LET’S MAKE A DEAL: PRICE SENSITIVITY AND OPTIMAL SUBSIDIES
AMONG BROADBAND NON-ADOPTERS
By
Dev Joshi, Chris McGovern, Raquel Noriega, Elizabeth Riesser, Hongqiang Sun, & John Walker1
August, 2012
ABSTRACT
Although nearly every American household has access to a broadband network, nearly one in three
American homes still do not subscribe to the service. For many households the cost of home broadband
service is the barrier that prevents them from subscribing. As the federal government joins with
broadband service providers and non-profit groups to determine the best way to overcome this cost
obstacle, key questions raised are how much are current non-adopters willing to pay for broadband
service at home, and, as a result, how large should effective price discounts or subsidies be in order to
overcome this barrier.
To that end, this study uses a rich dataset collected by Connected Nation in 2011 through Random Digit
Dial (RDD) telephone interviews with 15,082 adult heads of households who do not subscribe to home
broadband service across a heterogeneous selection of states. This survey research aims to better
understand demographic and socioeconomic characteristics of non-adopters and assess key barriers to
broadband adoption.
This study finds that a majority of non-adopters2 would not be willing to subscribe to home broadband
service at any price, indicating that there are other barriers to entry beyond affordability at play. Yet
price remains a key barrier to broadband adoption among many. Connected Nation’s data estimates that
39% of non-adopters report that they would subscribe to home broadband service if it were offered at a
price they consider acceptable. This study analyzes price sensitivity across non-adopters of various
demographic groups and finds that there are significant variations between different socioeconomic
groups.
Using a binary logistic regression, the study estimates the marginal impact of geographic and
socioeconomic factors on non-adopters’ price sensitivity. Results show that the marginal impact of
socioeconomic factors including race, ethnicity, age, rural/non-rural differences, the presence of children
in the household, educational attainment, and employment status all have significant marginal impacts
(positive or negative) on non-adopters’ willingness to subscribe to broadband at a price they consider
acceptable. By contrast, the impact of having a disability does not have a marginal effect on the
likelihood of a non-adopter being price sensitive.
In addition, this study employs a Van Westendorp Price Sensitivity analysis to measure how much nonadopters would be willing to pay to have broadband service at home. The goal of this research is to
inform the policy debate regarding consumers’ range of acceptable prices to promote universal adoption
1
The authors are members of Connected Nation, Inc., which has sponsored this study. For more information and contact
information about the authors see Appendix B. For more information about Connected Nation visit
www.connectednation.org.
2
In this study, a “non-adopter” is the head of a household that does not subscribe to home broadband service, defined as
Internet service at advertised speeds of at least 768 Kbps download.
Electronic copy available at: http://ssrn.com/abstract=2033415
of broadband technology. The study estimates an optimal target subsidy level or price discount of $25.30
per month to best promote home broadband adoption among price sensitive non-adopters. This same
test is applied to various demographic groups of non-adopters to examine price effects across different
demographic groups. Estimated optimal subsidies are relatively consistent across socioeconomic groups
that are characterized by lower rates of broadband adoption, such as low-income households,
minorities, people with disabilities, and the elderly. This result suggests that policies aimed to maximize
broadband adoption among price sensitive households that remain disconnected should focus on a
constant price subsidy or discount regardless of the socioeconomic group targeted.
By contrast, estimated optimal price subsidies across the jurisdictions studied are significant. In Alaska,
for example, a state showcasing the highest adoption rates across all studied, optimal price subsidies or
discounts are estimated at $40.19 per month. This result is partially driven by the fact that current retail
prices for broadband services are significantly higher in Alaska than in other jurisdictions. This result
suggests that a one-size-fits-all policy may not be optimal in some jurisdictions with divergent market or
socioeconomic conditions at play.
1. INTRODUCTION
Affordability is a key barrier to broadband adoption among many who have yet to embrace high-speed
Internet service in their homes. According to the Federal Communications Commission’s (FCC) National
Broadband Plan, in 2010, 35% of adults in the United States did not subscribe to broadband service at
home (Horrigan). Among these non-adopters, 36% said that prices must fall for them to switch to
broadband service, indicating that affordability is still a major issue for millions of Americans who remain
disconnected.
This affordability gap slows economic growth, reduces educational opportunities, and limits job creation
opportunities for Americans of any socioeconomic group. As more of these opportunities continue to
shift online, families and their children who remain disconnected from online resources will face an
increasing challenge to overcome poverty and become active and productive members of society.
Faced with this challenge, state and federal policy makers are debating what policies are appropriate to
address this affordability barrier. The Federal Communications Commission (FCC) is evaluating whether
to introduce a subsidy mechanism for home broadband subscriptions similar to the existing Low-Income
Lifeline program, which currently provides a subsidy to low-income families for the monthly subscription
of voice telephone service. Later this year, the FCC is scheduled to implement a Broadband Lifeline Pilot
program aimed to test various assumptions about how to structure such a program (FCC Lifeline).
Private sector stakeholders are also assessing these assumptions. Models such as Comcast’s Internet
Essentials, CenturyLink’s Internet Basics, Connect2Compete’s upcoming discount offerings, and
Connected Nation’s own Every Community Online program are all offering low-cost broadband
subscription service and computer equipment at various price points to vulnerable communities.
These initiatives share a degree of uncertainty regarding what the optimal price discount should be to
attract households that are not willing to subscribe to broadband at current market prices. In fact, the
FCC’s Broadband Lifeline Pilot program is specifically designed to test various assumptions regarding the
optimal subsidy amount per household (Connected Nation). How low must the monthly price of
broadband be in order to attract the largest number of non-adopters while not wasting taxpayer money
or Universal Service Fund contributions on households that would not subscribe to broadband at any
cost (for example, those who simply do not understand the benefits of broadband regardless of cost)?
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
Electronic copy available at: http://ssrn.com/abstract=2033415
2
For these programs to efficiently and sustainably invest public funds, they must estimate how many nonadopting households will be motivated to adopt broadband as a result of price subsidies, and perhaps
more importantly, how low monthly subscription fees must drop in order to overcome the price barrier.
In 2011, Connected Nation surveyed 15,082 adult heads of households who currently do not subscribe
to home broadband service. These non-adopter surveys were conducted between June 21 and August
17, 2011, across seven states including Alaska, Iowa, Michigan, Minnesota, Nevada, South Carolina and
Texas, and were funded as part of the State Broadband Initiative (SBI) federal grant program in those
states.
The SBI is a federal grant program administered by the National Telecommunications and Information
Administration at the US Department of Commerce with funds provided through the American Recovery
and Reinvestment Act of 2009 and the Broadband Data Improvement Act of 2008. Recognizing that
economic development, energy efficiency, and growth in the education and healthcare sectors rely on
individuals being able to access, adopt, and use broadband, the SBI is designed to be a comprehensive
program to facilitate broadband and information technology availability and adoption.
The goal of Connected Nation’s non-adopter surveys is to better understand the demographic
characteristics of digitally-lagging households, assess key barriers to broadband adoption, and analyze
technology-related purchasing decisions. Specifically, the surveys were designed to assess the degree to
which affordability is a key barrier to adoption and the price at which non-adopters would be willing to
subscribe to home broadband service. Based on these data, this study derives several price sensitivity
tests and estimates a range of efficient subsidies or price discounts to attract demand among these nonadopting households.
These non-adopter surveys offer a rich database of information regarding why households choose not to
have broadband and information technology in their home. The data indicate that 39% of non-adopting
respondents would be willing to subscribe to home broadband service if cost were not a factor,
suggesting that approximately 17 million American households would subscribe to broadband if it were
offered at a price they consider to be acceptable (United States Census Bureau, 2010 Census).3 But what
is the price at which they would be willing to subscribe to the service? And based on this, what range of
subsidies to incent adoption should policy makers consider?
This study addresses these questions using a Van Westendorp Price Sensitivity Analysis, which relates a
range of optimal prices that would maximize the number of non-adopting households who would
choose to subscribe to broadband at various price levels. The study also addresses price sensitivity and
optimal subsidy levels among non-adopters of different demographic groups including: race, ethnicity,
gender, geography, age, employment status, educational attainment, income level, disability status, and
the presence of children in the home.
2.LITERATUREREVIEW
In order to establish price perceptions and estimate acceptable price ranges, this study utilizes the Van
Westendorp Price Sensitivity analysis, a self-anchoring measurement system. Self-anchoring scales have
a long history in public opinion and psychosocial research (Batt and Katz). Within consumer research,
one of the first to develop the technique of having respondents establishing their own price ranges was
the Dutch economist Peter Van Westendorp.
3
For this study, non-adopters are defined as “price sensitive” if they responded “Yes” when asked “If you could subscribe to
home broadband service at a price you consider acceptable, would you do so?”
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
Electronic copy available at: http://ssrn.com/abstract=2033415
3
Before addressing the model used for the price sensitivity analysis, it is important to discuss the central
issue of why millions of residents are still not subscribing to broadband service. Various academic
institutions, research organizations, and government agencies have conducted research that focused on
the critical relation between broadband adoption and the cost associated with it.
According to the Federal Communications Commission’s (FCC) National Broadband Plan, in 2010, 35% of
adults in the United States did not subscribe to broadband service at home (FCC NBP). Furthermore, the
research shows that more than one-third of the non-adopters, or 36%, cited costs (including monthly
subscription, non-recurring costs, and equipment costs) as the biggest barrier to broadband adoption for
them, including 15% who cited the monthly cost of broadband as their main barrier to home broadband
adoption. According to the National Telecommunication and Information Administration (NTIA), 32% of
American households did not subscribe to broadband as of 2010, with 24% of those households
reporting that they did not subscribe due to affordability issues (NTIA 2011). Similarly, the NTIA’s 2010
Digital Nation report showed that among non-adopters, 53% of households with annual incomes below
$25,000 cited cost as the reason for not adopting broadband (NTIA 2010). Recent findings by the Pew
Internet and the American Life Project show that 34% of adults do not have broadband service in the
home, and 21% of those non-adopters do not subscribe due to cost (Pew).
One potential solution is the use of subsidies or targeted discounts, supported by the local or federal
government or by providers, to reduce the monthly price of broadband to attract more non-adopters
who could be convinced to subscribe at a cost below the current market value (Aron and Ingraham;
Hauge and Prieger). In order to implement such a strategy, we must address the question of how deep
the monthly subsidy should be. One method for answering this question is through the use of a Van
Westendorp Price Sensitivity Analysis.
The Van Westendorp Price Sensitivity model is used to determine a range of acceptable prices that
potential customers would be willing to pay for the good or service and, based on that, estimate an
optimal price point for a new or lagging product or service. This technique is often used during the new
product development phase to aid in price-setting, and has been used to determine optimal price ranges
for products such as e-reader devices (Epps) and software (Harmon), and services such as home Internet
service (Batt and Katz) and next-generation mobile broadband (Lee). Based on these studies, Connected
Nation also adopted the Van Westendorp Price Sensitivity model to determine the optimal price
discount that would entice the largest share of non-adopters to subscribe to home broadband service.
3.RESEARCHMETHODOLOGY
3.1. Data Collection
This study uses data from a random digit dial telephone survey of 15,082 adult heads of households
without broadband service at home (non-adopters) between June 21 and August 17, 2011 across seven
states (Alaska, Iowa, Michigan, Minnesota, Nevada, South Carolina, and Texas). This sample included
1,190 respondents who were contacted on their cell phones and 13,892 who were contacted via landline
phones. Multiple attempts were made to each working telephone number on different days of the week
and at different times of the day to increase the likelihood of contacting a potential respondent.
Interviews were conducted in English in every state; in addition, residents in Texas and Nevada were
offered the option to take the survey in Spanish. On average, after the respondent agreed to participate,
these surveys took approximately seven minutes to complete. The data from these surveys were
subsequently weighted by age, gender, state, and county of residence (rural or non-rural) using a rim
weighting process to account for minor variances between the sample and the population of nonLet’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
4
adopters. For this study, “rural” county or county equivalents are defined as those that are outside the
jurisdiction of any Metropolitan Statistical Area (MSA).
Out of the 15,082 non-adopters surveyed, 4,383 reported they would be willing to subscribe to home
broadband service at what they considered an acceptable price.4 These respondents were then asked a
series of four questions to determine what that acceptable price would constitute:
“At what monthly price would you consider a home broadband subscription to be ‘too expensive
to consider’? We’re looking for a price per month in dollars.”
“And at what monthly price would you consider a home broadband subscription to be ‘getting
expensive, but still worth the cost’? Here, we’re looking for a price that is less than [Response
from previous question].”
“Now, at what monthly price would you consider a home broadband subscription to be ‘a
bargain, definitely worth the money’? Here, we’re looking for a price that is less than [Response
from previous question].”
And
“And at what monthly price would you consider a home broadband subscription to be ‘so
inexpensive that you would question the quality of the service and not consider subscribing’?
Here, we’re looking for a price that is less than [Response from previous question].”
Based on the effective sample sizes, the margin of error for all heads of households without broadband =
+ 1.76% at a 95% level of confidence, while the margin of error for the 3,312 non-adopters who
answered all four price-related questions = +2.67% at the same level of confidence. As with any survey,
question wording and the practical challenges of data collection may introduce an element of error or
bias that is not reflected in these margins of error. The survey results were also reviewed by experts in
the fields of survey design and statistics in each state.
In addition, in 2012 Connect Puerto Rico conducted a random digit dial telephone survey of 2,400 adult
heads of households across Puerto Rico who do not subscribe to broadband service, including 1,200 who
were contacted on cell phones. At a 95% level of confidence, this sample provides a margin of error of
+2.16% among all respondents in Puerto Rico, while the margin of error for the 584 non-adopters who
answered all four price-related questions = +4.23 at the same level of confidence. However, due to this
survey being conducted at a different time interval and differences in broadband infrastructure and
economic conditions, the results from Puerto Rico are reported in this study separately, but are not
included as part of the Connected Nation average.
3.2. Data Analysis
First, survey responses were analyzed by cross-tabulation to explore whether demographic factors are
associated with the likelihood of a non-adopter’s willingness to subscribe at a price that s/he considers
acceptable. This includes factors such as race/ethnicity, educational attainment, age, annual household
income, employment status, county of residence (rural or non-rural), the presence of children in the
household, and disability status. All of these factors were self-reported by the respondent during the
course of the survey, and results were collected as categorical variables.
4
These respondents answered “yes” when asked “If you could subscribe to home broadband service at a price you consider
acceptable, would you do so?”
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
5
This cross-tabulated analysis provides insights into respondents’ willingness to pay for home broadband
service; however, since some demographic variables will be highly correlated, this analysis will not
quantify the marginal effect of each demographic factor on predicting whether a non-adopter would
subscribe to home broadband service at an acceptable price. To measure the marginal impact of these
same demographic factors on a respondent’s willingness to subscribe given a price reduction (i.e., to
answer “yes” when asked whether s/he would be willing to subscribe to home broadband service at a
price that is considered acceptable), this study uses a binary logistic regression. This regression measures
the probability of a home broadband non-adopter saying yes (value=1), s/he would be willing to
subscribe at a price that s/he considers acceptable, or no (value=0), s/he would not be willing to do so.
The model uses the form:
= β + log 1−
Where β is the intercept coefficient; is the coefficient for each individual independent variable, p is
the probability that a person would be willing to subscribe to broadband at a price s/he considers
acceptable, ( ) is the odds that a respondent would be willing to subscribe to home broadband service
at an acceptable price. Independent variables include annual household income, educational
attainment, age, race and ethnicity, county of residence, the presence of children at home, employment
status, and disability status.
When making a one-unit change in one of the independent variables, ceteris paribus, the log of odds
ratio of a person’s stated willingness to subscribe at a price s/he considers acceptable can be expressed
as follows:
log(
)
1 − = ( + 1 − ) = log(1 − )
So
)
(
)
(
= !
(1 − )
= !
(1 − )
Finally, the odds ratio of one’s self-reported willingness to subscribe to home broadband service at a
reduced price is equal to eβi when the independent variable increases or decreases one unit and holding
other independent variables constant. For binary or multi-level categorical independent variables, this
indicates how the odds of a respondent saying that s/he would be willing to subscribe at a price that
s/he considers acceptable changes based on that demographic factor, all other things held constant.
While the binary logistical analysis provides insight into which demographic factors can predict the
likelihood of subscribing to home broadband service as a result of price subsidies, it cannot be used to
determine the optimal price at which the largest number of non-adopters would subscribe to home
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
6
broadband service. To address this question, this study uses the Van Westendorp Price Sensitivity
analysis.
The Van Westendorp Price Sensitivity analysis is a technique to assess the willingness to pay for a good
or service among consumers who have not yet embraced a product or service. By asking a series of
questions to determine the value that respondents place on acquiring a good or service, as well as what
prices they feel would be too expensive to consider the product worthwhile, researchers are able to
derive a Van Westendorp “Optimal Price Point,” or VOPP. Although more complicated than simply asking
a person “How much would you be willing to pay for home broadband service,” Van Westendorp
analyses tend to provide a more accurate snapshot of what consumers are actually willing to pay for
goods or services (Market Strategies International).
Connected Nation’s 2011 non-adopter surveys asked broadband non-adopters a series of questions to
determine what value they would place on having home broadband service. The responses to those
questions were then used to assess the price point that would induce the optimal number of nonadopting households to embrace home broadband service. The value these respondents place on having
home broadband service would theoretically be less than a price they consider as “too expensive,” but
higher than a price they would consider so cheap that they would question the value of the service or
the integrity of the offer from the provider. Based on results from these questions the study determines
the broadband VOPP for non-adopting households. In addition, this model shows a Range of Acceptable
Prices (RAP) for policy makers to consider when developing a cost-reduction program to attract nonadopters; if the subsidized monthly price is higher than this RAP, many non-adopters would still not
subscribe because, even at the reduced prices, the service may still be beyond their reach or because
they are not price sensitive and face other barriers to broadband adoption (such as lack of digital literacy
or knowledge of the technology’s relevance upon their welfare). If broadband prices are subsidized so
much that they fall below the RAP, potential consumers may question the value of the service they are
being offered and excess tax-payer money would go to over-subsidize broadband adoption. Comparing
this VOPP and RAP with current average market prices for broadband services, we are able to estimate
the range of optimal subsidies across different demographic groups in order to most efficiently attract
the greatest number of current non-adopters.
Armed with these VOPP and RAP estimates and taking into account average retail market prices, the
study then estimates the optimal monthly subsidies or price discounts for a strategy aimed to overcome
price barriers affecting targeted, vulnerable populations. Optimal subsidies are estimated for the whole
non-adopter population as well as targeted demographic groups that remain disproportionately
disconnected from online resources. These optimal subsidy estimates address the key policy questions
currently under debate at the FCC and across private sector stakeholders working to close the digital
divide among demographic groups who have access to broadband networks but do not chose to
subscribe to the service.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
7
4.THEBROADBANDADOPTIONGAP
According to Connected Nation’s 2011 survey research across 7 states, 37%, or approximately 14 million
adults, do not subscribe to home broadband service (Table 1).5
TABLE 1.
PERCENTAGE OF NON-ADOPTERS BY JURISDICTION
Alaska
27%
Iowa
37%
Michigan
39%
Minnesota
28%
Nevada
33%
South Carolina
43%
Texas
38%
Seven State Average
37%
Puerto Rico
65%
Among these states, South Carolina had the
largest share of non-adopters at 43%, while
Alaska had the lowest share of non-adopters
with 27%. In Puerto Rico, a similar survey in
2012 (included here as a point of reference,
but not included in the Connected Nation
average) showed that 65% of adults still do not
subscribe to home broadband service.
Among those surveyed in 2011, 29% of adults
who do not subscribe to home broadband
service say cost is the main barrier to
broadband adoption (Table 2).
TABLE 2.
MAIN BARRIER TO HOME BROADBAND ADOPTION
Cost
Digital Literacy
Relevance
Availability
Access the Internet Someplace Else
Do Not Go Online Very Often from Home
Do Not Own a Computer, or Home Computer Does Not Work
Other
Refused/No Response
29%
19%
19%
8%
5%
3%
2%
9%
6%
This includes 18% of
non-adopters who
say the monthly cost
of broadband
service is too
expensive, 7% who
cite the cost of
purchasing a
computer, and 4%
who report that the
activation and
installation fees of
broadband service
are too expensive.
5.PRICESENSITIVITYFORHOMEBROADBANDSERVICEAMONGNON-ADOPTERS
While cost is a major barrier to broadband adoption, other factors contribute to the persistent digital
gap. Across the surveyed states, 73% of adults without home broadband service cite reasons other than
price for not subscribing. In fact, only 39% of surveyed adults who do not currently subscribe to home
broadband service say they would subscribe at a price that they consider acceptable. This suggests that
making broadband service more affordable could have significant impact upon home broadband
adoption. Extrapolating these results to the entire nation, if targeted price discounts were applied, an
estimated 17 million additional US households would subscribe to broadband.
5
Puerto Rico data comes from the 2012 Connect Puerto Rico Residential Technology Assessment and is included here as a point
of reference. This figure is not included in the average of other states surveyed.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
8
Reducing the monthly price of home broadband service, though, would not be a cure-all for the
broadband adoption gap. As such, it is important to learn not only how many non-adopters could be
convinced to subscribe at a reduced (subsidized) price, but what the subsidy range should be to induce
the optimal number of non-adopters to subscribe to the service.
5.1. Price Sensitivity for Home Broadband Service by Demographic
While the data show that cost reductions can attract new subscribers, social and economic factors have a
significant impact upon non-adopters’ willingness to pay for home broadband service. This section
examines the price sensitivity of non-adopters of different demographic groups shown by the literature
to be disproportionately disconnected from online resources.
Race/Ethnicity:
Minority non-adopters
TABLE 3.
are significantly more
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
likely to be price sensitive
BY RACE/ETHNICITY
(Table 3). Altogether, 48%
% of Non-Adopters Willing to Subscribe to
of minority non-adopters
Race/Ethnicity
Broadband at an Acceptable Price
said they would subscribe
Caucasian
35%
at a price they consider
African
American
45%
acceptable. This includes
Hispanic
50%
45% of African-American
non-adopters, 50% of
Other
43%
Hispanic non-adopters,
and 43% of non-adopters
who identified themselves as some other race or ethnicity (such as Asian, Native American, or Native
Alaskan). They are all significantly higher than 35% of Caucasian non-adopters who would be willing to
subscribe.
Gender:
There is no significant difference between male and female non-adopters in their willingness to
subscribe to broadband
TABLE 4.
service at a price they
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
consider acceptable; 39%
BY GENDER
of female non-adopters
% of Non-Adopters Willing to Subscribe to
would subscribe at a
Gender
Broadband at an Acceptable Price
price they consider
Male
40%
acceptable, compared to
Female
39%
40% of male nonadopters (Table 4).
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
9
Urban-Rural:
Non-adopters living in rural
TABLE 5.
counties or county
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
equivalents are slightly but
BY COUNTY (OR COUNTY EQUIVALENT) OF RESIDENCE
significantly more likely to say
% of Non-Adopters Willing to Subscribe to
County of Residence
they would be willing to
Broadband at an Acceptable Price
subscribe to home
Rural
42%
broadband service than nonNot
Rural
39%
adopters living in urban or
suburban portions of these
states (Table 5). This suggests that there is pent-up demand among non-adopters in rural areas where
broadband service may cost more than equivalent service in urban or suburban areas.
Households with Children:
Non-adopters with
TABLE 6.
children living at home
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
are nearly twice as likely
BY THE PRESENCE OF CHILDREN IN THE HOME
to report that they would
% of Non-Adopters Willing to Subscribe to
be willing to subscribe to
Presence of Children
Broadband at an Acceptable Price
home broadband service
Children at Home
59%
at a price they consider
No Children at Home
31%
acceptable (Table 6). This
shows that non-adopters
with children at home are
more price sensitive and more likely to be convinced to subscribe to home broadband service by lower
monthly prices than those with no children at home.
Age:
Younger non-adopters
TABLE 7.
tend to be more price
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE BY AGE
sensitive than older non% of Non-Adopters Willing to Subscribe to
adopters surveyed (Table
Age
Broadband at an Acceptable Price
7). More than one-half of
non-adopters between the
Age 18-34
57%
ages of 18 and 34 report
Age 35-54
47%
that they would be willing
Age 55-69
30%
to subscribe to home
Age 70 or Older
12%
broadband service if it
were offered at a price
they considered
acceptable, while non-adopters in every older age bracket were significantly less likely to say the same
thing. Among the oldest demographic group, non-adopters age 70 or older, only 12% of respondents
were willing to subscribe to home broadband service if it were offered at a price they considered
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
10
acceptable, indicating how barriers other than price are driving their decision not to subscribe to
broadband service at home.
Employment Status:
More than one-half of the employed non-adopters (51%) would subscribe to broadband service at a
price they consider acceptable, significantly higher than adults in any other self-reported employment
category (Table 8).
TABLE 8.
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
BY EMPLOYMENT STATUS
Employment Status
Employed (Full-Time or Part-Time)
Retired
Disabled, Unable to Work
Unemployed
% of Non-Adopters Willing to Subscribe to
Broadband at an Acceptable Price
51%
17%
39%
45%
Retirees are the least likely to report that they would subscribe at an acceptable price, echoing the trend
among all older adults.
Educational Attainment:
Non-adopters with at least some college education are more price sensitive than non-adopters who have
a high school diploma or less (Table 9).
TABLE 9.
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
BY EDUCATIONAL ATTAINMENT
% of Non-Adopters Willing to Subscribe to Broadband
Educational Attainment
at an Acceptable Price
Less than High School Diploma
35%
High School Diploma
38%
Some College
47%
College or Advanced Degree
42%
Breaking these data down further, non-adopters who report they had attended some college but did not
earn a college degree are the most likely to say they would respond to price subsidies. This can indicate
that while non-adopters with no college education may not recognize the value of having home
broadband service, those with college degrees may be stymied by barriers other than cost.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
11
Income:
More than two out of five non-adopters with annual incomes of less than $25,000 (41%) would
subscribe to home broadband service if the price were acceptable, while 49% of non-adopting
households earning $50,000 or more would do so (Table 10).
TABLE 10.
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
BY ANNUAL HOUSEHOLD INCOME
% of Non-Adopters Willing to Subscribe to Broadband
Annual Household Income
at an Acceptable Price
Less than $25,000
41%
$25,000-$49,999
45%
$50,000 or More
49%
Refused/No Response Given
23%
This result appears to be counterintuitive, since higher income households would in principle be more
able to afford broadband service in the home and, hence, be less price sensitive. This seemingly
counterintuitive result may be driven by factors other than price considerations, though. A greater
proportion of lower income households may be faced with barriers to broadband adoption other than
price, including lack of digital literacy and understanding of the relevance of online resources to their
welfare, lack of computer in the home or other. Price for such households may be a factor to broadband
adoption, but not the main factor. For higher income households, likely more digitally savvy on average,
price becomes the main barrier to broadband adoption.
Disability Status:
One out of three non-adopters
TABLE 11.
who report at least one
PRICE SENSITIVITY FOR HOME BROADBAND SERVICE
disability said that they would
BY DISABILITY STATUS
be willing to subscribe to home
broadband service at a price
% of Non-Adopters Willing to Subscribe
Disability Status
they consider acceptable (Table
to Broadband at an Acceptable Price
11).6 This is significantly lower
Adults with Disabilities
33%
than the 44% of non-adopters
Adults without Disabilities
44%
who did not report having any
disability. This is partially due to
the increase in disabilities reported by older respondents who have already been shown to be less likely
to report a willingness to be swayed to subscribe to home broadband service by price incentives.
6
Non-adopters with disabilities are defined as those who responded in the affirmative when asked “Do you have any long-term
physical, mental, or emotional conditions that make it difficult to do any of the following tasks? (Walking or climbing stairs;
concentrating, remembering, or making decisions; visiting a doctor’s office or shopping by yourself; dressing or bathing),” or
answered “yes” when asked “Are you blind or have serious difficulty seeing, even when wearing glasses” and/or “Are you deaf
or have serious difficulty hearing?”
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
12
6.THEMARGINALEFFECTOFDEMOGRAPHICFACTORSONNON-ADOPTERS’PRICE
SENSITIVITYFORHOMEBROADBANDSERVICE
In addition to cross-tabulations, this study explores the marginal effect of various demographic factors
on a respondent’s willingness to subscribe if offered home broadband service at an acceptable price.
This regression measures the marginal impact of demographic factors on a respondent’s willingness to
subscribe at a price they consider acceptable (i.e., to answer “yes” when asked whether s/he would be
willing to subscribe to home broadband service at a price that is considered acceptable). The dependent
variable for this regression is the likelihood of a home broadband non-adopter saying yes (value=1), s/he
would be willing to subscribe at a price that s/he considers acceptable, or no (value=0), s/he would not
be willing to do so. Results are shown in Table 12.
TABLE 12.
MARGINAL IMPACT OF DEMOGRAPHIC FACTORS ON NON-ADOPTERS’ PRICE SENSITIVITY
BINARY LOGIT MODEL ESTIMATES
B
Race (ref: Caucasian)
African American*
Asian
Hispanic*
Other
Age (ref: Age 18-34)
Age 35-54*
Age 55-69*
Age 70 or older*
Presence of children
(ref: Households with no children at home)*
Educational attainment (ref: Advanced or
Professional Degree earned)
Less than high school*
High school graduate*
Some college
College graduate
Annual household income
(ref: income $75,000 or more)
Less than $25,000*
$25,000 to $49,999
$50,000 to $74,99 9
Employment
(ref: employed full-time or part-time)
Retired*
Not working due to disability
Unemployed
Other
Rural status (ref: Non-Rural)*
Disability status (ref: no disability)
Constant
S.E.
Wald
Df
Sig.
Odds
Ratio
.322
-.212
.176
.202
.073
.230
.077
.114
24.353
19.270
.854
5.192
3.118
4
1
1
1
1
.000
.000
.356
.023
.077
1.380
.809
1.193
1.224
-.414
-.777
-1.876
.075
.082
.103
375.443
30.799
88.935
333.673
3
1
1
1
.000
.000
.000
.000
.661
.460
.153
.488
.061
63.215
1
.000
1.630
101.457
4
.000
40.695
26.546
1.448
1.315
1
1
1
1
.000
.000
.229
.252
20.050
3
.000
6.681
.088
.036
1
1
1
.010
.767
.849
47.316
4
.000
-.743
-.527
-.125
-.125
-.215
.024
.017
.116
.102
.104
.109
.083
.080
.091
.476
.590
.883
.883
.806
1.024
1.017
-.419
.081
-.047
-.016
.070
.088
.088
.106
36.085
.851
.290
.023
1
1
1
1
.000
.356
.590
.879
.657
1.084
.954
.984
.179
.048
13.747
1
.000
1.196
.081
.088
.851
1
.356
1.084
-.106
.066
2.585
1
.108
.900
*Independent variables denoted with a (*) and bold font are significantly different from the reference category at a 95%
level of confidence.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
13
The model is statistically significant at the 0.05 level. Race and ethnicity, age, the presence of children at
home, educational attainment, annual household income, employment status, and rural/non-rural
differences all have statistically significant marginal effects on whether a respondent voices a willingness
to subscribe to home broadband service at a price they consider acceptable, ceteris paribus. On the
other hand, disability status does not have a statistically significant impact on non-adopters’ willingness
to subscribe to broadband service at a price they consider acceptable.
In terms of race and ethnicity, all other things being equal, African American non-adopters are 38% more
likely than Caucasians to report that they would be willing to subscribe to home broadband service at a
price they consider acceptable. Hispanic non-adopters, ceteris paribus, similarly are 19% more likely than
Caucasians to be price sensitive. Asians and other minority groups are not significantly different from
Caucasians in their price sensitivity.
The odds that a non-adopter would be willing to subscribe to broadband at a price they consider
acceptable decreases among respondents in older age brackets. Compared to non-adopters age 18-34,
those age 35-54 are 34% less likely, those age 55-69 are 54% less likely, and non-adopters age 70 or older
are 85% less likely to be price sensitive. This suggests that among older non-adopters, other barriers
such as a lack of digital literacy skills or a lack of interest in broadband and online resources are more
influential than price in their decision on whether or not to subscribe to home broadband service.
Non-adopting households with children are 63% more likely than non-adopting households without
children to report that they would subscribe if offered broadband at an acceptable price. This suggests
that these non-adopter households with children tend to be more cognizant of the benefits of having
home broadband service and that their purchasing decisions are relatively more constrained by price
effects.
Along lines of educational attainment, there is no significant difference between non-adopters who have
earned advanced or professional degrees (the reference group in the regression) and those who have
earned college degrees or those who have attended college but did not earn a degree. On the other
hand, non-adopters whose highest educational attainment is a high school degree or less are
significantly less likely to report that they would be willing to subscribe at an acceptable price, all other
things held constant. Once again, other factors to broadband adoption appear to be more prevalent
amongst this latter group.
In terms of willingness to subscribe to home broadband service at an acceptable price, there is no
significant difference between non-adopting households with annual incomes of $25,000-$49,999, those
with annual incomes of $50,000-$74,999, and those with annual incomes of $75,000 or more, all other
factors held constant. By contrast, non-adopting households that report annual incomes below $25,000
are significantly less likely than those in the highest income bracket ($75,000 or more) to report that
they would be willing to subscribe to home broadband service at a price they consider acceptable. In
other words, the lowest income non-adopting households appear to be less price sensitive than the
higher income non-adopting households. This seemingly counterintuitive result suggests that other
barriers to adoption, such as relevance of the broadband service, the lack of a home computer, or a lack
of available broadband service (all of which are cited significantly more often among low-income nonadopters than those in the higher income brackets) may be at play.
Employed non-adopters are significantly more likely to be price sensitive than retired non-adopters, all
other things held constant. These two groups of non-adopters, though, are the only ones who show any
statistically significant difference along the lines of employment status.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
14
The difference between rural and non-rural non-adopters is statistically significant. Rural non-adopters
are 19% more likely to claim they would be willing to subscribe at an acceptable price than non-adopters
living in urban or suburban areas, all other factors held constant.
Conversely, disability status does not have a statistically significant impact on non-adopters in this model,
all other things held constant. This is in spite of the cross-tabulated results in section 5.1 showing that
non-adopters with disabilities are less likely to say they would subscribe to home broadband service at
an acceptable price.
These findings suggest that programs targeting non-adopters that subsidize the monthly cost of home
broadband service for households with children, rural households, African Americans, and Hispanics
should expect higher participation rates than monthly broadband subsidies targeting the elderly,
retirees, and those who have earned a high school degree or less. Further, according to these results,
non-adopters in with annual income below $25,000 would be less likely to respond to subsidies based on
monthly broadband costs, all other things held constant.
7.OPTIMALPRICEPOINTSAMONGNON-ADOPTERSANDOPTIMALSUBSIDIES
The Van Westendorp Optimal Price Point (VOPP) among non-adopters who said they would be willing to
subscribe at a price they considered acceptable is $21.00 per month (Figure 1).
FIGURE 1. VAN WESTENDORP OPTIMAL PRICE POINT FOR BROADBAND
AMONG NON-ADOPTERS
Too Expensive
Inexpensive
100%
90%
Expensive
Too Inexpensive
80%
70%
VOPP=$21.00
60%
50%
40%
30%
20%
10%
0%
$0
$5
$10
$15
$20
$25
$30
$35
$40
$45
$50
$55
$60
$65
RAP = $19.00-$29.00
To frame the best price for a low-cost broadband offering, it is also important to consider the Range of
Acceptable Prices (RAP). Among non-adopting households surveyed by Connected Nation in 2011, the
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
15
range of acceptable price is as low as $19.00 and as high as $29.00. This suggests that a program
designed to attract the most new subscribers should consider a price within the ten dollar difference of
the lowest acceptable price and the highest acceptable price.
By comparison, according to surveys conducted by Connected Nation, the average broadband subscriber
in these states pays $46.30 per month, more than twice the VOPP for non-adopters. This suggests that
an optimal, sustainable per month subsidy ranges from $17.30 and $27.30 per month, for an optimal
subsidy equaling $25.30 per month to attract the most non-adopters who could be convinced to
subscribe by lower monthly costs.
7.1. Optimal Subsidies by State or Territory
Across the seven states surveyed, 39% of non-adopters report that they would be willing to subscribe to
broadband if the service was offered at a price they considered reasonable (Table 13). 7
Texas non-adopters are significantly more likely to report
that they would subscribe to home broadband service at
a price they consider acceptable. Conversely, Minnesota
and Nevada have the smallest share of non-adopters
who are price sensitive. Puerto Rico survey results from
2012 are presented here as a point of comparison to the
2011 surveys in other states, but are not included as part
of the Connected Nation average percentages.
Using the average reported monthly price for home
broadband service in each state, this study calculated
optimal subsidies based on the VOPP that non-adopters
in each state are willing to pay for broadband. Based on
the results of the Van Westendorp analysis in each of
these seven states; the optimal monthly subsidy ranges
from $40.19 in Alaska to $23.00 in Texas (Table 14). 8
TABLE 13.
PERCENT OF NON-ADOPTERS
WILLING TO SUBSCRIBE TO BROADBAND AT AN
ACCEPTABLE PRICE
Alaska
44%
Iowa
34%
Michigan
33%
Minnesota
30%
Nevada
30%
South Carolina
37%
Texas
45%
Seven State Average
39%
Puerto Rico
32%
7
Puerto Rico data comes from the 2012 Connect Puerto Rico Residential Technology Assessment and is included here as a point
of reference. This figure is not included in the average of other states surveyed.
8
Ibid.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
16
TABLE 14.
OPTIMAL BROADBAND SUBSIDIES – BY JURISDICTION
State/Territory
Alaska
Iowa
Michigan
Minnesota
Nevada
South Carolina
Texas
Seven State Average
Puerto Rico
Average Reported Monthly
Broadband Price among
Current Broadband
Subscribers
$66.19
$47.90
$46.90
$49.46
$48.36
$47.93
$44.00
$46.30
$47.33
Monthly VOPP for
Non-Adopters
Optimal Monthly
Subsidy or Price
Discount
$26.00
$20.00
$21.00
$20.00
$21.00
$22.00
$21.00
$21.00
$40.19
$27.90
$25.90
$29.46
$27.36
$25.93
$23.00
$25.30
$26.00
$21.33
In Alaska, with the cost of living significantly higher than in other states, it is not surprising that the
average cost for home broadband service is also higher than in other states. On average, Alaskan home
broadband subscribers pay $66.19 per month for their home broadband service. Interestingly, the
estimated VOPP among non-subscribers in Alaska (estimated at $26.00 per month) is also higher than
across all other jurisdictions, except for Puerto Rico. As a result, at an estimated $40.19 per month, the
optimal target subsidy or price discount is significantly higher in Alaska than in any other jurisdiction
surveyed.
This result suggests that there is no “one size fits all” effective subsidy program. Current market
conditions, as well as geographic, income, and demographic make-up of a given area must be taken into
consideration to design a successful and efficient subsidy program.
7.2. Optimal Subsidies by Demographic
Based on the average monthly price of $46.30 currently paid for broadband service in these seven states
and the VOPP calculated for non-adopters who identified as belonging to each of the demographic
groups, Table 15 presents the optimal subsidy price that would convince the largest number of nonadopters in each demographic category to subscribe.
Results show that most demographic differences have little effect on estimated optimal price points and,
hence, estimated optimal monthly subsidies (Table 15). The optimal estimated subsidy across the
selected demographic groups shows little variance from the $25.30 average for all non-adopting
households, with one exception: race and ethnicity. Demographic effects are detailed below:
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
17
Household Income:
Household income has no impact upon the
estimated VOPP and, hence, no impact
upon the derived optimal subsidies. The
optimal subsidy for non-adopters living in
households with annual incomes below
$25,000 is the same as that among
households earning $25,000 or more a
year at $25.30. This optimal subsidy
coincides with the average optimal subsidy
for all non-adopters. This indicates that
while higher income non-adopting
households are more likely to be price
sensitive (see section 5 above), optimal
target retail prices are unaffected by
income level.
Age:
TABLE 15.
OPTIMAL MONTHLY BROADBAND SUBSIDIES FOR NON-ADOPTERS
BY DEMOGRAPHIC GROUP
Seven State Average
Income less than $25,000
Income $25,000 or more
Age 18 to 34
Age 35 to 54
Age 55 or Older
College Education
No College Education
Employed
Not Employed
Caucasian
Hispanic
Black, or African American
Households with Children
Adults with Disabilities
$25.30
$25.30
$25.30
$24.30
$24.30
$26.30
$26.30
$25.30
$25.30
$26.30
$26.30
$22.30
$25.30
$25.30
$26.30
The optimal subsidy for non-adopters age
55 or older is $26.30 per month. This is
slightly higher than the optimal price for non-adopters age 18-34 as well as those age 35-54, for whom
$24.30 is the optimal monthly subsidy. This suggests that any price subsidies that target older nonadopters will need to be more aggressive, as older non-adopters are less price sensitive than their
younger counterparts.
Education:
The optimal subsidy point for non-adopting adults with college degrees, estimated at $26.30 per month,
is slightly higher than for those without a college education, estimated at $25.30. This suggests that nonadopters with college educations who could be convinced to subscribe by price subsidies are willing to
pay only slightly more than those who have not attended a college or university.
Employment:
Employment status also shows little impact on the optimal subsidy amount for non-adopters. The
optimal subsidy among employed non-adopters is $25.30, compared to $26.30 for non-adopters who are
not employed. This suggests that employed residents have a greater recognition of the value of home
broadband service and require less of a financial incentive to adopt home broadband service.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
18
Race and Ethnicity:
The demographic characteristic that shows the greatest impact on the optimal monthly subsidy is along
the lines of race and ethnicity. Optimal monthly subsidies for Caucasian non-adopters are estimated at
$26.30, compared to an estimated subsidy of $25.30 for African Americans and $22.30 for Hispanics.
Hispanic and African American non-adopters would require smaller price incentives than Caucasian nonadopters to sway them to subscribe to home broadband service. This suggests that there is pent-up
demand and a relatively greater awareness of the benefits of broadband across these minority cohorts.
Presence of Children:
The optimal monthly subsidy for households with children is $25.30, the same as the average across all
non-adopting households in the seven states. By contrast, results from section 5 above indicate that nonadopting households with children are significantly more likely to be price sensitive than those without
children. Thus, even though households with children are more willing to say they would subscribe at a
price they consider acceptable, that acceptable price does not differ between non-adopters with
children and those without.
Disability Status:
The optimal monthly subsidy for respondents with disabilities is $26.30, which is slightly higher than the
seven state average. This suggests that non-adopting adults with disabilities, many of whom are also
older, will need greater financial encouragement to be convinced to subscribe to home broadband
service.
8. CONCLUSIONS
Across a heterogeneous selection of seven states, nearly two out of five broadband non-adopters report
that they would be willing to subscribe to home broadband service if it were made available to them at a
price they considered acceptable. Extrapolating this estimate to the entire nation, price is a key barrier
to adoption for nearly 17 million American adults. This indicates that targeted policies aimed to lower
the effective retail price points among non-adopters could impact millions of households who would be
swayed by lower prices to get connected to home broadband service.
Yet in order for any such subsidy program to be effective, it must target the right audience and be
prudent in its subsidy spending - any price incentives provided through state, federal, or private subsidy
mechanisms must be designed to ensure the greatest impact across the targeted population and result
in sustainable broadband adoption. In particular, any governmental subsidies, such as those currently
under examination by the FCC under the Broadband Lifeline Pilot program, must have a target price
range so the price-reduction programs will attract the most new subscribers while remaining careful not
to over-spend financial resources.
This study shows that programs that reduce the price of home broadband service for the targeted nonadopter population to $21.00 per month would have the greatest impact across non-adopters across the
overall population studied. Given current market prices, this implies an optimal subsidy equaling $25.30
per month. However, results change significantly across the various states and territories studied.
Alaskan non-adopters report an optimal target price for non-adopters of $26.00 per month, implying an
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
19
optimal subsidy of $40.19 per month. This result is partially driven by the fact that current retail prices
for broadband services in Alaska are significantly higher than in other jurisdictions. By contrast, optimal
estimated per month subsidies are $23.00 in Texas and $21.33 in Puerto Rico. These results suggest that
a one-size-fits-all policy may not be optimal in jurisdictions with divergent market or socioeconomic
conditions at play.
By contrast to these regional variations, optimal subsidy estimates across different demographic groups
analyzed, including income, age, educational attainment, employment status, race and ethnicity, the
presence of children in the home, and disability status, are very consistent, with optimal estimated
subsidies ranging from $22.30 to $26.30 per month. This suggests that policies aimed to maximize
broadband adoption among price sensitive households that remain disconnected should focus on a
constant price subsidy or discount regardless of the socioeconomic group targeted.
Subsidy programs that target more price sensitive groups, such as those who are younger, those who
have children at home, non-adopters with college educations, African Americans, and Hispanics, have
the potential to be relatively successful, though even within these demographic groups, not every
household can be swayed by price alone. As a result, price subsidies need to be paired with digital
literacy and local awareness campaigns that can address the myriad of issues that currently prevent
many residents from subscribing to broadband.
Overall, programs aimed to address the affordability barrier to broadband adoption through price
subsidies alone will not close the broadband adoption gap; other issues such as digital literacy and
perceptions of broadband service’s value must also be addressed. For many non-adopters, though,
strategies aimed to reduce the retail price of home broadband service through subsidies or other
mechanisms may provide a key incentive to bring broadband into their homes.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
20
REFERENCES
Aron, Debra J. and Allan T. Ingraham, “The Effects of Legacy Pricing Regulation on Adoption of
Broadband Service in the United States.” (April 1, 2012). <http://ssrn.com/abstract=2032797>
Batt, C., and J. Katz. "Consumer Spending Behavior and Telecommunications Services."
Telecommunications Policy. 22.1 (1998): 23-46.
Connected Nation. “Consumer Broadband Adoption Trends: 2011 Residential Technology Assessment.”
<http://www.connectednation.org/survey-results/residential>
Lee, Paul. "Addicted to Connectivity: Perspectives on the Global Mobile Consumer, 2011." Deloitte
Global (2011). Retrieved August 15, 2012. <http://www.deloitte.com/assets/Dcom-Global/Local
Assets/Documents/TMT/9314A_Mobile_Reports_sm5.pdf>
Epps, Sarah Rotman. "New Forrester Report: The eReader Price Squeeze." Forrester (September 2, 2009).
Retrieved August 15, 2012. <http://blogs.forrester.com/sarah_rotman_epps/09-09-02new_forrester_report_ereader_price_squeeze>
"FCC Solicits Applications to Fund Low-Income Broadband Adoption Pilots." Connected Nation (May 8,
2012). Retrieved August 15, 2012.
<http://www.connectednation.org/sites/default/files/bb_pp/lifeline_pilot_pn_brief_5_8_2012_
final1.pdf>
Federal Communications Commission. Report and Order and Further Notice of Proposed Rulemaking:
Lifeline and Link Up Reform and Modernization. WC Docket No. 11-42 (February 6, 2012).
<http://www.fcc.gov/document/fcc-reforms-modernizes-lifeline-program-low-incomeamericans-0>
Hauge, Janice A. and James E. Prieger, “Demand-Side Programs to Stimulate Adoption of Broadband:
What Works?” (October 14, 2009). <http://dx.doi.org/10.2139/ssrn.1492342>
Harmon, R.R., Unni, R.A., and R. Timothy. “Price Sensitivity Measurement and New Product Pricing: A
Cognitive Response Approach.” Portland International Center for the Management of
Engineering and Technology Proceedings: Management of Converging Technologies. Ed.
Kocaoglu, Dundar F., Anderson, Timothy R., and Tugrul U. Daim (August, 2007).
Horrigan, John. “Broadband Adoption and Use in America (OBI Working Paper #1).” Federal
Communications Commission (2010).
<http://online.wsj.com/public/resources/documents/FCCSurvey.pdf>
Market Strategies International. “Van Westendorp Price Analysis: an Overview.” (August, 2005).
<http://www.marketstrategies.com/user_area/content_media/Van%20Westendorp%20Price%2
0Analysis.pdf>
National Telecommunications and Information Administration. “Exploring the Digital Nation: Computer
and Internet Use at Home.” (November, 2011).
<http://www.esa.doc.gov/sites/default/files/reports/documents/exploringthedigitalnationcomputerandinternetuseathome.pdf>
National Telecommunications and Information Administration. “Exploring the Digital Nation: Home
Broadband Internet Adoption in the United State.” (November, 2010).
<http://www.ntia.doc.gov/files/ntia/publications/esa_ntia_us_broadband_adoption_report_110
82010_1.pdf>
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
21
Pew Internet and American Life Project. “Trend Data (Adults).” Retrieved on August 15, 2012.
<http://pewinternet.org/Trend-Data-(Adults)/Home-Broadband-Adoption.aspx>
United States Census Bureau. "SF1 File." 2010 United States Census American FactFinder. August 1, 2012.
<http://factfinder2.census.gov/faces/nav/jsf/pages/searchresults.xhtml?ref=geo&refresh=t>
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
22
APPENDIX A:
Select questions and sample sizes
2011 Connected Nation Non-Adopter Surveys
State/Territories (Residents without Home Broadband Service)
n=
Respondents
Alaska
1,751
Iowa
2,400
Michigan
2,400
Minnesota
1,900
Nevada
1,830
South Carolina
2,401
Texas
2,400
Seven State Total
15,082
Which one of these is the main reason why you do not subscribe to home
broadband service?
n=
Respondents
Affordability
3,909
Digital literacy
3,109
Relevance
4,046
Availability
1,171
Do not go online very often from home
463
Can get access somewhere else
652
I don’t own a computer or my computer doesn’t work
103
Other
427
Don’t know/Refused
If you could subscribe to home broadband service at a price you consider
acceptable, would you do so?
Yes
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
1,202
n=
Respondents
4,383
23
Samples for Cross-Tabulation and Logit Regression Analyses:
Demographic Groups
Seven State Total
Race and Ethnicity
Caucasian
African American
Hispanic
Other
Gender
Male
Female
County (or Equivalent) of Residence
Rural
Non-Rural
Presence of Children
Households with Children
Households without Children
Age
Age 18 to 34
Age 35 to 54
Age 55 to 69
Age 70 or older
Employment
Employed Full-Time or Part-time
Retired
Disabled, and Not Employed
Unemployed
Education
Less than High School Diploma
High School Graduate
Some College
College Graduate or Advanced Degree
Annual Household Income
Less than $25,000
$25,000 to $49,999
$50,000 or more
Disability Status
Adults with Disabilities
Adults with No Disabilities
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
n=
Respondents
15,082
10,396
1,494
1,449
911
6,134
8,948
5,429
9,653
2,571
12,271
1,641
3,502
4,528
5,405
5,031
6,582
1,312
1,017
2,364
5,631
3,516
2,971
5,076
3,614
2,674
4,760
8,958
24
Samples for Van Westendorp Analysis9
Seven State Total
Race/Ethnicity
Caucasian
African American
Hispanic
Gender
Male
Female
County (or Equivalent) of Residence
Rural
Non-rural
Presence of Children
With Children
Without Children
Age
Age 18 to 34
Age 35 to 54
Age 55 to 69
Age 70 or older
Employment
Employed Full-Time or Part-time
Retired
Disabled, and Not Employed
Unemployed
Education
Less than High School
High School Graduate
Some College
College Graduate or Advanced Degree
Income
Less than $25,000
$25,000 to $49,999
$50,000 or more
Disability Status
Adults with Disabilities
Adults with no Disabilities
n=
Respondents
3,312
2,189
410
455
1,425
1,887
1,225
2,087
1,124
2,175
776
1,276
971
286
1,835
606
330
296
396
1,097
963
823
1,072
996
876
823
2,366
9
For use in the Van Westendorp analysis, respondents had to provide answers for all four of the following questions:
“At what monthly price would you consider a home broadband subscription to be 'too expensive to consider'?”
“At what monthly price would you consider a home broadband subscription to be 'getting expensive, but still worth the cost’?”
“At what monthly price would you consider a home broadband subscription to be 'a bargain, definitely worth the money’?” and
“At what monthly price would you consider a home broadband subscription to be 'so inexpensive that you would question the
quality of the service and not consider subscribing’?”
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
25
2012 Connect Puerto Rico Non-Adopter Assessment
Island-wide Sample
Non-Adopters in Puerto Rico
Which one of these is the main reason why you do not subscribe to home
broadband service?
n=
Respondents
2,400
n=
Respondents
Affordability
695
Digital literacy
361
Relevance
555
Other
599
Availability
152
Don’t know/Refused
38
If you could subscribe to home broadband service at a price you consider
acceptable, would you do so?
Yes
Sample for Van Westendorp Analysis
Island-wide
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
n=
Respondents
590
n=
Respondents
584
26
APPENDIX B: ABOUT THE AUTHORS
Dev Joshi is a Research Analyst with Connected Nation, and before joining the Connected Nation team
he served as a Market Analyst for PearlDiver Technologies Inc., in Fort Wayne, Indiana. Mr. Joshi holds a
Masters of Business Administration from Indiana University and a Bachelor of Arts in Accounting from
Simpson College and he can be reached at [email protected]
Chris McGovern is Connected Nation’s Manager for Research Development. Prior to this role he served
as Research Analyst for ConnectKentucky, Senior Research Analyst for Connected Nation, and Research
Assistant in Murray State University’s Department of Economics and Finance. Mr. McGovern received his
Bachelor of Arts degree in political science from the University of Illinois at Chicago, and his Master of
Science in economics from Murray State University. Chris McGovern can be contacted at
[email protected].
Raquel Noriega is Connected Nation’s Director of Public Policy. Prior to joining Connected Nation, Ms.
Noriega was an economic consultant in Washington, DC, Madrid, Spain, and London, UK, and has
provided consulting services to telecommunications and media firms in the USA, Europe, and Latin
America. Ms. Noriega received her Master in Arts and Sciences for Economics from Northwestern
University. She is a graduate of the University of Michigan in Ann Arbor with a degree in economics.
Raquel Noriega can be contacted at [email protected].
Elizabeth Riesser is a Research Analyst for Connected Nation. Before joining Connected Nation, Elizabeth
worked as an Education Coordinator for the University Of Louisville Office Of Medical Education, a
Graduate Research Assistant for the Kentucky State Data Center, and as a Neighborhood Planner in
Columbus, Ohio. She holds a Bachelor of Arts degree in history and a Masters of City and Regional
Planning from The Ohio State University. Elizabeth Riesser can be contacted at
[email protected].
Hongqiang Sun is a Research Analyst for Connected Nation, Inc. Before joining Connected Nation,
Hongqiang was a Statistician at the University of Illinois in the Department of Crop Science. Mr. Sun
holds a Bachelor of Science in Chemical Engineering from Liaocheng Normal University, a Master of
Science in Applied Chemistry from Beijing Institute of Technology, a Master of Science in Statistics from
the University of Illinois at Urbana-Champaign and is a Ph.D. candidate in Chemical Engineering from the
University of Illinois at Chicago. Hongqiang Sun can be contacted at [email protected].
John Walker is a Research & GIS Analyst for Connected Nation. Mr. Walker previously worked in
Connected Nation’s GIS Services Division where he used GIS technology to develop statewide broadband
inventory maps and analyses associated with broadband availability and deployment. Mr. Walker holds a
Bachelor of Science from the University of Louisiana at Monroe and a Master of Science from Western
Kentucky University. John Walker can be contacted at [email protected].
Further information about Connected Nation and its research can be found at
www.connectednation.org.
Let’s Make a Deal: Price Sensitivity and Optimal Subsidies among Broadband Non-Adopters
27