The effective reproduction number in Japan is regressed on real household expenditures per household for eating out, traveling, and clothing and footwear, as well as a measure of mobility in public transportation, using publicly available daily nationwide data from February 2020 to February 2021. The effects of absolute humidity, the declaration of states of emergency, and the year-end and new-year holiday period are controlled through dummy variables. The lagged effect due to incubation is also taken into account. The regression result indicates that household expenditures for cafe and bar had larger effects on an increase in the effective reproduction number per value of spending than the other household expenditures in the explanatory variables. The posterior means of simulations of the estimated regression model, however, imply that if the government aims to keep the effective reproduction number below one on average only by restricting household expenditures for cafe and bar, then it is likely that these expenditure items must be cut by more than 80% of the 2019 levels, or that it is impossible to achive the goal, depending on the degree of self-restraint on traveling.