high-speed broadband internet in north carolinacompared to member nations of the organisation for...
TRANSCRIPT
High-Speed Broadband Internet in North Carolina:
It’s fast, but is it effective?
By
Eleanor Blake
March 14, 2011
A paper submitted to the faculty of
The University of North Carolina at Chapel Hill
in partial fulfillment of the requirements for the degree
Master of Public Administration
This paper represents work done by a UNC MPA student. It is not a formal report of the School of
Government, nor is it the work of School of Government faculty.
Executive Summary
Improved access and faster speeds of broadband Internet are predicted to transform life for
Americans. As access to and adoption of high-speed broadband Internet has increased in the past
decade, how has it impacted the economy, education and quality of life for North Carolinians? This
research looks at the effect of broadband adoption in North Carolina and finds that the current
adoption rate of broadband Internet has significant influence on select economic factors, but no
measurable influence on a select set of social or educational indicators.
2 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Background
High-speed broadbandi Internet is no longer a convenience or just for entertainment. It is now a national goal
that is predicted to improve education, quality of life and economic performance. The extensive National
Broadband Plan refers to it as essential 21st century infrastructure that is “transforming the landscape of
America more rapidly and more pervasively than earlier infrastructure networks” (Federal Communications
Commission, 2010, p. 3). The Pew Internet and American Life Project’s Home Broadband Adoption 2009
(Horrigan, 2009) report found that 43% of Americans believe lack of Internet to be a disadvantage for
conducting particular activities, primarily job seeking and building career skills.
North Carolina, like many other states, and the federal government are working to improve the speed and
availability of broadband in hopes of achieving an array of positive outcomes. On a global scale, the United
States has room to improve. While in the year 2000 the U.S. had the highest percentage of population
connected to the Internet, it is now ranked 18th for both adoption and access
ii and 24
th for advertised speed
compared to member nations of the Organisation for Economic Co-operation and Development.iii North
Carolina ranked 14th among U.S. states for broadband adoption rates in 2009.
iv The Federal Communications
Commission (FCC) adoption data show the average broadband subscription in North Carolina as just below
400 homes per 1,000 (at a speed of 768 kbps). See Appendix 2: Broadband Adoption Map.
North Carolina is active in expanding broadband infrastructure and improving broadband speeds, primarily
to underserved areas and key institutions.v In 2010, North Carolina received over $275 million in federal
grants for broadband expansion, and initiatives such as the e-NC Authority,vi MCNC
vii and other businesses
are rolling out fiber to connect North Carolina homes and institutions. As the state moves forward with
broadband initiatives, an understanding of the effects of broadband is needed. This research explores whether
broadband has had a measurable impact on 2009 economic, social and educational outcomes. Empirical
investigation into the impact of broadband efforts can inform North Carolina practitioners and policy makers
working to build and implement the state’s Internet strategy.
Previous Research
Research on broadband began in earnest in the early 2000s after the technology-influenced economic boom
of the late 1990s. During this time Internet deployment was in its early stages and much of the research was
hypothetical and forward-looking, predicting billions of dollars in GDP growth and over a million jobs
created (Gillett, Lehr, & Osorio, 2006, p.6). More recent reports have analyzed the impact of broadband in
municipalities, states, regions and the nation and have primarily utilized longitudinal methods, analysis of
growth and change rates, and examination of comparable communities.
A number of studies have reported economic growth related to broadband. Measuring Broadband’s
Economic Impact (Gillett, Lehr, & Osorio, 2006) found that, nationwide, broadband contributed to job
creation, the number of businesses (with greater growth in the IT-intensive sector), and property values, but
no statistical impact was found for wages. A study by Jed Kolko (2010) also found mixed results nationwide,
reporting broadband was not associated with employment rate or wages, but did impact business activity. An
increase in gross sales in a community in Florida was attributed to its municipally-owned broadband network
(Ford & Koutsky, 2005). And a study of the commercial benefits of broadband in the rural Appalachia region
found modest impact on firm productivity and wages in particular regions and scenarios (Burton & Hicks,
2005).
3 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Research on the social benefits of broadband is more prognostic, predicting improved access to health care,viii
ix reduced energy consumption,
x and improvement of public safety.
xi A survey of the residents of New
Brunswick rural areas of Canada found that broadband improved quality of life through streamlined daily
operations and enhanced access to information, without negatively impacting longstanding social
relationships (Selouani & Hamam, 2007).
Educational research has focused primarily on Internet access within schools and its effect on educational
performance, highlighting the opportunities and challenges of technology in primary and secondary schools.
Effects on test scores have been mixed (Underwood, et al., 2005); a study by Goolsbee and Guryan (2005) on
the impact of E-rate government Internet subsidies in schools found no significant effect on performance
scores.
Beyond survey-based studies and data collected by the e-NC Authority, the impact of broadband has not
been empirically evaluated for North Carolina. This research seeks to provide greater understanding of the
influence broadband Internet has had on North Carolina economic, social and educational attainment.
Methodology
Data Analysis Methods
This study uses correlations and regression analyses to examine the relationship between broadband Internet
adoption in North Carolina and a set of economic, social and educational indicators. Correlations were
examined between broadband and indicators proven or hypothesized to be affected by broadband in prior
research. xii
xiii
In order to test for spurious relationships in the correlations, multiple regression analyses were
used to determine if broadband is a significant influencing factor on the dependent variables when
controlling for other significant independent variables and controls, such as population density, educational
attainment and income per capita. Broadband adoption data, along with other influencing factors and
controls, were tested in multiple models to measure the influence and power with which broadband affects a
range of economic, social and educational indicators.
Eight regression models were run to test for broadband’s impact. The dependent variables selected for the
models are those that have shown to be influenced by broadband in previous research or are key indicators of
economic, educational and social wellbeing. For economic influence, three dependent variables were
selected. Income per capita and median home value were used as measures of current personal wealth,xiv
while retail sales was used as a measure of current economic productivity.xv
Educational impact is gauged
using End-of-Grade (EOG) school test scores and End-of-Course (EOC) school test scores.xvi
The social
dependent variables used for this research were crime index, teen pregnancy and poverty.xvii
Previous
research was used to inform model construction and the inclusion of control variables with careful attention
to avoiding multicollinearity. See Appendix 3: Variable Definitions.
Data Collection Methods
Broadband adoption data were collected at the county level from the FCC’s form 477 reports, the standard
for broadband subscribership data.xviii
This data represent fixed lines of broadband Internet in homes and
does not include access to the Internet via wireless or mobile connections. Adoption rates are reported in
ranges, 0 to 5,xix
and indicate households per 1,000 with broadband connections. Rates are connections at the
768 kbps speed, the FCC’s high-speed definition in 2009 when the data were collected. North Carolina
4 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
adoption rates range from 2 to 5 with an average of 2.83, indicating adoption at the higher end of the 200 to
400 connections per 1,000 households range.
The three economic dependent variables, income per capita, median home value and retail sales, were
collected from the NC Department of Commerce and are calculated based on U.S. Census data and statistical
projections conducted by the geographic information systems company ESRI. School test scores and all other
educational data are from the NC Department of Public Instruction and were aggregated to the county level
based on reporting by school districts in the NC School Report Cards. For the social dependent variables,
teen pregnancy rate is collected by the NC Department of Health and Human Services and is measured as the
number of babies born to teenagers age 15 to 19 per 1,000 births. Crime index, a combination of eight FBI-
defined crimes, comes from the NC Department of Commerce and data calculated by the company Applied
Geographic Solutions. Poverty rate, the percent of people earning less than an income threshold set by the
federal government, also comes from the Department of Commerce and statistical calculations by ESRI.
Other demographic data were included in the research as independent variables and controls and were
collected from a variety of sources, including the U.S. Census, NC Department of Commerce and the NC
Department of Public Instruction. See Appendix 3: Variable Definitions and Appendix4: Data Descriptive
Statistics.
Findings
Results of the correlation analysis indicate significant relationships with seven of the eight outcome
variables. Broadband adoption has a strong positive correlation with the economic indicators income per
capita and median home value, and a weaker, but significant and positive correlation with retail sales. The
educational indicators, EOG and EOC school test scores, both have a relatively weak, positive relationship
with broadband adoption. There is a significant negative relationship between broadband adoption and
poverty rate and between broadband and teen pregnancy, while the correlation with crime index is not at a
statistically significant level (Table 1). With the exception of crime index, relationships were confirmed in
the expected direction for economic, educational and social indicators. These initial results provide evidence
of a potential influence of broadband adoption on key economic, social, and educational outcomes.
Table 1: Correlations Results
Outcome Correlations with Broadband Adoption
Outcome Variable Correlation with Broadband, R
Income per Capita -0.591**
Median Home Value -0.644**
Retail Sales -0.306**
EOG Test Scores -0.347**
EOC Test Scores -0.288**
Crime Index -0.172**
Teen Pregnancy -0.455**
Percent Poverty -0.475**
**significant at the 0.01 level
5 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
While initial correlations indicated statistically significant relationships between broadband economic, social
and educational measures, regression analysis allowed for more sophisticated investigation of these
relationships. Regression results for the eight models found that the adoption of broadband was significant in
only two, those measuring economic attainment: income per capita and median home values. The model
examining income per capita explained 80.6% of the variance, and broadband as an independent variable was
significant at the <0.01 level, influencing income along with (but with less power than) educational
attainment and population density. The median home value model accounted for 65.4% of the variance in
home values, and broadband was a significant predictor along with income and urban-rural designation.
Contrary to findings in other regions of the U.S.,xx
broadband was not a significant influencing factor on
retail sales when combined with the major factors of income, urban-rural designation and poverty. These
findings show that broadband does, to some degree, positively influence personal economy, but the lack of a
statistically significant relationship with retail sales does not indicate impact on current business activity. The
presence and adoption of broadband in North Carolina counties, therefore, is shown to be important for the
particular economic outcomes of income and home values.
Broadband was not found to be a statistically significant predictor in the models testing selected educational
and social outcomes, despite the significant correlations. The major drivers of EOG and EOC test scores
were poverty, population density and teachers with advanced degrees. Crime index, teen pregnancy and
poverty rate were largely influenced by income (Table 2). See Appendix 6 for detailed results.
Broadband adoption, as it currently exists, is not shown by this research to be an important factor for
improvement and advancement of education and social wellbeing or for all economic outcomes at this time.
There is a possibility that certain economic measures respond more quickly to high-speed broadband
infrastructure and that the positive influence on educational attainment and social wellbeing may yet to be
fully realized. Future studies are key, but this research does not support North Carolina’s recent and past
investment in broadband availability in terms of generating all the expected positive results.
Table 2: Multiple Regression Results
Multiple Regression Results: Broadband’s Influence on Dependent Variables
Dependent Variable Broadband Beta Coefficient Broadband Significance
Income per Capita -0.198 <0.0100
Median Home Value -0.283 <0.0100
Retail Sales -0.117 0.158
EOG Test Scores -0.027 0.761
EOC Test Scores -0.000 0.999
Crime Index -0.193 0.061
Teen Pregnancy -0.069 0.497
Poverty -0.043 0.664
6 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Limitations
Because income and home values are themselves important determinants of wellbeing, reverse causality is a
potential limitation. This research does show broadband adoption to be a significant influencing factor on
income and home values. However, greater wealth allows an ability to purchase broadband and related
technologies, and there is a possibility that income as an independent factor determines broadband adoption.
Limitations also exist with the current broadband data. The changing speed and technology of high-speed
Internet means that the definition and measurement of broadband are also frequently changing. The FCC’s
1999 minimum broadband speed was 200 kbps downstream, a speed incapable of managing the degree of
high-quality video, voice, graphics and data sent and received on the Internet today. The 2010 Broadband
Plan increased the speed minimum from 768 kbps to 4 Mbps, a new definition that discounts 66% of U.S.
Internet connections.xxi
These inconsistent definitions, as well as the difficulty in gathering complete access
and adoption data from private providers, limit broadband impact studies.
Additionally, the particular data set and methods used in this research do not explore broadband’s influence
on economic productivity factors such as employment, indicators with varying relationships to broadband
according to other studies.xxii
Conclusion & Recommendations
This study finds that broadband adoption is important for North Carolina economy, influencing residents’
personal economics through increased income and home values. With no evidence of broadband’s influence
on the selected educational and social indicators, the findings are not completely consistent with the
predictions and stories of the full economic, social and educational benefits of broadband.
Further research is needed to substantiate North Carolina’s current investment and pursuits, and definitive
results will depend on the accumulated results of continued studies utilizing a variety of approaches and
improved data. Time-series analyses in prior research did find measurable broadband impact; a longitudinal
study is required to test for a potential lag in the fuller impact of broadband in North Carolina. Additionally,
comprehensive studies must include Internet connections to businesses and institutions, as well as
households, to fully capture impact. Wireless connections and mobile 3G and 4G networks also contribute to
connectedness and may prove to be more popular Internet services in the future.
With millions of dollars and other resources dedicated to broadband, North Carolina policy makers and
practitioners can evaluate this study not as validation or invalidation, but as benchmark data for current and
past broadband initiatives, as well as evidence of greater need for improved data measurement and tracking,
continued research, and strategic broadband initiatives based on measurable results.
i The FCC defines broadband, or high speed Internet, as “advanced communications systems capable of providing high-speed transmission of services such as data, voice, and video over the Internet and other networks.” Technologies include digital subscriber line (DSL), fiber optic cable, coaxial cable, wireless technology and satellite. The 2010 National Broadband Plan sets the speed standard of these technologies at 4 Mbps actual download and 1 Mbps actual upload. http://www.fcc.gov/broadband/ ii Adoption, a.k.a. subscribership or penetration, is the number or percentage of households with purchased
broadband lines. Access refers to the number or percentage of households able to access broadband, regardless of subscription.
7 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
iii 2009 data. Data collection and interpretation, as well as rankings, are controversial.
http://www.oecd.org/sti/ict/broadband iv http://www.ntia.doc.gov/reports/2010/ESA_NTIA_US_Broadband_Adoption_Report_11082010.pdf
v Gaps in availability in North Carolina occur according to rural-urban areas and by tiers. Nationwide, disparity is also
attributed to differences in racial and ethnic backgrounds. http://www.pewcenteronthestates.org/ report_detail.aspx?id=59149 vi The e-NC Authority’s mission is to ensure that residents and businesses in North Carolina, particularly in rural and
distressed urban areas, are aware of high-speed Internet services and capabilities, have affordable access to them, and know how to use them. e-NC.org vii
MCNC received two federal grants for expansion of the NC Research and Education Network. https://www.mcnc.org/btop viii
Pew Internet and the American Life Project’s Internet “health seekers” survey. http://www.pewInternet.org/ Reports/2000/The-Online-Health-Care-Revolution/Summary.aspx ix Report on the impacts of telemedicine. http://www.broadband.gov/docs/ws_adoption_unserved/
ACLP_Broadband_Telemedicine_Overview.pdf x Smart grid technology can save energy. http://www.e-renewables.com/documents/Smart%20Grid/
Smart%20Grid%20Enabling%20the%2021st%20Century%20Economy.pdf xi The 2010 National Broadband Plan (The Federal Communications Commission, 2010, Chapter 16).
http://www.broadband.gov/plan/ xii
“Like electricity a century ago, broadband is a foundation for economic growth, job creation, global competitiveness and a better way of life.” The National Broadband Plan (The Federal Communications Commission, 2010, p. xi). http://www.broadband.gov/plan/ xiii
The e-NC broadband report (Baller & Lide, 2008, p. 2). xiv
Gillett, et al. (2006) found no impact on wages (a proxy for income), but it is used in this paper’s research as a strong indicator of long-term economic performance. Kelly (2003) showed home values are higher in communities with greater broadband infrastructure. xv
Ford & Koutsy (2005) found greater growth in gross sales in a Florida county that provides broadband utility than in comparable counties. xvi
Goolsbee & Guryan (2005) used standardized end-of-year test scores as a measure of educational attainment. xvii
The NC Rural Center Data Dictionary (2009) collects data on poverty rate, teen pregnancy rate and crime as measures of wellbeing. xviii
Purchased broadband home connections of at least 768 kbps downstream advertised speed reported by telephone companies, cable system operators, terrestrial wireless service providers, satellite service providers and other facilities-based providers of advanced telecommunications capability. xix
0 = zero, 1 = zero < x <= 200, 2 = 200 < x <= 400, 3 = 400 < x <= 600, 4 = 600 < x <= 800, 5 = 800 < x xx
Ford & Koutsy (2005) found greater growth in gross sales in a Florida county that provides broadband utility than in comparable counties. xxi
FCC 9/10 news release. http://www.fcc.gov/wcb/iatd/comp.html xxii
Kolko (2010) found no impact of broadband on the unemployment rate. Gillett et al. (2006) found increased employment due to broadband.
8 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 1: Bibliography
Balanskat, A., Blamire, R., & Kefala, S. (2006). The ICT Impact Report: A review of studies of ICT
impact on schools in Europe. European Schoolnet.
Baller, J., & Lide, C. (2008). Bigger Vision, Bolder Action, Brighter Future: Capturing the Promise of
Broadband for North Carolina and America. The Baller Herbst Law Group for e-NC.
Bifulco, R., & Ladd, H. F. (2006). The Impacts of Charter Schools on Student Achievement: Evidence
from North Carolina. American Education Finance Association.
Boyd, D., Lankford, H., Loeb, S., Rockoff, J., & Wyckoff, J. (2008). The Narrowing Gap in New York
City Teacher Qualifications and Its Implications for Student Achievement in High-Poverty
Schools. Journal of Policy Analysis and Management, 27(4), 793-818.
Burton, M. L., & Hicks, M. J. (2005). The Residential and Commercial Benefits of Rural Broadband:
Evidence from Central Appalachia. West Virginia Development Office, A Vision Shared.
Crandall, R. W., & Jackson, C. L. (2001). The $500 Billion Opportunity: The Potential Economic Benefit
of Widespread Diffusion of Broadband Internet Access. Criterion Economics, L.L.C.
Federal Communications Commission (2010). Connecting America: The National Broadband Plan.
Feser, E. (2005). North Carolinians Online: Trends from the Citizen Surveys 1999-2004. The e-NC
Authority.
Ford, G. S., & Koutsky, T. M. (2005, April). Broadband and Economic Development: A Municipal Case
Study from Florida. Applied Economic Studies.
Futernic, K. (2010). Incompetent Teachers or Dysfunctional Systems? Re-framing the Debate on Teacher
Quality and Accountability. WestEd.
Gillett, S. E., Lehr, W. H., & Osorio, C. A. (2006). Measuring Broadband's Economic Impact. U.S.
Department of Commerce, Economic Development Administration.
Goolsbee, A., & Guryan, J. (2006). The Impact of Internet Subsidies in Public Schools. Review of
Economics and Statistics, 88(2), 226-347.
Horrigan, J. (2009). Home Broadband Adoption 2009. Pew Internet & American Life Project,
Washington, D.C.
Kelly, D. J. (2003). A Study of the Economic and Communitiy Benefits of Cedar Falls, Iowa's Municipal
Telecommunications Network. Iowa Association of Municipal Utilities, Telecommunications
Division.
Kolko, J. (2010). Does Broadband Boost Local Economic Development? Public Policy Institute of
California.
9 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Machin, S., McNally, S., & Silva, O. (2006). New Technology in Schools: Is There a Payoff? The
Institute for the Study of Labor.
National Telecommunications and Information Administration (2010). Exploring the Digital Nation:
Home Broadband Internet Adoption in the United States. U.S. Department of Commerce.
Pew Center on the States (2010). Bringing America Up to Speed: States' Role in Expanding Broadband.
Pew Charitable Trusts.
Selouani, S.-A., & Hamam, H. (2007). Social Impact of Broadband Internet: A Case Study in the
Shippagan Area, a Rural Zone in Atlantic Canada. Journal of Information, Information
Technology, and Organizations, 2, 79-94.
Underwood, J., Ault, A., Banyard, P., Bird, K., Dillon, G., Hayes, M., et al. (2005). The Impact of
Broadband in Schools. Nottingham Trent University. Becta ICT Research.
10 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 2: Broadband Adoption Map
11 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 3: Variable Definitions
Variable Definition Source Prior Studies Study Findings Used in research?
Broadband
Adoption
Lines in service, or purchased household Internet connections, of at least 768 kbps downstream
FCC Form 477
Used as the measure of broadband for all studies consulted
All reviewed studies found that broad-band improved select indicators
Yes
Access
Percent of households with access to at least 4 Mbps downstream
National Broadband Plan Maps
N/A None No. Not recommended for analysis due to statistical pre-diction methods
Economic
Retail Sales
Derived from receipts (net of sales taxes, refunds, and returns) of businesses primar-ily engaged in the re-tailing of merchandise
NC Depart-ment of Commerce EDIS database
Ford & Koutsky, Broadband and Economic Development: A Municipal Case Study from Florida (2005)
Greater growth of gross sales was shown in a Florida county with broadband utility than in comparable counties
Yes
Job creation The number of new jobs (based on a survey of NC indust-rial and commercial companies)
NC Depart-ment of Commerce EDIS database
Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Greater growth in employment ob-served in commun-ities with greater broadband
No. Complete data for all industries not available
Kolko, Broadband and Local Growth (2010)
Found a positive relationship between broadband and employment growth, especially in IT-intensive businesses and in areas with lower population density
Wages Average gross annual wages across all industries. (Includes salaries, bonuses, stock options, tips and other gratuities, and the value of meals and lodging, where supplied)
US Bureau of Labor Statistics
Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Broadband does not improve wages
Yes, income used
Kolko, Broadband and Local Growth (2010)
Unaffected by broadband
12 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Burton & Hicks, The Residential and Commercial benefits of rural broadband: Evi-dence from Central Appalachia (2005)
Hypothesizes that broadband will increase wages due to observed improve-ment in firm product-ivity
Number of Establish-
ments
An economic unit that produces goods or services
US Bureau of Labor Statistics
Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Growth in estab-lishments, especially IT-related, due to broadband from 1998-2002
No. Little correlation
Small Employers
Businesses with few or no employees (and not represented in Census establish-ment data)
N/A Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Used as a control for the effect of broad-band on other econ-omic indicators
No. Unable to obtain data
Median Home Value
Median value of owner-occupied housing
NC Depart-ment of Commerce EDIS database
Kelly, A Study of the Economic and Community Benefits of Cedar Falls, Iowa’s Muni-cipal Telecommun-ications Network (2003)
Higher in a com-munity with municipal broadband infrastruc-ture compared to similar communities
Yes
Kolko, Broadband and Local Growth (2010)
Hypothesized that broadband would increase home values and tax base, but not increase employment rate or wages
Unemploy-ment Rate
The number of unemployed divided by the labor force
NC Depart-ment of Commerce EDIS database
Kolko, Broadband and Local Growth (2010)
Unaffected by broadband
Yes
Income per Capita
Average income per individual
NC Depart-ment of Commerce EDIS database
Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Broadband does not significantly impact wages
Yes. Is a strong measure of long-term economic performance
Labor Force The total number of people employed and seeking employment
NC Depart-ment of Commerce EDIS database
Gillett et al., Measuring the Economic Impact of Broadband Deployment (2006)
Used as a control for the effect of broad-band on other econ-omic indicators
No. Unable to obtain partici-pation rate data
IT-Intensive Firms
The share of establishments in the IT-intensive sector
N/A Gillett et al. Measuring the Economic Impact of Broadband Deployment (2006)
Used as a control for the effect of broad-band on other econ-omic indicators
No. No standard definition
13 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Social
Voter Turnout
The number who voted as a percentage of registered voters for the November 2008 election
Democracy North Carolina
O’Brian, Measuring Citizen Engagement: The North Carolina Civic Index (2003)
Voter turnout and the improvement in citizen engagement build and improve communities
No. Not correlated and entire civic index not available at county level
Jacksonville Com-munity Council, Inc., Quality of Life Progress Report for Jacksonville and Northern Florida (2010)
Use voter turnout rate as a measure of quality of life
NC Rural Center Data Dictionary (2009)
Used as a measure of community health
Teen Pregnancy
The number of babies born to teenagers (females age 15-19) per 1,000 births
NC DHHS, State Center for Health Statistics
Adam Mellows-Facer, Social Indicators (2004)
Use teen pregnancy statistics as a snap-shot of social health, among many indi-cators
Yes
Jacksonville Com-munity Council, Inc., Quality of Life Progress Report for Jacksonville and Northern Florida (2010)
Use teen pregnancy rate as a measure of quality of life
Crime Index
An index of the eight crimes the FBI combines to produce its annual crime index
NC Depart-ment of Commerce EDIS database
Adam Mellows-Facer, Social Indicators (2004)
Use number of crimes as a snapshot of social health, among many indicators
Yes
NC Rural Center Data Dictionary (2009)
Used as a measure of community health
Infant Mortality
Infant deaths per 1,000 live births
State Center for Health Statistics, NC DHHS
NC Rural Center Data Dictionary (2009)
Used as a measure of community health
No
US Dept. of Health and Human Services
Considers infant mortality to be one of the most important indicators of health
Jacksonville Community Council, Inc., Quality of Life Progress Report for Jacksonville and Northern Florida (2010)
Used as a measure of community health
14 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Percent Uninsured
Percentage of population under age 65 without health insurance
UNC Sheps Center for Health Services Research
Jacksonville Community Council, Inc., Quality of Life Progress Report for Jacksonville and Northern Florida (2010)
Used as a measure of community health
No
Percent in Poverty
The percent of people earning less than the income threshold set by the federal government
NC Depart-ment of Commerce EDIS database
NC Rural Center Data Dictionary (2009)
Used as a measure of community health
Yes
Education
End of Grade & End of
Course scores
The percentage of students that score at or above grade level on end-of-grade and end-of-course tests
NC DPI School Report Cards
Bifulco & Ladd, The Impact of Charter Schools on Student Achievement: Evidence from North Carolina (2006)
Used as a measure of student performance for comparing different schools
Yes
Machin et al., New Technology in Schools: Is there a payoff? (2006)
An indicator of school performance used to measure the effects of technology funding
Goolsbee & Guryan, The Impact of Internet Subsidies in Public Schools (2005)
Use the California standardized end-of-year tests for math, reading and science for students in primary and secondary schools. Find no relation to Internet in the school and test scores
School funding
The total amount of funding from local, state and federal levels to each county, per student, in dollars
NC DPI School Report Cards
Machin et al., New Technology in Schools: Is there a payoff? (2006)
Technology funding found to positively impact test scores for certain grades and subjects, but other support and training needed for full impact
Yes
Students per Internet-
Connected Computer
Number of students per Internet-connected computer in each county
NC DPI School Report Cards
Balanskat et al., The ICT Impact Report: A review of studies of ICT impact on schools in Europe (2006)
Technology and connectivity in schools has positive impact on some grades and subjects
Yes
15 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Goolsbee & Guryan, The Impact of Internet Subsidies in Public Schools (2005)
Use “previous techno-logy histories” to control for historical, longitudinal, differen-ces that may affect improvement in stu-dent achievement
Teachers with
Advanced Degrees
The percentage of teachers who have completed an advanced college degree, including a master's or doctoral degree in elementary, middle and high schools
NC DPI School Report Cards
Futernick, Incom-petent Teachers or Dysfunctional Sys-tems? Reframing the debate on teacher quality and accountability (2010)
Advocates that a quality and supportive teaching environment is more necessary than individual teacher attributes
Yes
Boyd et al., The Narrowing Gap In New York City Teacher Qualifica-tions and Its Impli-cations for Student Achievement in High-Poverty Schools (2008)
Advanced degrees tend to be found in more economically prosperous areas and therefore may be highly correlated with economic status
Dropout Rate
Number of students grades 7-13 dropping out in one year, divided by total students grades 7-13
NC DPI Goolsbee & Guryan, The Impact of Internet Subsidies in Public Schools (2005)
Found no significant impact on dropout rate
Yes
Educational Attainment
The percentage of the population to attain at least a bachelor’s degree
NC Depart-ment of Commerce EDIS database
Bifulco & Ladd, The Impact of Charter Schools on Student Achievement: Evi-dence from North Carolina (2006)
Used as a control for student scores
Yes
Controls
Population density
Average number of people per square mile
NC Depart-ment of Commerce EDIS database
Kolko, Broadband and Local Growth (2010)
Found that broad-band, not population growth was the cause of economic growth
Yes
Tier Designation
Classification by the NC Department of Commerce indicating economic disadvan-tage (1) and economic prosperity (3)
NC Rural Center
NC Rural Center Data Dictionary (2009)
Used as a measure of community health
Yes
Urban-Rural Designation
Designation as urban or rural as defined by the NC General Assembly
NC Rural Center
Yes
16 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 4: Variable Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Broadband
Adoption (lines in service) 100 2 5 2.83 .711
Access (percent served) 100 .22 1.00 .7879 .17464
Economic
Retail Sales 100 9605091 12112045802 9.25E8 1.873E9
Wages 100 24660 61104 31692.07 5330.944
Number of Establishments 100 97 32369 2420.91 4573.950
Median Home Value 100 78284 214644 122434.62 30516.357
Unemployment Rate 100 .066 .161 .11063 .021836
Income per Capita 100 1.600400E4 3.633900E4 2.19294900E4 3.926953096E3
Educational
EOG Index 100 .5475 .9125 .750426 .0805034
EOC Index 100 .5210 .9270 .799070 .0742824
Dropout Rate 100 .67 5.02 2.9174 .85555
Funds per Student 100 6980 18468 9223.26 1565.747
Teacher Degree Index 96 .11 .43 .2543 .05924
Educational Attainment 100 .0858 .5430 .176116 .0838571
Social
Voter Turnout 100 .5783 .7773 .692916 .0398298
Turnout Gain 100 -.0461 .2795 .059601 .0439009
Teen Pregnancy 100 18.5 55.3 33.566 6.8529
Number of Physicians 100 .0 93.3 14.174 12.9401
Crime Index 100 6 239 88.74 41.562
Infant Mortality Rate 100 .0 20.8 7.605 4.5396
Percent Uninsured 100 .166 .288 .20239 .020691
Percent in Poverty 100 8.9 32.3 18.323 4.8959
Controls
Tier Designation 100 1 3 1.79 .743
Urban-Rural Designation 100 0 1 .15 .359
Population Density 100 9.0 1734.0 191.421 254.6498
Valid N (listwise) 96
17 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 5: Correlations of All Collected Variables with Broadband Adoption
Variable Correlation with Broadband, R
Retail Sales -0.306**
Annual Wages -0.365**
Number of Establishments -0.310**
Median Home Value -0.644**
Unemployment Rate -0.407**
Income Per Capita -0.591**
Population Density -0.336**
EOG 3rd
Grade Reading -0.411**
EOG 3rd
Grade Math -0.300**
EOG 8th
Grade Reading -0.370**
EOG 8th
Grade Math -0.201**
EOG Overall Reading -0.389**
EOG Overall Math -0.281**
EOG Subjects Index -0.347**
EOC English 1 -0.252**
EOC Algebra -0.228**
EOC Biology -0.323**
EOC Subjects Index -0.288**
Percent Students Taking SAT -0.412**
Average SAT Score -0.535**
Dropout Rate -0.101**
School Funds – Local -0.594**
School Funds – State -0.410**
School Funds – Federal -0.496**
Total Funds per Student -0.221**
Students per Internet Computer -0.165**
Teachers with Degrees – Elementary Schools -0.150**
Teachers with Degrees – Middle Schools -0.153**
18 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Teachers with Degrees – High Schools -0.158**
Teaching Degree Index -0.178**
Educational Attainment – High School Degree -0.684**
Educational Attainment – Bachelor’s Degree -0.544**
Voter Turnout 2008 -0.052**
Percent Improved Turnout 2008 -0.137**
Teen Pregnancy Rate -0.455**
Number of Physicians -0.330**
Crime Index -0.172**
Infant Mortality Rate -0.207**
Percent Uninsured -0.109**
Percent Poverty -0.475**
**significant at the 0.01 level
*significant at the 0.05 level
19 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
APPENDIX 6: Regression Results
Variables Results
Economic Model 1: Income per Capita Beta (Sig.)
- Broadband adoption
- Educational attainment (Bachelors)
- Unemployment rate
- Urban-rural designation
- Population density
-0.198 (<0.01)**
-0.591 (<0.01)**
-0.054 (0.322)
-0.042 (0.551)
-0.320 (<0.01)**
R2
-0.816
R2
Adj. -0.806
Economic Model 2: Median Home Value Beta (Sig.)
- Broadband adoption
- Population density
- Urban-rural designation
- Income per capita
- Poverty rate
-0.283 (<0.01)**
-0.055 (0.630)
-0.214 (0.023)*
-0.581 (<0.01)**
-0.149 (0.068)
R2
-0.672
R2
Adj. -0.654
Economic Model 3: Retail Sales Beta (Sig.)
- Broadband adoption
- Urban-rural designation
- Income per capita
- Poverty rate
- Unemployment rate
-0.117 (0.158)
-0.320 (<0.01)**
-0.730 (<0.01)**
-0.212 (0.013)*
-0.038 (0.585)
R2
-0.627
R2
Adj. -0.608
**significant at the 0.01 level
*significant at the 0.05 level
20 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Variables Results
Education Model 1: End-of-Grade Test Scores
Beta (Sig.)
- Broadband adoption
- Poverty rate
- Students per Internet computer
- Teachers with degrees
- Funds per student
- Population density
-0.027 (0.761)
-0.556 (<0.01)**
-0.067 (0.473)
-0.328 (<0.01)**
-0.002 (0.985)
-0.202 (0.017)*
R2
-0.522
R2
Adj. -0.490
Education Model 2: End-of-Course Test Scores
Beta (Sig.)
- Broadband adoption
- Poverty rate
- Students per Internet computer
- Teachers with degrees
- Funds per student
- Population Density
-0.000 (0.999)
-0.549 (<0.01)**
-0.000 (0.997)
-0.179 (0.065)
-0.066 (0.572)
-0.141 (0.152)*
R2 -0.345
R2
Adj. -0.301
**significant at the 0.01 level
*significant at the 0.05 level
21 | H i g h - S p e e d B r o a d b a n d I n t e r n e t i n N C
Variables Results
Social Model 1: Crime Index Beta (Sig.)
- Broadband adoption
- Income per capita
- Population density
- Dropout rate
-0.193 (0.061)
-0.454 (0.002)**
-0.784 (<0.01)**
-0.160 (0.056)
R2 -0.380
R2
Adj. -0.354
Social Model 2: Teen Pregnancy Rate
- Broadband adoption
- Income per capita
- Population Density
- Unemployment rate
- Urban-rural designation
-0.069 (0.497)
-0.678 (<0.01)**
-0.140 (0.341)
-0.129 (0.141)
-0.059 (0.626)
R2 -0.436
R2
Adj. -0.406
Social Model 3: Poverty Rate
- Broadband adoption
- Income per capita
- Population density
- Median home value
-0.043 (0.664)
-0.654 (<0.01)**
-0.295 (0.01)**
-0.224 (0.072)
R2 -0.483
R2
Adj. -0.461
**significant at the 0.01 level
*significant at the 0.05 level