Skip to main content

A spatial model of growth relationships and Latino-owned business

Abstract

The expansion of ethnic minorities evokes policy debate about their impact on the local economy, driving a need to measure their effects. This article introduces a spatial econometrics approach to Deller et al.’s expansion of the Carlino-Mills growth model. We employ the confidential US Census data to investigate drivers of local economic performance with emphasis on the role of Latino-owned businesses (LOB) on convergence. The model also includes a number of controls. The model produces direct, indirect, and total impact estimates, and expected values for the non-LOB controls. The estimated total impact of LOB employment on county-level average annual growth rates is significant and positive, but a rurality interaction carries the opposite sign, such that the total impact in rural areas is negative. The negative effect in rural areas is due to negative spatial spillovers captured by the model. The spatial Durbin error model empirical results indicate that although LOB employment interacted with rurality significantly impacts county-level growth rates of population, employment, and income, they do not change the equilibrium relationship between these variables captured by the speed of convergence.

This is a preview of subscription content, access via your institution.

Notes

  1. Although Latinos in the USA are the population of interest, some data collection agencies gather data using the term “Hispanic,” and where that is the case we maintain their wording. Given the population of interest is Latino and that most individuals in the USA who are Hispanic are also Latino, this article assumes results for Hispanic entrepreneurs in the USA would be similar to results for Latinos.

  2. Autor and Dorn (2013) find that automation of routine tasks plays a leading role in rising employment and wage polarization.

  3. We are grateful to the anonymous referee who provided these comments.

  4. Log transforming the initial periods also helps with the spatial lags, since the regressions relate changes in variables (population, employment, and income) that have very different scales. For example, the spatial lags constructed from the log-transformed variables \(W\log E_{0} ,\;W\log I_{0}\) in the population growth equation will be much better scaled in cases where we have a large population urban county with small rural counties as neighbors. We are grateful to the anonymous referee for these comments.

  5. As the SBO is a stratified sample of firms based on racial/ethnic status of the owner, the model uses the SBO weights to make sure LOB employment ratio at the county level is representative of the firm population.

  6. We use the USDA rural–urban classification codes (RUCC), which define “rural” as codes 7, 8, and 9.

  7. To find the total impact estimate, one must simply sum the direct and indirect coefficient estimates.

  8. Note that the US Census Bureau suppresses the exact number of observations. Though we use the continental US counties, we must round the number of observations to \(n = \sim3000\).

References

Download references

Acknowledgements

Thanks to Anil Rupasingha, Myriam Quispe-Agnoli, Julie L. Hotchkiss, Melissa Banzhaf, Margaret C. Leventsin, J. Clint Carter, Mark Fossett, Bethany DeSalvo, the Interuniversity Consortium for Political and Social Research, the MRDC, the TXRDC Consortium, numerous conference participants, and an anonymous referee. Support for this research from the North Central Regional Center for Rural Development and the Department of Agricultural, Food, and Resource Economics at Michigan State University is also gratefully acknowledged. The project was supported by the Agricultural and Food Research Initiative Competitive Program of the USDA National Institute of Food and Agriculture (NIFA), Award Number 2017-67023-26242, and by Hatch project 1014691. Michigan State University Institutional Review Board reviewed this work and declared it to have exempt status. Any opinions and conclusions expressed herein are those of the author and do not necessarily represent the views of the US Census Bureau. All results have been reviewed to ensure that no confidential information is disclosed.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Craig Wesley Carpenter.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Carpenter, C.W., Loveridge, S. A spatial model of growth relationships and Latino-owned business. Ann Reg Sci 63, 541–557 (2019). https://doi.org/10.1007/s00168-019-00942-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00168-019-00942-x

JEL classification

  • R1
  • R11
  • R12
  • L26