What to do about grade inflation?

Friend crystalghetto brings up an interesting essay and his personal experience with grade inflation. His post got me interested in the topic, and it turns out there's quite a bit of information out there on the subject.

• This Washington Post contributor, a professor, notes that grades like D and F "went by the wayside in the Vietnam era, when flunking out meant becoming eligible for the draft."
• His interest in grade inflation motivated him to collect data (sample size 22) on average grades at several postsecondary institutions, showing that average grades have risen as much as 0.20 on the 4-point GPA scale.
  
• On the same site, note that the average GPA at a private college is three-tenths of a point higher at private colleges. Do you pay for better grades by going private? Or are private college students that much brighter, on average?

Going back to 2001, there's a truly excellent essay on grade inflation by Harvard government professor Harvey C. Mansfield. In the essay, he addresses many of the excuses for grade inflation, from meeting student expectations, to lower faculty morale, to the theory of a smarter student.

[Grade inflation] signifies that professors care less about their teaching. Anyone who cares a lot about something -- for example, a baseball fan -- is very critical in making judgments about it. Far from the opposite of caring, being critical is the very consequence of caring.

The problem, Mansfield says, is that grade inflation makes it difficult to distinguish truly excellent work.

In a healthy university, it would not be necessary to say what is wrong with grade inflation. But once the evil becomes routine, people can no longer see it for what it is...Grade inflation compresses all grades at the top, making it difficult to discriminate the best from the very good, the very good from the good, the good from the mediocre. (emphasis mine)

In addition, grade inflation represents a fear of challenging brighter students to reach even greater achievement:

Some say Harvard students are better these days and deserve higher grades. But if they are in some measures better, the proper response is to raise our standards and demand more of our students. Cars are better-made now than they used to be. So when buying a car, would you be satisfied with one that was as good as they used to be? (emphasis mine)

...When bright students take a step up and find themselves with other bright students, they should face a new, higher standard of excellence.

It also seems that grade inflation is more evident in courses that depend on less objective measures of achievement.

Grades in humanities courses are notably higher than those in the social sciences, and both are higher than grades in the natural sciences.

Ultimately, Mansfield has no solution, but he does finger one particular cause: a culture of self-esteem that puts a priority on self-confidence over honest assessment. He also says that the passion of professors for affirmative action in the late 1960s led to grade inflation for minority students that eventually spread across all ethnic groups. And once in this cycle of up-grading, it's hard to go back.

In fact, Mansfield himself continues to give higher grades than he used to, although his attempt to be honest is to offer students two grades: one for the registrar (inflated) and one given in private. The latter is almost always lower and is intended to more readily reflect the level of achievement.

Clearly, grade inflation creates problems. But does anyone have a good way to solve it?