally or always move things in the intended direction, we thought it helpful to highlight the risk of unintended consequences. The alternative formulation would not have made that point as clearly.