How Mobile Homes Correlate with Per Capita Income
This Article contains an appendix, which can be found in the PDF linked below.
This study explores the nature of the relationship between the number of state-regulated mobile homes and per capita income, so as to determine whether higher-income parts of Illinois have more mobile homes than would be predicted by a recent BBC News article. It does so by identifying a simple way to determine the direction and strength of any relationship between mobile homes and per capita income, which that article assumes to be negative, if only at the county level in Illinois. The study, specifically, collects and analyzes mobile home data from Illinois and per capita income data from the U.S. Census. After combining these data, then using correlation coefficients, it finds a positive relationship—albeit with different intensities, based on whether the proxy for mobile homes in each county is the number of mobile home spaces or the number of mobile home parks in 2017.
Table of Contents Show
Introduction
Do lower-income parts of the U.S. have more mobile homes than higher-income ones? One might think so, based on a 2013 report from BBC News.
My study carries out some of this follow-up work by creating, and later analyzing, a new State of Illinois dataset. It does so by drawing upon information about the 853 licensed mobile home parks and 53,291 licensed mobile home spaces that were hosted by 94 of Illinois’s 102 counties.
Such a study may yield valuable insights into where mobile homes are located at the county-level and, potentially, into why Illinois residents choose this type of affordable housing over other competing alternatives.
My study, therefore, is justified in using correlation coefficients. By definition, correlation coefficients are a simplified statistical approach that determines the direction and strength of the relationship between variables.
Within this context, the study proceeds in four additional parts (Parts I-III). Part I contains this study’s positive analysis. Part II describes its methodological approach. Part III contains this study’s normative analysis, which builds upon a finding that mobile homes may be positively and relatively weakly correlated with per capita income in Illinois counties. Part V is this study’s conclusion, which summarizes its key findings and recommendations.
I. Positive Analysis
Mobile homes, which are a historically overlooked form of U.S. housing tenure, have become an increasingly popular way for governments to provide affordable housing.
For example, the Illinois General Assembly enacted its original Mobile Home Park Act, the state’s primary means of conveying indirect affordable housing subsidies to certain low-income areas.
Unfortunately, few scholars and reporters bother to examine whether this state-level regulation actually lives up to its promise.
As a result, the existing literature on mobile homes still does not answer a basic question: How does the number of mobile homes correlate with traditional measures of economic development such as per capita income?
This study does so by creating and using a new Illinois dataset, which makes three contributions to the mobile home literature. First, it identifies every licensed mobile home in Illinois and matches each with the per capita income of its host jurisdiction (i.e., county area).
II. Methodology
This study creates and analyzes a new Illinois dataset, in order to determine how the number of state-licensed mobile homes correlate with per capita incomes at the county level.
These two data sources, subsequently, were combined and used to compute group-level averages by county location.
This study later uses these data to make a series of findings based upon its interpretation of correlation coefficients. Correlation coefficients are used for at least three practical reasons. First, this approach provides a way to examine the relationship between the number of licensed mobile homes and per capita incomes, assuming that these two variables are useful proxies.
I note, however, that this study’s findings will not be accurate if it fails to account for selection effects, omitted variables and other methodological issues.
These methodological safeguards, if properly employed in this study, help to overcome a range of theoretical and practical issues.
III. Normative Analysis
To summarize, this study draws on mobile home data and per capita income information from the most recent American Community Survey.
The use of correlation coefficients requires this study to rank each county by number of licensed mobile home spaces and per capita income.
As a result, this study is able to explain the relationship between the number of licensed mobile home spaces and per capita incomes.
The initial results of this analysis are as follows, at least in Fiscal Year 2017. There is a positive and weak-to-moderate relationship between the number of licensed mobile home spaces and per capita income (.35).
This result does not mean that the number of licensed mobile home spaces are always positively and weakly-to-moderately correlated with per capita incomes. Rather, it merely indicates that scholars need to do additional work. One approach may consist of finding out if the number of licensed mobile homes correlate with per capital income when using different proxies.
In employing the first option, this study seeks to further interrogate its sole research question. Specifically, the goal is to further establish whether and how mobile homes may correlate with per capita income by substituting the “licensed mobile homes” proxy variable with the “licensed mobile home parks” proxy variable. The study period (Fiscal Year 2017) and unit of analysis (94 counties with a mobile home and/or mobile home park) are the same for the initial and the secondary analysis (i.e., for mobile homes and for mobile home parks).
Within this context, my follow-up research yields a weaker but similarly-positive result (.21).
These initial and secondary results indicate that higher-income counties may have higher numbers of mobile homes, and higher numbers of mobile home parks, than would be predicted. If these findings prove to be true, then contemporary views of where mobile homes are located may not hold up under scrutiny. Thus, more research will be needed on Illinois and other U.S. states.
Other normative implications of such a finding are equally straightforward. For example, Illinois should determine why some residents of higher-income counties choose to live in mobile homes instead of other types of affordable housing such as limited equity cooperatives.
Conclusion
This study looks at the nature of the relationship between the number of state-regulated mobile homes and per capita income, so as to determine whether higher-income parts of Illinois have more mobile homes than would be predicted by a recent BBC News article. It does so by identifying a simple way to determine the direction and strength of the relationship between mobile homes and per capita income, correlation coefficients, which my study uses to show that this relationship is positive and relatively weak in 2017. However, more research is needed to find out precisely how mobile homes correlate with per capita income.
Copyright © 2020 Randall K. Johnson, Professor of Law and Director of the Public Service Law Center, Mississippi College, School of Law. Special thanks to Dean Patricia A. Bennett, Professor Lisa Bernstein, Professor Christophe Henkel, Professor Angela Kupenda, Jayeeta Kundu, Taimoor Aziz, David Bunts, the participants in the 2019–2020 Legal Scholarship Workshop at the University of Chicago Law School and in the 2019-2020 Property Law Works-In-Progress at the Northeastern University School of Law. DOI: https://doi.org/10.15779/Z38X63B59D.
- See Tom Geoghegan, Why Do So Many Americans Live In Mobile Homes?, BBC News, West Virginia (Sep. 24, 2013), https://www.bbc.com/news/magazine-24135022 [https://perma.cc/A8MB-MDPK]. ↑
- Geoghegan, supra note 1. ↑
- See infra Appendix at Tables 4 and 5; See generally Jill Watts, As Coronavirus Magnifies America’s Housing Crisis, FDR’s New Deal Could Offer a Roadmap Forward, Time Magazine (Apr. 24, 2020), https://time.com/5826392/coronavirus-housing-history/?utm_source=newsletter&utm_medium=email&utm_campaign=history&utm_content=20200501&xid=newsletter-history (The national emergency and consequent economic crisis triggered by COVID-19 has exposed one of America’s greatest needs: adequate and safe housing.”). ↑
- Counties are an appropriate unit of analysis because this local level of government is present in every U.S. state. See, e.g., David Kenney & Barbara L. Brown, Basic Illinois Government: A Systematic Explanation 143-45 (3rd ed. 1993). As such, counties are especially useful for fully examining the relationship between mobile homes and per capita income within and among U.S. states. ↑
- See State of Illinois, Department of Public Health, Response To Freedom Of Information Act Request (Jan. 10, 2019) (“This email is in response to your recent Freedom of Information Act request for location information and number of spaces available for every licensed manufacturing home community.”); See County of Cook, Department of Public Health, Response To Freedom Of Information Act Request (Feb. 27, 2019) (“Please see attached letter in response to your FOIA request.”); See David Lee Matthews, Plan Would Almost Triple Units In City’s Only Mobile-Home Park, Crain’s Chicago Business, (Apr. 9, 2014), https://www.chicagobusiness.com/article/20140409/CRED03/140409750/marc-realty-wants-to-triple-units-at-chicago-s-only-mobile-home-park [https://perma.cc/V8HF-KRXR] (“Marc Realty LLC seeks a zoning change that would allow it to nearly triple the number of units at the City’s only mobile home-park . . . to 747 units, from its current 190, according to its zoning application with the city.”); See United States Census Bureau, B19013 Median Household Income In The Past 12 Months (In 2017 Inflation-Adjusted Dollars) [https://perma.cc/5DE4-ENY4]; See 2017 American Community Survey 5-Year Estimates, American Community Survey (Dec. 6, 2018) [https://perma.cc/7UFJ-YWW5]; See United States Census Bureau, 2017 Data Release: New And Notable, https://www.census.gov/programs-surveys/acs/news/data-releases/2017/release.html. ↑
- Cf. Randall K. Johnson, Uniform Enforcement or Personalized Law? A Preliminary Analysis of Parking Ticket Appeals in Chicago, 93 Ind. L.J. Supp. 34, 43 (2018) (“Although this Article does not try to establish if any observed differences are statistically-significant, which is a valid way of determining how much confidence may be placed in a given research finding, it could serve as a point of departure for future work that does so using regression.”). ↑
- Id. ↑
- Some of my past work with simplified approaches has inspired follow-up research, which uses regression analysis to look at the residential property tax appeals process in the City of Chicago and Cook County as a whole. Compare Randall K. Johnson, Who Wins Residential Property Tax Appeals?, 6 Colum. J. Of Tax L. 209 (2015) (applying percentage analysis to determine who wins residential property tax appeals in Cook County) with Robert Ross, The Impact Of Property Tax Appeals On Vertical Equity In Cook County, IL, U. Chi. Harris Pub. Pol’y (Unpublished Manuscript, 2017), http://apps.chicagotribune.com/news/watchdog/cook-countyproperty-tax-divide/data/harris-study.pdf [https://perma.cc/U8W6-RH7Y] (applying regression analysis to determine who wins residential property tax appeals in Cook County). This follow-up work has garnered attention from local, state, national, and international publications, and substantiated my preliminary research findings about residential property tax appeals in Cook County. See, e.g., Jason Grotto & Sandhya Kambhampati, The Tax Divide: Commercial Breakdown, Chicago Tribune, Dec. 7, 2017, http://apps.chicagotribune.com/news/watchdog/cook-county-property-tax-divide/index.html [https://perma.cc/UL8M-FW2W] (“Owners of residential properties, as a group, also ended up paying more in property taxes than they would have if the assessor’s office had done its work properly. The total amount of property taxes levied in a given year is fixed, so if one group of property owners doesn’t pay its fair share, others have to make up the difference.”). My past work on parking tickets, and parking ticket appeals, has also inspired a host of follow-up research in Chicago that adopts my methodological approach. Compare Johnson, supra note 6 (using percentage analysis to find out how parking tickets and other related matters are distributed by zip code in Chicago) with Woodstock Institute, The Debt Spiral: How Chicago’s Vehicle Ticketing Practices Unfairly Burden Low-Income and Minority Communities (Unpublished Manuscript, Jun. 21, 2018), http://woodstockinst.org/wp-content/uploads/2018/06/The-Debt-Spiral-How-Chicagos-Vehicle-Ticketing-PracticesUnfairly-Burden-Low-Income-and-Minority-Communities-June-2018.pdf [https://perma.cc/28WW-RVPR] (using percentage analysis to examine parking tickets, and other related matters are distributed in Chicago). This other follow-up work has garnered attention from local, state, national, and international publications, and substantiated my preliminary research findings about parking tickets, parking ticket appeals, and win rates on appeal in Chicago. See, e.g., Melissa Sanchez & Sandhya Kambhampati, How Chicago Ticket Debt Sends Black Motorists into Bankruptcy, ProPublica Illinois (Feb. 27, 2018), https://features.propublica.org/driven-into-debt/Chicago-ticket-debt-bankruptcy/ [https://perma.cc/G8A8-TJ8N] (describing how excessive ticketing has especially severe consequences for disadvantaged groups in the State of Illinois). ↑
- Correlation coefficients are “a statistical method of quantifying the association . . . between two variables.” Marcin Kozak, Wojtek Krzanowski & Malgorzata Tartanus, Use of Correlation Coefficient In Agricultural Sciences: Problems, Pitfalls And How To Deal With Them, 84 Annals Brazilian Acad. Sci. 1147 (2012). This study uses Microsoft Excel’s correlation function (CORREL) to create correlation coefficients that help to identify the nature of any relationship between mobile homes and per capita income in Illinois during Fiscal Year 2017. See, e.g., CORREL Function, Microsoft (last visited May 2, 2020) https://support.office.com/en-us/article/correl-function-995dcef7-0c0a-4bed-a3fb-239d7b68ca92 [https://perma.cc/2ZQE-YJ3N]. ↑
- Correlation coefficients also are used for three additional reasons. First, this simplified statistical approach is widely viewed as a valid way to establish the direction and strength of a relationship. Second, correlations are also easy to verify using Microsoft Excel. Lastly, this simplified statistical approach may lay a solid foundation for future research that looks at mobile homes and per capita income. Cf. Randall K. Johnson, How Tax Increment Financing (TIF) Districts Correlate with Taxable Properties, 34 N. Ill. U. L. Rev. 39, 41 n.19 (2013). ↑
- See Jeff Andrews, Can Manufactured Housing Ease America’s Affordable Housing Crisis?, Curbed (Mar. 2, 2018), https://www.curbed.com/2018/3/2/17058882/mobile-manufactured-homes-affordable-housing-crisis [https://perma.cc/3XXT-4GS3] (“[Fannie Mae] . . . will purchase around 30,000 manufactured housing mortgage loans over [the three years between December 2016 and 2019, it] . . . will also develop a pilot program for buying chattel loans and for supporting the financing of manufactured housing communities, whether owned by governments, nonprofits, or residents.”). ↑
- See, e.g., Will Ferguson, Los Angeles County Supervisors OK “County Mobile Home Program” To Put Affordable Homes Within Reach Of The Homeless, ManufacturedHomes.com (May 17, 2019), https://www.manufacturedhomes.com/blog/los-angeles-county-mobile-home-program/ [https://perma.cc/9RRD-F8EW] (“The Los Angeles County Board of Supervisors has recently unanimously approved a motion to pursue a plan, calling it the County Mobile Home Program, that would utilize manufactured homes to address the affordable housing crisis.”). ↑
- See, e.g., Geoghegan, supra note 1. ↑
- See generally Manufactured & Modular Homes/Mobile Structures, Illinois Department of Public Health (IDPH) (2001), http://dph.illinois.gov/topics-services/environmental-health-protection/manufactured-modular-homes-mobile-structures [https://perma.cc/832J-QRPZ]. ↑
- See generally 210 ILCS 115; Public Act 77-1472. ↑
- See generally 77 Ill. Admin. Code 860.400, Required Documents (requiring owners of mobile home communities to give new owners, and new renters, the following information: “a copy of the manufactured home community rules,” “a copy of the . . . publication ‘Living in a Manufactured Home Community,’” “[access to] . . . a copy of the Mobile Home Park Act and the Manufactured Home Community Code” and “the name address, and telephone number of the manufactured home community manager whom residents are to notify of a problem within the manufactured home community.”). ↑
- See IDPH, supra note 14 (“To ensure quality living conditions for people who reside in manufactured home communities, IDPH licenses all parks with 5 or more sites (except those located in home rule units). Staff inspects each park annually for license renewal, at which time they check the water supply sewage disposal system, electrical system, lighting, road conditions, spacing of homes and garbage disposal.”). ↑
- See, e.g., David Ray Papke, Keeping The Underclass In Its Place: Zoning, The Poor, And Residential Segregation, 41 Urb. Law. 787 (2009). ↑
- See, e.g., Esther Sullivan, Dignity Takings and “Trailer Trash”: The Case of Mobile Home Park Mass Evictions, 92 Chi-Kent. L. Rev. 937 (2017). ↑
- See, e.g., Anika Singh Lemar, The Role of States in Liberalizing Land Use Regulations, 97 N. C. L. Rev. 293 (2017). ↑
- See, e.g., Soham Dhesi, Protecting Mobile Homes as Affordable Housing, UCLA. L. Rev. Seminar (2018), https://www.uclalawreview.org/protecting-mobile-homes-as-affordable-housing [https://perma.cc/CAW6-K4M7]. ↑
- See infra Appendix at Tables 4 and 5. ↑
- For example, in keeping with a passage from a popular statistics textbook, “[i]f the two variables are associated, we will reduce our errors when our predictions about one of the variables are based on the knowledge of the other.” Joseph F. Healey, Statistics: A Tool For Social Research 341 (6th ed. 2002). ↑
- See infra Appendix at Table 1. ↑
- See infra Appendix at Tables 4 and 5. ↑
- Id. ↑
- Id. ↑
- See infra Appendix at Tables 4 and 5. ↑
- See State of Illinois, supra note 5; See County of Cook, supra note 5; See Matthews, supra note 5. ↑
- See U.S. Census, supra note 5. ↑
- This validity testing included “spot testing” for each data source. ↑
- See infra Appendix at Table 3. ↑
- Id. ↑
- See James Chen, Normal Distribution, Investopedia (May 7, 2019) (“Normal distribution, also known as the Gaussian distribution, is a probability distribution that is symmetric about the mean, showing that data near the mean are more frequent in occurrence than data far from the mean.”). ↑
- Cf. Johnson, supra note 10. ↑
- Kenney, supra note 4. ↑
- See Geoghegan, supra note 1. ↑
- See, e.g., John Antonakis, Samuel Bendahan, Philippe Jacquart & Rafael Lalive, On Making Causal Claims: A Review and Recommendations, 21 Leadership Q. 1086 (2010). ↑
- Cf. Johnson, supra note 10. ↑
- See infra Appendix at Tables 2 and 3. ↑
- Id. ↑
- Cf. Johnson, supra note 10. ↑
- Id. ↑
- See infra Appendix at Tables 4 and 5. ↑
- Id. ↑
- See State of Illinois, supra note 5; See County of Cook, supra note 5; See Matthews, supra note 5; See U.S. Census, supra note 5. ↑
- See infra Appendix at Tables 4 and 5. ↑
- See infra Appendix at Table 4. ↑
- Cf. Johnson, supra note 10. ↑
- Id. ↑
- See infra Appendix at Table 4. ↑
- See generally Deborah J. Rumsey, How to Interpret a Correlation Coefficient r, Statistics For Dummies, 2nd Edition (June 2016), https://www.dummies.com/education/math/statistics/how-to-interpret-a-correlation-coefficient-r/ [https://perma.cc/6V83-NLHH]; See U.S. Census Bureau, supra note 5. Fiscal Year 2017 is our focus because it is the last year that per capita income data is provided by the U.S. Census; See generally Watts, supra note 3. ↑
- See infra Appendix at Table 4. ↑
- Statistical significance may be determined through the use of probability values, which are computed at the 0.05 level and at the 0.10 level. See, e.g., Daniel Soper, Statistics Calculators, https://www.danielsoper.com/statcalc/related.aspx?id=44 [https://perma.cc/2YSD-2H4E] (last visited May 2, 2020). Probability values cannot exceed 0.05 for one-tailed probability values or 0.10 for two-tailed probability values, in order to be considered statistically significant. The initial result had probability values of 0.000272 (one-tailed test) and 0.000544 (two-tailed test), so it can be considered statistically significant. ↑
- Parameters for probability value computations draw on correlation values (+0.35) and sample sizes (94). ↑
- Cf. Geoghegan, supra note 1. ↑
- Id. ↑
- Id. ↑
- Id. ↑
- Id. ↑
- See infra Appendix at Table 5. ↑
- The initial result had probability values of 0.021102 (one-tailed test) and 0.042205 (two-tailed test). Thus, only modest conclusions may be drawn about the relationship between these variables. These conclusions are that there is some evidence that the null hypothesis in this study cannot be accepted (i.e., that there is some relationship between the number of mobile homes and per capita income). ↑
- Parameters for probability value computations draw on correlation values (+0.21) and sample sizes (94). ↑
- Compare infra Appendix at Tables 4 and 5 with Geoghegan, supra note 1. ↑
- See generally Julie D. Lawton, Limited Equity Cooperatives: The Non-Economic Value of Homeownership, 43 Wash. U. J. L. & Pol’y 187, 207 (2014) (“A limited equity cooperative restricts the amount of equity appreciation, or the resale price above the owner’s purchase price, that the cooperative owners may obtain upon resale of the cooperative share . . . The over-arching intended benefit of an LEC is to preserve the property’s affordability by ensuring the cooperative share price does not increase to a level unaffordable to future low- and moderate-income buyers [that want to buy in].”). ↑
- See generally Alana Semuels, The Case for Trailer Parks, The Atlantic, (Oct. 24, 2014), https://www.theatlantic.com/business/archive/2014/10/the-case-for-trailer-parks/381808 [https://perma.cc/9JSL-44VH]. ↑
- Cf. Frank Mulholland, Property Tax Code on Mobile and Manufactured Homes Changed, Shelbyville Daily Union, (Jan. 12, 2011), https://www.shelbyvilledailyunion.com/news/property-tax-code-on-mobile-and-manufactured-homes-changed/article_a9ddc740-af68-531c-aa1b-10f1a14a9aae.html [https://perma.cc/9KM2-KH2Q] (“Public Act 96-1477 . . . went into effect on January 1, 2011 and changes the tax code on mobile and manufactured homes . . . [because] . . . people in [conventional residential properties and mobile homes] . . . use the same services and yet . . . mobile home and manufactured homes paid much less taxes.”). ↑
- See Wendy Plotkin, Rent Control in Chicago After World War II: Politics, People & Controversy (1998), http://wbhsi.net/~wendyplotkin/Prologue.pdf [https://perma.cc/S3FE-2396]. ↑