Why the Advantage Fades—and What Else Could Explain It

The falling US local bias of entrepreneurship is a symptom, not a diagnosis. The question is whether the decline reflects a changed mix of businesses, more portable finance, more individual mobility, weaker attachment to birthplace, or some combination. Separating those possibilities matters because each one implies a different meaning for local founders and for the regions in which they operate.

The symptom is not the cause

The observed result is clear enough: self-employed people became less unusually likely than wage workers to live in their state of birth between 1970 and 2018. The US paper also shows a major shift away from high-startup-capital self-employment, especially among local entrepreneurs. These are two related facts, but a relationship between them is not the same as a completed causal explanation.

The same fall in LBE can be generated by different changes. Suppose local entrepreneurs remain common in a particular kind of business, but that kind becomes a smaller share of all self-employment. The aggregate local share may fall even though the local advantage within that business is unchanged. That is a composition story. Alternatively, suppose the business mix stays similar but movers can now borrow from distant lenders on terms closer to those available to locals. The finance advantage shrinks, and the LBE falls without any change in the founder's desire to stay home. That is an institutional story.

A third possibility is behavioral. People may move more often, or feel less attached to their birthplaces, so both entrepreneurs and wage workers become more mobile. If entrepreneurs become more mobile than workers, LBE falls; if both groups become more mobile by the same amount, the relative gap may not change. The LBE statistic cannot distinguish those cases without information about each group's absolute location pattern.

A fourth possibility concerns selection. The people entering self-employment may have changed. If the population contains more individuals whose skills, wealth, customers, or business models are portable, the local bias could decline even if local finance remains important for other founders. Changes in the classification of self-employment or in the mix of occupations can have a similar effect. The robustness of the US result across incorporated and unincorporated businesses and full-time status is useful, but it does not make the population composition fixed over five decades.

flowchart TD A["Observed symptom: falling relative LBE"] --> B{"Which condition changed?"} B --> C["Less high-capital self-employment"] B --> D["Less locally segmented finance"] B --> E["More individual mobility"] B --> F["Home attachment plus weak local jobs"] B --> G["Learning and sector specialization"] B --> H["Selection or measurement change"] C --> C1["Fits the documented US composition shift"] D --> D1["Fits the mechanism but is not directly measured in the US trend"] E --> E1["Possible but LBE alone does not measure absolute moves"] F --> F1["Italian regional direction weighs against it as the complete cause"] G --> G1["Italian density and specialization evidence is weak"] H --> H1["Requires comparable population and business measures"] C1 --> I["Provisional diagnosis, not a full decomposition"] D1 --> I E1 --> I F1 --> I G1 --> I H1 --> I