4 ms·

Thing with this is that causality is so much harder to prove so any type of evidence-based policy is reduced to stabbing around in the dark hoping that the odds

Thing with this is that causality is so much harder to prove so any type of evidence-based policy is reduced to stabbing around in the dark hoping that the odds of the next happiness poll are not diluted by some random corruption scandal. The whole point of having measurable metrics and models is that they allow the use of the scientific method in policy making, whilst asking random people in the street might just get different answers whether you are asking them during or after their lunch break.

makeitdouble 4y ago
It’s the usual tension between an easy measure of something highly misleading and a harder to measure actually meaningful index. IMO going the easier route invariably brings along many other dangerous shortcuts. For instance here you mention “scientific method” for GDP when the measure in itself isn’t that reliable: what fits inside GDP isn’t fixed nor objective, and a lot of decisions happens upon survey. We have countries (most countries ?) regularly reshaping how they measure GDP to fit the narrative they need.
TheDong 4y ago
> causality is so much harder to prove [...] The whole point of having measurable metrics and models is that they allow the use of the scientific method in policy making A happiness index is hard to measure, agreed... but that's also not how GDP is used. Policy-makers will say "our research shows lowering taxes will do X and increase GDP", or "our research shows raising taxes will do Y and increase GDP". That happens now. I don't think we usually actually manage to tie an actual law or policy, as implemented, back to a change in GDP pretty much ever. It's not really the scientific method. What GDP gives us is a way to frame policy. If our number were "National Happiness", policies would be framed in terms of that. "Increasing taxes lets us make this work better, and increases happiness" and "Decreasing taxes gives people more money, which increases happiness". The same policies, but framed differently, and with different arguments as a result. That framing seems better to me, and how measurable and "scientific" it is seems like it would barely change.
- gruez 4y ago
  >What GDP gives us is a way to frame policy. >Policy-makers will say "our research shows lowering taxes will do X and increase GDP", or "our research shows raising taxes will do Y and increase GDP". That happens now. I don't think we usually actually manage to tie an actual law or policy, as implemented, back to a change in GDP pretty much ever. It's not really the scientific method. Strong disagree on this. Policies aren't just framed on "GDP goes up, we should do it", they're framed on "GDP goes up by $x at the cost of $y, so we should do it". A policy that claims to increase GDP, but where $x is smaller than $y is going to get shot down even if it sounds good. With GDP there's at least some way of checking your estimates. You can fudge your estimates up or down by changing your assumptions, but in the end you're still trying to pin down how much economic activity will occur and that can be cross-checked with tax records or government statistics. How are you going to do the same with GNP?

makeitdouble 4y ago

It’s the usual tension between an easy measure of something highly misleading and a harder to measure actually meaningful index. IMO going the easier route invariably brings along many other dangerous shortcuts. For instance here you mention “scientific method” for GDP when the measure in itself isn’t that reliable: what fits inside GDP isn’t fixed nor objective, and a lot of decisions happens upon survey. We have countries (most countries ?) regularly reshaping how they measure GDP to fit the narrative they need.

TheDong 4y ago

> causality is so much harder to prove [...] The whole point of having measurable metrics and models is that they allow the use of the scientific method in policy making A happiness index is hard to measure, agreed... but that's also not how GDP is used. Policy-makers will say "our research shows lowering taxes will do X and increase GDP", or "our research shows raising taxes will do Y and increase GDP". That happens now. I don't think we usually actually manage to tie an actual law or policy, as implemented, back to a change in GDP pretty much ever. It's not really the scientific method. What GDP gives us is a way to frame policy. If our number were "National Happiness", policies would be framed in terms of that. "Increasing taxes lets us make this work better, and increases happiness" and "Decreasing taxes gives people more money, which increases happiness". The same policies, but framed differently, and with different arguments as a result. That framing seems better to me, and how measurable and "scientific" it is seems like it would barely change.

gruez 4y ago
>What GDP gives us is a way to frame policy. >Policy-makers will say "our research shows lowering taxes will do X and increase GDP", or "our research shows raising taxes will do Y and increase GDP". That happens now. I don't think we usually actually manage to tie an actual law or policy, as implemented, back to a change in GDP pretty much ever. It's not really the scientific method. Strong disagree on this. Policies aren't just framed on "GDP goes up, we should do it", they're framed on "GDP goes up by $x at the cost of $y, so we should do it". A policy that claims to increase GDP, but where $x is smaller than $y is going to get shot down even if it sounds good. With GDP there's at least some way of checking your estimates. You can fudge your estimates up or down by changing your assumptions, but in the end you're still trying to pin down how much economic activity will occur and that can be cross-checked with tax records or government statistics. How are you going to do the same with GNP?