01. Effective simplification is the first principle of effective modeling and analysis.
02. All learning involves modeling: there is no learning without generalization.
03. Underlying any good prescriptive model is a good descriptive model.
04. Before you start work on an empirical problem, think about what you would do if you had all the data in the world.
05. To get your thinking straight: 1) Start any model or analysis with the simplest case; 2) Consider behavior in the extreme cases.
06. The more complex the practical problem, the greater the value of analytic tools.
07. Do not take refuge in complexity.
08. Most analysis too little considers uncertainty. Uncertainty goes both ways.
09. A well-behaved model has well-behaved solutions (and conversely).
10. If an optimum is wide and flat, your precise choice doesn't matter that much.
11. When people adjust the discount rate in a model, it is generally to avoid explicitly addressing other assumptions.
12. Be aware of regression toward the mean--a widespread phenomenon whose consequences are frequently mis-attributed to other factors.
13. When making subjective estimates of any unknown quantity, broaden your range. Most people are much too confident in their judgements.
14. Watch out for situations in which the coin lands on its edge (i.e., consider the hypothesis that you have fundamentally misunderstood what is going on.)
15. Beware if the "Aha, the externality!" argument: some externalities matter, some don't. Same with uncertainties. Ritualistic invocation of uncertainties (often by the right) or externalities (often by the left) is a debating gambit. For that matter, same with adverse selection (left) and moral hazard (right) in insurance debates.
16. Inflation is the great weapon of the administrator.
17. When negotiating for a new job, push for a higher salary than you think you should. (An instance of the general rule: push to claim more when the pie is expanding.)
18. When the resolution of a political issue is clear and you cannot influence another's view, look for comparative advantage not absolute advantage.
19. Watch out for the principal-agent problem: it crops up nearly every day, and can be dealt with through a variety of creative methods.
20. Government should be risk-neutral if and only if: a) the risk is small compared to the total budget; b) resources are fungible across budget categories; c) government is the actual party at risk, not acting as an agent for the risk-averse.
21. Evaluation of outcomes in comparative rather than absolute terms is a) omnipresent; b) nearly always destructive, and c) to a significant extent correctible. Indeed, possibilities for preference adaptation are far stronger and more widespread than we imagine. "The essence of a satisfied life, grasshopper, lies less in getting what you want than in learning to want what you have."
22. Every choice in life is a tradeoff.
3 comments:
Good post and this post helped me alot in my college assignement. Thank you seeking your information.
what was the assignment?
Good post and this fill someone in on helped me alot in my college assignement. Gratefulness you as your information.
Post a Comment