Digital photography is like digital audio or anything digital. It has NO tolerance to errors if you exceed the recording range (and usually you notice it in the high limit and not in the low). It happends in audio when you clip the audio peaks (awful high frequency distortions) and in photography you ended with unnaturally-blown highlights.
The analog world (audio tapes, film negatives or slides) usually tend to smooth the high peaks, compressing the "signal" in a more pleasing way. Digital has no such smooth efect, it only clips it to the largest possible value, so you must to keep all the signal in valid ranges (usually underexposing), and then, if you want, modify the curve to increase the lower levels and compress the high ones, to obtain a similar image. This has advantages and disavantages. The good news is: you are actually INCREASING the effective camera sensitivity when you underexpose, but the bad ones are you are also increasing noise, during the level increase in dark tones at posprocessing. Also, you must shoot in 12 bits (usually that means RAW or TIFF) to get some extra values to use during the level processing.
So, digital photography has the advantage of inmediate review (or in digicams, preview), but this review is also more necesary than in film photography, because is more easy to destroy a picture in digital than in film.