Naturally, economists wondered about this productivity paradox. It
seemed inconceivable that computers could be as unimportant as statistical
analysis suggested. Several possible explanations were posited [2]. First,
computers likely have their biggest impact within the service sector, but we
measure output in services less well than we measure output in agriculture
or manufacturing. A second and related problem is that computers may have
their biggest effect in increasing the quality of output rather than the amount
of output. Economists struggle to capture the value of quality improvements.
If these are under-estimated then the causes of quality improvement will be
under-appreciated