1. If you cannot be unbiased, at least be consistent. Consistency is often easier to achieve and more important.
2. With the resources you have, there is a limit on how good you can ever get. We call it the Cramér-Rao lower bound.
3. Update your expectations as news come in. The state space model is just a systematic way of doing it.
4. Very often you just cannot be a perfectionist. To address endogeneity is to risk having weak instruments; relaxing assumptions and address them pave ways to inefficiencies - probably when you don't need to. You make trade-offs.
5. All estimates come with a standard error. And even the standard error may be inappropriate and needs to be adjusted.
6. Don't wow at complex models and laugh at simple ones. As things get complicated, the chance that you made inappropriate assumptions and committed errors gets higher.
2. With the resources you have, there is a limit on how good you can ever get. We call it the Cramér-Rao lower bound.
3. Update your expectations as news come in. The state space model is just a systematic way of doing it.
4. Very often you just cannot be a perfectionist. To address endogeneity is to risk having weak instruments; relaxing assumptions and address them pave ways to inefficiencies - probably when you don't need to. You make trade-offs.
5. All estimates come with a standard error. And even the standard error may be inappropriate and needs to be adjusted.
6. Don't wow at complex models and laugh at simple ones. As things get complicated, the chance that you made inappropriate assumptions and committed errors gets higher.
