As an minor addition to the already accepted answer, it may be said that since the late 1860s or so 'it has always been that way'. The oldest 'rulebook' for chess problem composition that I know of is from 1878, and its second rule is:

2. Capture of a man on the first move of white is not regarded as good unless intricate, beautiful, and numerous variations result from it.

(See http://www.anders.thulin.name/posts/the-art-of-problem-composition/ for the full text: the rules are at the end. It is quite dated.)

Earlier, in the 1850s, many 'good' problems were of the capture-early-and-often type, and were probably designed to show off a solver as someone who could find a forced mate in 18 moves or so. As taste changed towards more modern style (in the 1860s), this became old-fashioned and crude.

As an minor addition to the already accepted answer, it may be said that since the late 1860s or so 'it has always been that way'. The oldest 'rulebook' for chess problem composition that I know of is from 1878, and its second rule is:

2. Capture of a man on the first move of white is not regarded as good unless intricate, beautiful, and numerous variations result from it.

(See http://www.anders.thulin.name/posts/the-art-of-problem-composition/ for the full text: the rules are at the end. It is quite dated.)

Earlier, in the 1850s, many 'good' problems were of the capture-early-and-often type, and were probably designed to show off a solver as someone who could find a forced mate in 18 moves or so. As taste changed towards more modern style (starting in the 1860s), this became old-fashioned and crude.