“As commander in chief, I’m responsible,” David Frye once said, doing his impression of Richard Nixon during the height of the Watergate scandal. “But I’m not to blame. Let me explain the difference: People who are to blame go to jail. People who are responsible do not.”

The Nixon administration sliced the bologna pretty thinly back then. No thinner, though, than many of us do today when we walk the fine lines that separate responsibility, accountability, and blame.

Let me explain the difference: When you hold employees accountable, you help them succeed. When you assign (or allocate) blame, you ensure everyone’s failure.

Start with the terms “responsible” and “accountable.” Never mind the dictionary; there are two similar but distinct ideas to keep track of when you delegate; we can use these two words to make the difference perfectly clear.

First, let’s talk about responsibility. If you’ve been assigned a responsibility you’re supposed to figure out what has to get done and go do it. If you’re a responsible soul, you do so. Key to responsibility is matching it with authority: Giving someone responsibility for the summer picnic, for example, but not letting them choose what brand of hot dog you buy just isn’t kosher.

Then there’s accountability, which we’ll attach to a related but different idea. If you’re accountable, your boss is aware of how you are doing. If a problem arises, your boss is aware of that, too, and wants to know how it arose and what you’re going to do about it. People are (or are not) responsible; you hold them accountable. The distinction may seem like a fine line, but it’s really a broad chasm.

People are responsible for results, not problems. That means they’re responsible for solving problems within their area of authority, not for the problems themselves. It also means you hold them accountable for results, not for problems that arise.

The exercise of determining who caused a problem doesn’t assign responsibility, nor does it hold people accountable. It simply assigns blame, determining who gets punished. This “heads will roll!” mentality is always a bad idea, because if heads may roll, heads will be kept down, which means problems will go unsolved.

Asking who caused a problem is, in fact, the wrong question because most problems are caused not by a “who” but by insufficient process design, resource constraints, earthquakes, tornadoes … and yes, occasionally by incompetence, laziness, poor judgment, or just a bad guess. Don’t ask who caused a problem, ask what caused it. If you find the problem arose from the actions of an employee or team, punishment is appropriate only on rare occasions. What matters is figuring out how to avoid repetition.

Because of how a problem arose you may decide an employee is unsuited to a job, or even unsuited to employment with your company. If that’s the right answer, make the right decision – not to punish the employee, but to help both the employee and your company succeed, neither of which will happen unless the employee changes roles or employers.

Punishment is only appropriate in the case of malfeasance.

So … If you’re responsible, you make sure things get done. If you’re accountable, your boss is in the loop. If you’re to blame, your boss asked who was at fault, decided it was you, and made chopped liver out of you for it.

Another bit o’ semantics: Many people call the combination of responsibility and authority “ownership.” The upside: People who own things understand that they’re both responsible for them and have authority over them. The downside: They guard them as part of their territory. Personally, I prefer the term “stewardship,” which means you’re both responsible and have authority, but over somebody else’s property.

Responsibility, authority, ownership, stewardship, accountability, blame … what matters isn’t how thinly you slice these words. You may define them beautifully, or do the wurst job ever.

What matters is, as the Japanese would say, whether you fix the problem or just affix blame.

“Practice makes perfect,” we learned as kids.

“Do things right the first time,” we learned in Total Quality Management.

Doing things right the first time is for manufacturing. It means quality happens during assembly, not during inspection, when you’d have to do it over. Outside of the factory, you have to make mistakes first.

Practice is as useful for leadership techniques as for methodologies like Use Case Analysis, or really difficult skills like welding.

Perhaps that’s why so many people turn into jerks when they’re promoted to leadership positions.

Let me explain. An important part of leadership is the ability to deliver bad news — small, but important nonetheless. Unfortunately, it’s as unnatural as the embouchure for the oboe (trust me) so we usually flub it the first time out.

What happens next? Some leaders avoid repetition. They either leave management or just skate by difficult situations. Others overreact the other way, giving lots of people bad news, whether or not there’s bad news to give. It’s that practice thing we were talking about earlier.

Most bad news doesn’t have to be given. Two open-ended questions do a better job: “How do you think you performed?” and “How could you do better next time?” Usually, you’ll find that the employee already recognizes the problem and, with encouragement, can figure out what would have worked better. Just empathize, so the employee knows this isn’t a catastrophe, applaud the employee’s professionalism in recognizing a substandard result, and reiterate the employee’s value to you and the company.

This doesn’t always work, though, and when it doesn’t, be prepared. Here are some tips and techniques on how to give bad news (but don’t expect to be excellent at this until you’ve been through it a few times):

Tip No. 1: Make sure this isn’t just a personality conflict. Employees don’t have to like you but do have to respect you. You owe them the same attitude.

Tip No. 2: Scale your delivery. If it’s trivial, don’t bother. If it’s not that big a deal, say so and explain why you’re taking the trouble to discuss it at all. If it’s a significant issue, make that clear, too.

Tip No. 3: Be sure of your facts. It’s demotivating for the employee, and embarrassing for you, if you find out the events you’re reacting to never really happened, or happened very differently from what you’ve surmised.

Tip #4: Plan the conversation. Don’t trust your natural eloquence. Make sure you know what you’re going to say and how you’re going to say it.

Tip #5: Be resolute. Yes, resolute, corny as it may sound. And keep the conversation focused. For example, you may have an employee who, through passive resistance, is subverting a goal you’ve set. If the employee questions your goal, say, “We can schedule a meeting for that conversation if you’d like. Right now, that is the goal and we’re discussing your decision to not achieve it.”

Tip #6: Listen, but don’t let the conversation drag out. Your goal is to make sure the employee understands you – liking you or the message probably isn’t feasible. Paraphrase the employee’s position and confirm you got it right. Then ask the employee to paraphrase your message. When you’re sure the employee understands you, end the meeting.

Tip #7: Retain your composure. Anger is unprofessional. Exhibit the behavior you expect. If the employee becomes angry, don’t respond in kind. If it continues, say, “We’ll continue this conversation when you’ve regained your composure.”

Tip #8: Bad news is private. Be careful, though. While it’s never appropriate to chew someone out in public, many experts suggest, and some HR departments insist on, a witness to any difficult performance-related discussion. Since sexual harassment itself, unwarranted accusations of harassment, and wrongful termination lawsuits of all kinds are real, private conversations can be risky, and not just when you and your employee are of different genders. Consult your HR department for advice and to determine your company policy.

One last thought: While it’s important to be able to give bad news effectively, it isn’t important to like it. If you find that you enjoy the experience … you’ve stepped across a dangerous line.