I’d like to try to make a point here that I’ve made before, but hopefully make it more clearly this time. My point is: trend tracking and policy analysis have little relevance for each other. You can discuss education policy, or you can discuss education trends. You can discuss medical policy or you can discuss medical trends. You can discuss immigration policy, or you can discuss immigration trends. And you can discuss redistribution and inequality trends, or you can discuss redistribution and inequality policy. But in all of these cases, and many more, the trend and policy topics have little relevance for each other.
Is not the reason for a policy to change a trend? If there are no trends in need of changing, what use is policy? Sure, arguing about the trend implications of policy might be hard and rife with many of the sociological facts that you often note. But even so, without trend implications, what use is policy? Another way of putting it might be that trend measurement is a way of making policy beliefs pay rent. Of course folks might enforce that rent payment badly, or with biases, or for status gains, but there is still a kernel of purpose behind it.
Another point you mention is that policies are changed rarely and so we should not expect trends to be much affected by policy changes, at least not after controlling for many other trend changers. But there is something normative in your use of "rarely." One might say that while the number of policy changes that have had suitably large trend impacts is low, the trend impact of the few trend-impacting policy changes was extremely high. I am not making such a claim but I could envision folks making such a claim with regard to e.g. civil rights, gender wage gap, measure of freedom for affected minorities, etc. If someone derived utils from the improvement in the trends of these topics, and the degree to which they derived utils was super huge, then even that small amount of trend-changing policy would have been vastly worth all the cost of noisy, unproductive trend measurement and trend debate on a host of many other less possibly efficacious policies.
Another way of reading your post is that people simply have lottery-like preferences when it comes to how they attach utils to policy-induced trend changes. Spend lots of small costs analyzing many trends; always analyze policy in terms of the trends it might impact (even if there is weak evidence that it can actually impact the trend); and for the few lightning-strike policies that really do change trends dramatically this way, celebrate them lots; and for the others, oh well. Just another scratch off ticket that, while it didn't pay, was fun to scratch.
Trends usually are used to persuade outsiders that action should be taken. But the particularly form of action is always what the insiders already favored. Even if the trends later change 180 degrees, it's still the same set.
Theoretical policy discussions are almost completely irrelevant except maybe 1 generation later.
Trend metrics tend to be near the desired outputs. The satisfactory or unsatisfactory nature of desired outputs, and particularly, whether they are, in total, getting better or worse, is a critical element in determining the generally correct level of conservatism vs. radicalism in policy.
We rarely change policies, so most trends aren't caused by policy changes. And even when we do change policies, those are usually only a small contribution to the trends, which have many other causes. I didn't at all say that we cant measure things relevant for policy. I said time trends in such measures aren't usually directly relevant.
This seems offbase. If you are changing policy it should be to make an improvement. If you have no way to measure that improvement, you shouldn't be changing policy.
Mortality and rate of (effective) drug introduction are things we want to control, via policy or other means. Risk aversion, info asymmetry and customer irrationality are also measurable and trend able (even if the trend is expected to be flat). The fact that policy also needs to consider difficult to measure items (such as actual meddling preferences) doesn't mean policy discussions should be isolated from measurable reality.
Is not the reason for a policy to change a trend? If there are no trends in need of changing, what use is policy? Sure, arguing about the trend implications of policy might be hard and rife with many of the sociological facts that you often note. But even so, without trend implications, what use is policy? Another way of putting it might be that trend measurement is a way of making policy beliefs pay rent. Of course folks might enforce that rent payment badly, or with biases, or for status gains, but there is still a kernel of purpose behind it.
Another point you mention is that policies are changed rarely and so we should not expect trends to be much affected by policy changes, at least not after controlling for many other trend changers. But there is something normative in your use of "rarely." One might say that while the number of policy changes that have had suitably large trend impacts is low, the trend impact of the few trend-impacting policy changes was extremely high. I am not making such a claim but I could envision folks making such a claim with regard to e.g. civil rights, gender wage gap, measure of freedom for affected minorities, etc. If someone derived utils from the improvement in the trends of these topics, and the degree to which they derived utils was super huge, then even that small amount of trend-changing policy would have been vastly worth all the cost of noisy, unproductive trend measurement and trend debate on a host of many other less possibly efficacious policies.
Another way of reading your post is that people simply have lottery-like preferences when it comes to how they attach utils to policy-induced trend changes. Spend lots of small costs analyzing many trends; always analyze policy in terms of the trends it might impact (even if there is weak evidence that it can actually impact the trend); and for the few lightning-strike policies that really do change trends dramatically this way, celebrate them lots; and for the others, oh well. Just another scratch off ticket that, while it didn't pay, was fun to scratch.
Trends usually are used to persuade outsiders that action should be taken. But the particularly form of action is always what the insiders already favored. Even if the trends later change 180 degrees, it's still the same set.
Theoretical policy discussions are almost completely irrelevant except maybe 1 generation later.
So when metrics are lower than hoped for, does that say we should have more or less radical policy?
Trend metrics tend to be near the desired outputs. The satisfactory or unsatisfactory nature of desired outputs, and particularly, whether they are, in total, getting better or worse, is a critical element in determining the generally correct level of conservatism vs. radicalism in policy.
We rarely change policies, so most trends aren't caused by policy changes. And even when we do change policies, those are usually only a small contribution to the trends, which have many other causes. I didn't at all say that we cant measure things relevant for policy. I said time trends in such measures aren't usually directly relevant.
This seems offbase. If you are changing policy it should be to make an improvement. If you have no way to measure that improvement, you shouldn't be changing policy.
Mortality and rate of (effective) drug introduction are things we want to control, via policy or other means. Risk aversion, info asymmetry and customer irrationality are also measurable and trend able (even if the trend is expected to be flat). The fact that policy also needs to consider difficult to measure items (such as actual meddling preferences) doesn't mean policy discussions should be isolated from measurable reality.