This hypothesis has its own paper, by Robert Wassner and David Edwards. Here is the bottom line:
The more land use decisions in a state are driven by fiscal considerations, the more likely may be the degree of sprawl observed in the state’s urban areas. The reason is that outlying local governments in an urban area, with a greater likelihood of possessing undeveloped land, are more likely to use revenue considerations in choices related to land use. As discussed next, the extent that outlying local governments in a state’s urban areas weigh fiscal considerations in land use decisions is expected to depend on the relative reliance by local governments in the state on the different possible forms of local revenue. That is, greater statewide reliance on a form of local revenue that rises and falls with differences in land use choices can result in forms of land use decisions in an urban area that generate greater sprawl. The theoretical connection between sales tax reliance by local governments in a state and greater urban decentralization centers on the ability of local government officials to influence the location of retail activity in an urban area. The amount of overall retail activity in an urban area is determined by factors such as population, income, age distribution, etc. in the region and is unlikely to change due to local government influences. But local land use decisions can shape the distribution of overall retail development in an urban area. Local this land for retail activity if local greater retail generates for them. This surplus comes in the forms of (1) non-residents paying local sales taxes and (2) the lower alternate land uses.
The paper offers a good bit of evidence, though the authors are very careful to claim this is not the only or even the major factor behind sprawl. Lots of little local governments, of course, also enhance educational competition and the quality of the local schools, which further encourages sprawl.