Deontological vs. consequentialist ethics

Deontological = rule-based. Consequentialist = consequence-oriented. Or, deontological = focused on the means, whatever the end; consequentialist = focused on the end, whatever the means.

An absurd example of deontological ethics is making it a rule never to lie and not lying to a Nazi asking you whether you're hiding any Jews. The result (people killed) is intuitively recognized as horrific, but you have adhered to the rule you committed yourself to. The point of this ad absurdum argument is that not caring about the consequences, as long as you adhere to the rules, is sometimes ridiculous. Or, in Latin, fiat justitia ruat caelum (“let the justice be done though heavens may fall").

One can also think, of course, of consequentialist ad absurdums. For instance, one achieves some worthy goal through lying or bullying others: you invest your client's money into something against his wishes, and that earns profit for him. Yes, you have achieved a positive result, but you have done so through an immoral act. Or, for instance, if you see some guy who owns some land that he doesn't use or uses badly. You confiscate the land from him, pay him very well for it, and put the land to a much better use, enriching the community. The guy was paid more than he would ever earn, and everyone benefited so much, but you have violated his property rights.

So, it would seem that it's important to take into account both consequentialist and deontological approaches.

One way to reconcile them is to treat deontological approach as producing long-term consequences with certain high but not 100% probability. Deontological rules can be framed in terms of slippery slopes. On the other hand, consequentialist approach worries about short-term consequences that have very high, near-100% probabilities (such as someone drowning). Usually it's more responsible to worry about long-term over short-term consequences, unless short-term consequences are totally terrible (and possibly unforeseen or an exception/emergency). So, this is the way to frame the balance between them in consequentialist terms (long-term vs. short-term consequences). Deontological framing would be to make a rule: e.g., "generally speaking, we don't lie, unless lying can save someone's life." Or: "generally speaking, we don't break into people's houses, unless we need to save someone's life from a fire." Etc.

Things get turned on their heads when allowance to break the long-term rules for short-term emergencies gets confused for not caring about the long-term rules. It's like a cousin who asked to sleep on the couch for the night and stayed for a month. I.e., the time frame of both the rule and the consequences gets confused. For instance, "we can confiscate someone's house to pay for someone else's cancer treatment". (Note, also, how the usual criticism of libertarian arguments involves life-boat situations.)