Skip to main content
SpringerLink
Log in

From universal service to universal connectivity

  • Original Article
  • Published:
Journal of Regulatory Economics Aims and scope Submit manuscript

Abstract

Two features of the century-old policy goal of promoting universal telephone service in the United States have been enduring. Policymakers have focused on (1) wireline telephone (and more recently, fixed-line broadband) services and (2) households. The widespread adoption of mobile telephones compels a fresh examination of this focus. We construct a new measure of universal connectivity which accounts for consumers’ choices of communications technologies and for their geographic mobility over the course of the day. This measure, in turn, compels a conceptual and empirical investigation of the determinants of mobile telephone diffusion within families. Our estimations of intra-household demand for mobile service permit us to develop simulations that estimate the economic impact of modernizing a key element of existing universal service policy (viz., the Lifeline Program) to reflect the goal of improving individual connectivity. We find that a policy expansion from a single subsidy per household to multiple subsidies per eligible household members would increase mobile subscriptions by 2.25 million and Lifeline costs by $250 million.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Notes

  1. See Riordan (2002) for a detailed review of the universal service economics literature.

  2. See, e.g., the Broadband Data Improvement Act: Pub. L. 110-385, October 10, 2008, which reports that “[c]ontinued progress in the deployment and adoption of broadband technology is vital to ensuring that our Nation remains competitive and continues to create business and job growth.”

  3. For a discussion of the recent extension of public efforts to promote universal service via wireless connectivity, see Ukhaneva (2015).

  4. See, Macher et al. (2016) for a discussion of consumers’ propensities to substitute wireless for wireline telephone service.

  5. FCC, 19th Annual Wireless Competition Report, Sept. 23, 2016, Para. 45, https://apps.fcc.gov/edocs_public/attachmatch/DA-16-1061A1.pdf.

  6. We assume that individuals have their cellphones with them, have their cellphones turned on, and are located in a mobile service area.

  7. Consistent with the historical measurement of universal service we abstract from the communications potential for individuals who have wireline access to telephone service through their workplace. In this sense, our measure of universal connectivity (and similarly prior measures of universal service) is a conservative measure of the actual connectivity that individuals may enjoy. It parallels, however, the similar omission in the historical measurement of universal service, which has never incorporated workplace access to communications.

  8. See http://www.cdc.gov/nchs/nhis/about_nhis.htm for a detailed overview.

  9. Surveyed households track U.S. population demographic characteristics closely (Macher et al. 2016). We employ CDC-established sampling weights as a robustness check to confirm that the empirical results are not affected by NCHS sampling methods.

  10. Figure 6 in the online Appendix provides the average number of cellphones per household member at different household age categories and over time. See Macher et al. (2017).

  11. This allocation is driven by: (1) observed empirical regularities in household mobile telephone distribution patterns; (2) survey data of mobile telephone ownership patterns by gender, age, race and education [e.g., Rainie (2013)]; and (3) estimation results described below.

  12. Alternative mobile telephone assignments quantitatively alter the universal connectivity measure only slightly; the qualitative features and implications of the construct remain intact. For example, an assignment based solely on individual household member age (e.g., oldest members of a household get priority in cellphones assignment) produces indistinguishable results from those reported.

  13. Cellphones are assumed to provide complete coverage and connectivity for individuals throughout the entire day. As this assumption is less plausible in the early sample years than in the latter sample years, this measure may accordingly overstate individual connectivity in our early sample years. As the ATUS only samples individuals at least 15 years old, we exclude those under 15 years old in constructing the connectivity measure. Finally, ATUS data are reported as averages across several years: statistics are available for the 2003–2007 period at http://www.bls.gov/tus/tables/a3_0307.pdf; for the 2007–2011 period at https://www.bls.gov/tus/tables/a3_0711.pdf; and for the 2009–2013 period at http://www.bls.gov/tus/tables/a3_0913.pdf.

  14. For instance, 95.6% of individuals are at home at midnight: Hence, the subsample drawn contains 95.6% of individuals.

  15. Figure 7 in the online Appendix aggregates across a 24-h day to provide an average connectivity in each sample year. Average universal connectivity increases significantly: from 81% in 2003 to 92% in 2013. That is, by 2013 over the course of any day an average of 92% of Americans were connected to the telephone network. See Macher et al. (2017).

  16. See, e.g., Dekimpe et al. (1998), Gruber and Verboven (2001) and Rouvinen (2006).

  17. See, e.g., Ward and Woroch (2010), Macher et al. (2016) and Grzybowski and Verboven (2016).

  18. See, e.g., Taylor (2012) for a detailed discussion.

  19. We abstract from how usage prices may affect calling intensity. Specifically, because most fixed and mobile subscription plans include a “bucket” of minutes, the marginal price of an additional call is zero as long as the subscriber has not exhausted her allowance. We thus consider the effective marginal price of usage to be zero so that every urge to call is price unconstrained.

  20. The urge to communicate likely varies by time of day. The model described here reflects communications demand at a certain point in time.

  21. If i particularly values i to j communication (i.e., \(u_{ij}>> u_{ji}\)), but j is frequently away (i.e., \(\phi _j\) is low), i’s marginal utility increases with j’s mobile subscription. This condition may thus lead to inter-personal “side-payments” to support j’s subscription even when—absent those payments—j would chose not to subscribe. Such side-payments are most frequent between family members. See Becker (1974, 1981) for discussion.

  22. This objective function is implied by assumptions given in Harsanyi (1955, 1978). Aribarg et al. (2010) provide empirical support for this approach in the case of cellphone adoption models.

  23. Some firms offer discounts for multiple cellphone subscriptions. Our data are not sufficiently granular to capture this phenomenon empirically, and as such, we make the simplifying assumption that household mobile expenditures are additive in these different charges.

  24. These data were provided to us by Rosston et al. (2008). While many local phone carriers offer “measured service” in which customers pay smaller monthly subscription charges and—after a call or minute allowance—marginal charges per minute, industry sources indicate that the percentage of customers who avail themselves of this option is de minimis. We accordingly focus on monthly rate variations.

  25. Population weights are necessary because we only know the county where each household resides but not the household’s relevant wire center within that county. The price measure thus represents a population-weighted county-level average of wire center prices.

  26. The FCC Reference Book was produced by the Industry Analysis and Technology Division within the FCC’s Wireline Competition Bureau. This annual publication provided local exchange carrier (LEC) landline rates in 95 U.S. urban areas until it was discontinued in 2008.

  27. ARPU includes revenue related to service provision, such as roaming charges, long distance toll calling, usage-related charges, activation fees, voicemail and other services fees. ARPU does not include revenue related to handset rental or purchase charges.

  28. FCC Competition Reports indicate national carriers’ market share of total mobile service revenue increased from 87% in 2007 to 96% in 2013. These national carriers largely implemented uniform pricing over this period. Any limited time promotional pricing offers were generally available on a nationwide basis. While price variation due to regional carriers might have existed in the earliest years of the sample, it has largely disappeared along with these regional carriers. Any variation in local price indices that arises from regional carriers is small relative to the variation that arises from state and local taxes.

  29. COST reports are available beginning in 1999 and every three years thereafter (i.e., 2001, 2004, 2007 and 2010). See COST (1999, 2002, 2005, 2008, 2011) and Mackey (2008, 2011) for specific mobile telecommunications service information.

  30. The first two COST reports provide a single tax rate that blends state and local taxes for fixed and mobile service. The latter COST reports separate taxes levied on fixed and mobile service.

  31. Voice Over Internet Protocol (VoIP) introduction over 2003–2013 allowed households to utilize a computer or cable connection instead of traditional Time Division Multiplex (TDM) connection for communication. While VoIP raises the possibility that households do not identify a computer- or cable-connected telephone as a “traditional” landline, we discount this possibility given the NHIS question wording: “Is there at least one telephone inside your home that is currently working and is not a cellphone?” This question does not capture communications involving computer connections or cellphones, but potentially does capture communications involving cable connections.

  32. We estimated the model without exposure variable, and confirm that estimation results are nearly identical to those presented in Table 3.

  33. The model is estimated using Stata via the ivpoisson routine.

  34. These data were generously provided to us by Fremeth et al. (2014). Data are collected daily and then averaged over the year for all states except Alaska (which is assumed 100% Republican). See Fremeth et al. (2014) for a detailed description.

  35. USGS topography data are posted in the Natural Amenities Scale file and available at http://www.ers.usda.gov/data-products/natural-amenities-scale.aspx.

  36. These percentage changes are calculated using the following formula: \(\frac{Y_h^1}{Y_h^0} = e^{\hat{\beta }_k\Delta X_{hk}}\), where \(Y_h^0\) is the initial number of cellphones per household; \(Y_h^1\) is the number of cellphones per household when the independent variable of interest \(X_{hk}\) changes; \(\Delta X_{hk}\) is the change in the independent variable; and \(\hat{\beta _k}\) is the estimated coefficient associated with \(X_k\).

  37. We determine goodness of fit in Poisson model estimation using the squared coefficient of correlation between fitted and observed values of the dependent variable.

  38. The possibility exists that other unobserved family characteristics, such as whether children living in the home are with someone other than a father or a mother, may impact cell phone adoption rates. Our data, however, do not permit us to test for these effects.

  39. It is estimated that 80% of the cellphones in use as of 2016 were smartphones. See, FCC 19th Wireless Competition Report, Para. 121, Chart VII.A.1.

  40. Another demand-side program in place during our data window was “Link-Up America,” which subsidizes initial phone service subscriptions for low-income households. Three federal supply-side programs designed to encourage universal service deployment are: High Cost, Libraries and Schools, and Rural Health Care. These programs respectively provide subsidies to telecommunications carriers, schools, and healthcare facilities to increase connectivity. An analysis of whether these supply-side programs may impact intra-household mobile telephone service demand is left for future research. Additionally, telephone companies were historically encouraged to deploy payphones to provide telephone access. With the introduction and diffusion of mobile telephone service these policy efforts have subsided.

  41. These qualifying programs include Medicaid; Supplemental Nutrition Assistance Program (Food Stamps or SNAP); Supplemental Security Income; Federal Public Housing Assistance (Section 8); Low-Income Home Energy Assistance Program’ Temporary Assistance for Needy Families; National School Lunch Program’s Free Lunch Program; Bureau of Indian Affairs General Assistance; Tribally-Administered Temporary Assistance for Needy Families; Food Distribution Program on Indian Reservations; Head Start (if income eligibility criteria are met); and other applicable state-level assistance programs.

  42. A 2016 FCC Order further expanded the Lifeline Program to allow eligible households subsidies on fixed broadband service subscription. See https://www.fcc.gov/document/fcc-modernizes-lifeline-program-digital-age for more detail.

  43. Eligible households are considered those: (1) with incomes at or below 124% of the poverty threshold (i.e., the closest threshold among income categories in our data to the 135% level used by the FCC); or (2) with at least one member participating in a federal (i.e., SNAP, Medicaid, SSI) or state or county low-income qualifying program. NHIS data indicates whether a household participates in one of these portals, but there are a few miscellaneous state-specific eligibility programs that exist which are impossible to capture. The vast majority (84%) of Lifeline subscribers enter the program via the eligibility criteria identified. See http://usac.org/_res/documents/about/quarterly-stats/LI/Subscribers-by-Eligibility-Program.pdf. As our data include a small percentage (i.e., below 1%) of Native Americans households, we exclude them from the simulation analysis due to insufficient sample size.

  44. See Burton et al. (2007) for a detailed discussion and analysis of the history and determinants of observed take rates. USAC statistics indicate 33% Lifeline Program take-rates in November, 2015. See http://usac.org/_res/documents/about/quarterly-stats/LI/Subscribers-by-Eligibility-Program.pdf.

  45. In limiting any increase in cellphone subscriptions to eligible households, the simulation abstracts from any positive network effects that changes in subscriptions among eligible housseholds may have on subscriptions among non-eligible households. Consequently, our simulation is a conservative estimate of the total subscription increase associated with the policy change.

  46. See Ukhaneva (2015) for more detail.

  47. This increase does not include changes in the costs of administering the change in the Lifeline program or any economic distortions that may occur as a consequence of alternative public financing mechanisms to fund the program expansion. Additionally, we do not examine the potential for the de novo costs of a universal service policy to be reduced by reductions in expenditures on existing universal service funds other than Lifeline.

  48. See statistics published by the International Telecommunications Union http://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx.

References

  • Aribarg, A., Arora, N., & Kang, M. Y. (2010). Predicting joint choice using individual data. Marketing Science, 29(1), 139–157.

    Article  Google Scholar 

  • Becker, G. S. (1974). A theory of social interactions. Journal of Political Economy, 82(6), 1063–1094.

    Article  Google Scholar 

  • Becker, G. S. (1981). Treatise on the family. Cambridge, MA: Harvard University Press.

    Google Scholar 

  • Burton, M., Macher, J., & Mayo, J. W. (2007). Understanding participation in social programs: Why don’t households pick up the lifeline? The BE Journal of Economic Analysis & Policy, 7(1), 1–28.

    Google Scholar 

  • Cameron, A. C., & Trivedi, P. K. (2005). Microeconometrics: Methods and applications. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  • Council on State Taxation. (1999). 50-state study and report on telecommunications taxation.

  • Council on State Taxation. (2002). State study and report on telecommunications taxation. Washington DC: Telecommunications Tax Force Special Report.

  • Council on State Taxation. (2005). State study and report on telecommunications taxation. Washington DC: Telecommunications Tax Force Special Report.

  • Council on State Taxation. (2008). State study and report on telecommunications taxation. Washington DC: Telecommunications Tax Force Special Report.

  • Council on State Taxation. (2011). State study and report on telecommunications taxation. Washington DC: Telecommunications Tax Force Special Report.

  • Dekimpe, M. G., Parker, P. M., & Sarvary, M. (1998). Staged estimation of international diffusion models—An application to global cellular telephone adoption. Technological Forecasting and Social Change, 57(1–2), 105–132.

    Article  Google Scholar 

  • Federal Communications Commission. (2016). Mobile wireless competition report (19th annual).

  • Fremeth, A. R., Holburn, G. L., & Spiller, P. T. (2014). The impact of consumer advocates on regulatory policy in the electric utility sector. Public Choice, 161(1–2), 157–181.

    Article  Google Scholar 

  • Goolsbee, A., & Klenow, P. J. (2002). Evidence on learning and network externalities in the diffusion of home computers. Journal of Law & Economics, 45(2), 317–343.

    Article  Google Scholar 

  • Gruber, H., & Verboven, F. (2001). The diffusion of mobile telecommunications services in the European Union. European Economic Review, 45(3), 577–588.

    Article  Google Scholar 

  • Grzybowski, L., & Verboven, F. (2016). Substitution between fixed-line and mobile access: The role of complementarities. Journal of Regulatory Economics, 49(2), 113–151.

    Article  Google Scholar 

  • Harsanyi, J. C. (1955). Cardinal welfare, individualistic ethics, and interpersonal comparisons of utility. Journal of Political Economy, 63(4), 309–321.

    Article  Google Scholar 

  • Harsanyi, J. C. (1978). Bayesian decision theory and utilitarian ethics. American Economic Review, 68(2), 223–228.

    Google Scholar 

  • Macher, J. T., Mayo, J. W., Ukhaneva, O., & Woroch, G. (2016). Demand in a portfolio-choice environment: The evolution of telecommunications. Georgetown University Working Paper.

  • Macher, J. T., Mayo, J.W., Ukhaneva, O., & Woroch, G. (2017). Online appendix to “From Universal Service to Universal connectivity”. Available at http://cbpp.georgetown.edu.

  • Mackey, S. (2008). Excessive taxes and fees on wireless service: Recent trends. State Tax Notes, 47(7), 519–531.

    Google Scholar 

  • Mackey, S. (2011). A growing burden: Taxes and fees on wireless service. State Tax Notes, 14, 475–487.

    Google Scholar 

  • Rainie, L. (2013). Cell phone ownership hits 91% of adults. Fact Tank: News in the Numbers (Vol. 6).

  • Riordan, M. H. (2002). Universal residential telephone service. In M. Cave, S. Majumdar, & I. Vogelsang (Eds.), Handbook of telecommunications economics. Amsterdam: Elsevier.

    Google Scholar 

  • Rosston, G., Savage, S., & Wimmer, B. (2008). The effect of private interests on regulated retail and wholesale prices. Journal of Law & Economics, 51(3), 479–501.

    Article  Google Scholar 

  • Rouvinen, P. (2006). Diffusion of digital mobile telephony: Are developing countries different? Telecommunications Policy, 30(1), 46–63.

    Article  Google Scholar 

  • Taylor, L. D. (2012). Telecommunications demand in theory and practice. Berlin: Springer.

    Google Scholar 

  • Ukhaneva, O. (2015). Universal service in a wireless world. Georgetown University Working Paper. https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnx1a2hhbmV2YW9sZ2F8Z3g6NTA5YjUyYzM5NmM0NjRmNg.

  • Ward, M., & Woroch, G. (2010). The effect of prices on fixed and mobile telephone penetration: Using price subsidies as natural experiments. Information Economics and Policy, 22(1), 18–32.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. McDonough School of Business, Georgetown University, Washington, DC, USA

    Jeffrey T. Macher, John W. Mayo & Olga Ukhaneva

  2. Department of Economics, University of California, Berkeley, CA, USA

    Glenn A. Woroch

Authors
  1. Jeffrey T. Macher
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. John W. Mayo
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Olga Ukhaneva
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Glenn A. Woroch
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to John W. Mayo.

Additional information

We are grateful to Carol Corrado, Amy Farmer, Anna-Maria Kovacs, Peter Fox-Penner, J. Bradford Jensen, Scott Savage, Scott Wallsten, Michael Ward and seminar participants at the Federal Communications Commission (FCC) for comments on earlier drafts. Any errors or omissions remain our own.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 66 KB)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Macher, J.T., Mayo, J.W., Ukhaneva, O. et al. From universal service to universal connectivity. J Regul Econ 52, 77–104 (2017). https://doi.org/10.1007/s11149-017-9336-8

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11149-017-9336-8

Keywords

JEL Classification

Search

Navigation