community access to communications: 10 questions – provisional answers michael l. best...
TRANSCRIPT
Community Access to Communications:10 Questions – Provisional Answers
Michael L. Best
26 August 2004
Ten Questions
1. Typology2. Technologies3. Public policies4. Business models5. Internal capacity6. External capacity7. Sustainability8. Equity9. Evaluation10. Holism
Ten Questions
1. Typology2. Technologies3. Public policies4. Business models5. Internal capacity6. External capacity7. Sustainability8. Equity9. Evaluation10. Holism
Question 3: Public Policies
To truly realize broad community access to communications requires a range of supportive public policies.
Why Regulate?
Positive network externalities (the benefits a new consumer accrues from connecting - private benefits - are less than the total benefits to society.)
Merit goods Political and regional development objectives
Universal Service Obligations
Goals (might) include:
Geographic component (high-cost areas)
Economic component (low-income areas)
Democratization of provision
Key is find a basket of policies that are
non-discriminatory, competitively
neutral, and inexpensive (efficient)
Universal Service Provision in India
RSP licenses for operation outside of metros
Barrier-free entry for RSPs
Revenue sharing terms for RSPs that are as attractive
as the ones BSOs enjoy
Fee-free spectrum licensing for RSPs using wireless
technologies
Rationalized and reduced taxes and duties on ICT
goods
(with A. Jhunjhunwala, C. Maclay, B. Ramamurthi)
Rural Service Provision: VoIP
International
Domestic
International
Domestic
International
Domestic
International
Domestic
PC
PC
PSTN
PSTN
Terminate
Ori
gina
te
Public Policy: License-exempt Radio Usage
License exemption in USA at 2.4 GHz ISM band
Users do not require license to broadcast using approved radios.
Lowers entry barriers. Enhances innovation and
creative local uses.
License-Exempt Radio Bands in Africa: A Comparative Analysis
(with I. Netto & S. Gillet)
Most African countries permit use but require license for 2.4 GHz
License-Exempt Radio Bands in Africa: A Comparative Analysis
50 of 54 African countries categorized.
Corresponding ITU study categorized only 12 countries.
51% of countries require full licenses for broadcast in 2.4 GHz.
Unlicensed vs. Registration vs. Licensed
Most African countries permit use but require license for 2.4 GHz
Harmonization!
Significant diversity in regulations across the continent inhibits economies of scale and may discourage large entrants.
Lack of clarity and enforcement discourages innovation and small entrepreneurs.
We propose that NEPAD, working with regional economic communities and international players, work to harmonize regulations across the continent and build personnel and enforcement capacity.
Question 4: Business Models
One particularly exciting business model innovation is the micro- and small-enterprise provisioning basic and value-added communication services.
The Village Information Center
Telecenters, telecottages, community technology centers, community communication shops, village knowledge centers, public call offices, networked learning centers, multipurpose community telecenters, digital clubhouses, cabinas públicas, infocentros, community access centers, ….
SARI: Sustainable Access in Rural India
SARI aims to….
demonstrate the sustainability of the Internet and Internet-enabled systems and services in poor rural communities.
and show a linkage between such technologies and social and economic development.
(with A. Jhunjhunwala, C. Maclay & J. Sinha)
The SARI Village Information Centers
Provide Internet via WLAN, PC, and application suite to villages - many that are off the phone grid
Each village information center is locally owned and operated (franchise model)
Pilot Project Scope
Working in Madurai District, Tamil Nadu, South India
Madurai city not included Pilot project undertaken in
the Taluk of Melur covering the two Panchayat Unions of Melur and Kottampatti
Service area 2,000 sq km, 32,000 people
Pilot Status
80 connections in over 50 villages Average village size of 1,000 households;
smallest is 300 households Highest density of rural Internet kiosks
connections anywhere In catchment area 23% of population has
used the Internet (national average 1.5%, world 9%)
Connected Villages
Padinetankudi Karungalakudi Keelavalavu Vellalur Urranganpatti Thaniamangalam Alagarkovil Neaythanpatti T.Ulagpitchanpatti Sengarampatti Othakadai Attapatti Kottampatti Chittampatti Pudhutamaraipatti Pulimalaipatti Mankulam Karpuooravahini A.Vellalapatti
Navinipatti Kelaiyur Kallampatti Arittapatti Narasingampatti Therkutheru Kottakudi T.Vellalapatti Thiruvadhavur Arasappanpatti Vellaripatti Andipattipudur Thumbaipatti Melur- Kalanjiyam Tr Centre Palayasukkampatti Kuthappanpatti Kidaripatti Kattayampatti Pullipatti
Connected Villages
Current Applications
Education & Training (Windows, Office, etc.)
Cybercafe applications (e-mail, voicemail, chat)
E-government services (caste, income, birth, death
certification, pension schemes, complaints and petitions) Over
600 applications processed in 10 months.
Entertainment applications (Tamil movies, astrology, games)
Tele-health, tele-agriculture, tele-veterinary services
Current Research Inputs Include
household surveys operator surveys user surveys instrumented PC’s ISP meter reads maintenance logs daily usage reports government usage reports baseline surveys payment reports
There is an extraordinary challenge in collecting solid data on usage, outputs, and outcomes from rural facilities.(Colle & Roman)
An ICT4D Sustainability Framework
Economic sustainability (Heeks)
Social/Cultural sustainability (IDRC)
Political/Institutional sustainability (IDRC)
Technological sustainability
Economic Sustainability: A Micro Business Model
Capital costs: wiring, furniture $ 300 kiosk equipment 1,000 other 300
Recurrent costs (monthly): rent, electricity, maintenance 25 Internet 15 Interest and depreciation 28
Break-even revenue $68 (per month) $ 2.70 (per day)
Scaling the Results
Tamil Nadu’s rural population density is 297 people per square kilometer
Most of rural Tamil Nadu is within 50 kilometers of fiber backbone
The quality of grid electricity in Madurai district is fair The physical terrain of the Madurai district is fairly favorable
for terrestrial wireless systems Although these communities are poor (with individual incomes
averaging below $1 per day) and agricultural based, high levels of awareness and sensitization to the value of ICT’s helps to drive interest in Internet services
Sensitivity Analysis
Recall current break-even point is $2.70 per day
$300 Internet appliance $2.10 100 km Microwave backhaul 2.95 Population density of 100 people per sq km 4.31 VSAT 4.68 PV Solar Cells 5.06
Sensitivity Analysis
$2.10
$2.70$2.95
$4.31$4.68
$5.06
$0.00
$1.00
$2.00
$3.00
$4.00
$5.00
$6.00
Cheapappliance
Current Backhaul 100 people persq km
VSAT PV Solar
Question 5: Internal Capacity
By “internal capacity” I mean the training, skills, and capabilities of the community access center’s personnel and in particular the local manager and operator.
Selection
The fat density ellipses (red circles) and rollercoaster smoothing spline fit (green line) indicate the lack of a correlation between the two explanatory variables (months of prior computer experience and months of prior work experience) and the single response variable (average number of hours of Internet use per month per site).
Training
Work-based projects or assignments Observing experts and colleagues Reading books and journals Open learning and self-managed learning Workshops, courses, seminars, or conferences
(Murray, Murray & Brooks 2001)
Training
Peer-to-peer networks!
Retention
“I'd say an operator lasts on an average 6 months. Maybe that's how long it takes to learn computer skills and get another job! [One operator], for example, is onto her fifth job in three years. [Another of one of our most successful owners] hires a new person perhaps every 4 months.”
Question 9: Evaluation
“[E]valuation studies are urgently needed to provide an assessment of the role and impact of community telecentres, as organizations and donors are implementing these facilities… without an adequate understanding of how well they respond to the communication and information needs… or of their impacts on social equity and economic development.”(Whyte, 2000)
Critical Elements to Evaluation
Some of the critical elements to monitoring and evaluation include: Start early and monitor often. It may make sense to begin baseline
and preliminary assessments well before the facilities are ever in place.
Be sure to bring into the evaluation mix control communities (for instance comparable communities that are not part of the intervention).
Develop methodologies that are based on clear theoretical foundations (e.g. Roman 2003).
Develop consistent research frames, instruments, and a broad powerful range of indicators (Whyte, 2000).
Critical Elements to Evaluation
Disaggregate data along appropriate dimensions of equity (see Question 8).
Pay very close attention to data quality (which is often quite poor) (Roman & Colle, 2003).
Quantitative and objective measures that are amenable to statistical modeling can help ensure robust results (though it is easy to lie with statistics).
Qualitative and observational data can reveal truths that are not available to quantitative approaches (thought it is easy to be misled by anecdotes).
Usage Indicator
DEC01
JAN02
FEB02
MAR02
APR02
MAY02
JUN02
JUL02
AUG02
SEP02
OCT02
NOV02
DEC02
JAN03
FEB03
MAR03
APR03
MAY03
JUN03
JUL03
AUG03
SEP03
S1
0
10
20
30
40
50
60
70
80
90
100
Ave
rag
e h
ou
rs p
er c
on
nec
tio
n
Months
Overall Average Usage of Internet
(with J. Thomas)
I have performed a 22 month study of over 50 village information centers.
Revenues Indicator
Linear Multivariate Model
. exbxbby nno ...ˆ 11
.
We are seeking a model of the form:
Where ŷ is the response variable of interest (internet use) the x’s are the various explanatory variables, and the b’s are coefficients or parameters of the explanatory variables.
Least Squares Regression
. .
NumHousholds The number of households within the village (village population). %Hindu The percentage of the village that is Hindu (as opposed to Christian, Muslim, or other religions) Teledensity The number of phone lines, per capita, in the village. OperatorLandowner A dummy variable (1 or 0) representing whether the telecentre operator owns any land. OperatorWorkExp The number of months of prior work experience that the telecentre operator has from
other jobs. OperatorOwnKiosk A dummy variable (1 or 0) representing whether the telecentre is owned by the operator.
)(5.40)(6.1)(5.31
)(449)(%2.372)(23.4.22ˆ
nKioskOperatorOwrkExpOperatorWondownerOperatorLa
yTeledensitHindudsNumHousholy
Least Squares Regression
. .
Observed average Internet use plotted against predicted average Internet use (R2 = 0.94, p = .0005, n=51).
0
50
100
150
200A
vera
ge In
tern
et U
se A
ctua
l
0 50 100 150 200
Average Internet Use Predicted
Acknowledgments
I have received the generous support of The World Bank, UNEP, ITU, USAID and the
Engineering Information Foundation
SARI has been supported in part by USAID, ICICI, the Internet Business Capital
Corporation, n-Logue Communications Pvt. Ltd. and a grant from Ray Stata
Shameless Advertisement
Please subscribe, submit your work, and readInformation Technologies and International Development
Edited at the University of Maryland and Georgia Tech and published at MIT.
http://mitpress.mit.edu/ITID
Community Access to Communications:10 Questions – Provisional Answers
Community Access to Communications:10 Questions – Provisional Answers
Michael L. Best
Thank You!