Disaggregated Spending and the Productivity Bias Hypothesis