I think in part the fact that rejected Starbucks openings are factored in as a control that this is partially counted into the measurement. Additionally, it appears that startup density per year increases with time since the space opens, which indicates that it is partially a causal, or at least time-correlated factor.
I think in part the fact that rejected Starbucks openings are factored in as a control that this is partially counted into the measurement. Additionally, it appears that startup density per year increases with time since the space opens, which indicates that it is partially a causal, or at least time-correlated factor.
A logical counter explanation, to your point, would be that geographic locations sponsoring the opening of a Starbucks may also be making a general investment in local businesses, and as a result we see startup density in general increase.
“ A logical counter explanation, to your point, would be that geographic locations sponsoring the opening of a Starbucks may also be making a general investment in local businesses, and as a result we see startup density in general increase.”
Yes, and you also need to explore whether rejected Starbucks openings are really random and whether those neighborhoods are in fact good controls.
I think in part the fact that rejected Starbucks openings are factored in as a control that this is partially counted into the measurement. Additionally, it appears that startup density per year increases with time since the space opens, which indicates that it is partially a causal, or at least time-correlated factor.
A logical counter explanation, to your point, would be that geographic locations sponsoring the opening of a Starbucks may also be making a general investment in local businesses, and as a result we see startup density in general increase.
“ A logical counter explanation, to your point, would be that geographic locations sponsoring the opening of a Starbucks may also be making a general investment in local businesses, and as a result we see startup density in general increase.”
Yes, and you also need to explore whether rejected Starbucks openings are really random and whether those neighborhoods are in fact good controls.