Jolle Toledano ARCEPs Commissioner, Chair of the Postal Economics Project Group, Professor of Econom

1
Postal economics in developing countries Towards
new models
Joëlle Toledano (ARCEPs Commissioner, Chair of
the Postal Economics Project Group, Professor of
Economics)joelle.toledano_at_arcep.fr José Anson
(Economist, International Bureau, Universal
Postal Union)jose.anson_at_upu.int
2
Postal economics for developing countries (DCs)
a lack of models

• Much strategic and economic work on postal
  reform in industrialized countries
• Little study of the reforms undertaken in
  less-developed countries
• Other network sectors (telecoms, energy,
  transport, etc.) studied in DCs
• World Bank GICT report and other isolated
  studies
• Trend to project the results of postal economics
  for industrialized countries onto developing
  countries (e.g. the analysis of delivery, the
  relationship between income and mail volumes,
  regulation, etc.)
• Our warning this kind of extrapolation may not
  be appropriate

3
Industrialized countries postal stylized facts

• The postal model in industrialized countries
• high letter-post volumes
• B2C-driven traffic
• high salaries and heavy automation
• large economies of scale for home deliveries.
• the sender pays for home delivery (Rowland Hill)
• generally low share of international mail in
  their total volume of mail
• Universal service means delivery everywhere, at
  affordable prices, and with a good quality of
  service

4
DCs postal stylized facts

• In contrast, in DCs, and particularly in Africa
• low levels of letter-post volumes
• no high salaries and no heavy automation
• delivery may be organized differently than in
  industrialized countries
• sometimes no delivery at home at all
• the sender and addressee may both pay for postal
  services
• possible high share of international mail in
  their total volume of mail
• Postal sector not developed under
  laissez-faire (e.g. many countries in Latin
  America).
• First result applying the analytical framework
  and postal regulation model of the industrialized
  countries to countries with lower traffic,
  different composition of mail flows, and with a
  different economy of delivery is not appropriate

5
Two questions on the future of the postal sector
and USO in DCs

• What are the viable growth models for the postal
  sector in DCs which will result in a reliable,
  affordable postal service accessible to all?
• Why promote the development of USO? What is the
  contribution of the postal sector to economic
  development (financial services, new information
  and communication technologies)?
• To respond to these questions, we must develop
  new postal development models that draw lessons
  from appropriate economic analysis.

6
Are traffics and GDP linked for low-income
countries?
R2 0.69
7
The classical view
Not really!
R2 0.69
8

• Clustering our economic analyses of the postal
  development in developing countries ...

9

• For low-income countries (up to 812 USD per
  capita in our sample), preliminary econometric
  results account for 73 of the differences in
  domestic traffics between countries
• three basic explanatory factors, in order of
  importance
• (populationchildrenilliterate
  people)/population share of potential users in
  the population-gt positive effect
• percentage of the population having to collect
  mail from the Post -gt negative effect
• potential users per post office (not true for
  ICs)-gt negative effect (thus positive effect of
  a larger network)

10

• For middle-income countries (between 847 and
  9,568 USD per capita in our sample), preliminary
  econometric results explain 86 of the
  differences in domestic traffics
• four basic explanatory factors, in order of
  importance
• per capita gross domestic product (GDP in terms
  of ppp) -gt positive effect
• the proportion of the population who are
  potential or actual users of the letter post-gt
  positive effect
• the price of the domestic letter-post service
  (price in terms of ppp)-gt negative effect
• potential users per post office (not true for
  ICs)-gt negative effect (and thus positive effect
  of a larger network)

11

• For low-income countries (up to 812 USD per
  capita in our sample), preliminary econometric
  results also suggest that
• domestic traffic is evolving simultaneously with
  three other variables
• number of Internet connections per 1,000
  inhabitants -gt positive correlation -gt not
  substitutes, but rather complements
• number of mobile telephones per 1,000
  inhabitants -gt positive corr. -gt not substitutes,
  but rather complements
• amount of value added to GNP by the service
  sector -gt positive corr.

12

• For middle-income countries (between 847 and
  9,568 USD per capita in our sample), preliminary
  econometric results also suggest that
• domestic traffic is evolving simultaneously with
  the following five variables
• number of international letter-post items
  received per potential or real user of the letter
  post -gt positive correlation (corr.)
• population density -gt positive corr.
• number of Internet connections and mobile
  telephone subscribers per 1000 inhabitants -gt
  positive corr. -gt no substitution, but rather
  complementarity with new ICTs
• degree of competition, legal or de facto -gt
  negative corr.

13

• An historical perspective may also be insightful
  ...

14
Yesterday (1876) number of letters per
inhabitant in industrializing countries
8.7
15
And today (2004) in industrialized countries
404.9
Over a century volume of mail 50X whereas GDP
25X and GDP/hab. 5X
16
And today (2004) in DCs
37.6
13.4
5.2
3.2
18.1
DCs Posts today what is a postal model for
them that will be as much effective as the
ICs public monopoly over the last century?
17

• The next step what are the viable growth models
  for the postal sector for different clusters of
  developing countries as given by the results of
  our econometric and economic analyses?

18

• First understanding and evaluating what already
  exists
• A look at a first cluster of developing countries
  with intermediate levels of letter-post volumes
  the case of Latin American countries ...

19
Latin America (except Brazil) and destructive
competition

• Very often unregulated postal markets in Latin
  America
• Sometimes lack of concern of the government for
  the postal sector
• Development of destructive competition
• Lower volumes, higher prices for individuals
• High levels of market instability (turn-over
  of illegal competitors)
• Excessive capacity leading to infrastructure
  destruction over time
• Widespread price discrimination
• Unstable regulatory frameworks
• See Rudy Cuadras presentation during CA 2005

20

• Understanding and evaluating what already exists
  and focusing on delivery
• A look at a second cluster of developing
  countries with low to very low levels of
  letter-post volumes the case of Sub-Saharan
  African countries
• Current questionnaire and interviews of
  Sub-Saharan African countries focusing on
  delivery

21

• Why delivery?
• 50 of total costs in industrialized countries
• Free delivery introduced between 1840 and 1850 in
  many countries
• But what about delivery in Sub-Saharan African
  countries today?

22
A snapshot from a mission to Niger

• Much strategic and economic work on postal
  reform in industrialized countries
• Little study of the reforms undertaken in
  less-developed countries
• Other network sectors (telecoms, energy,
  transport, etc.) studied in DCs
• World Bank GICT report and other isolated
  studies
• Trend to project the results of postal economics
  in industrialized countries onto developing
  countries
• This kind of extrapolation may not be
  appropriate

23
On the other side

• Much strategic and economic work on postal
  reform in industrialized countries
• Little study of the reforms undertaken in
  less-developed countries
• Other network sectors (telecoms, energy,
  transport, etc.) studied in DCs
• World Bank GICT report and other isolated
  studies
• Trend to project the results of postal economics
  in industrialized countries onto developing
  countries
• This kind of extrapolation may not be
  appropriate

24
Beyond a first look at 2004s UPU statistics
25
Carrying out research on postal delivery is
paramount for building up new postal models for
developing countries (DCs)

• No actual universal delivery in many developing
  countries (DCs)
• Very little home delivery in lots of least
  developed countries (LDCs), particularly in
  Africa
• Very often receivers paying to access delivery
  services
• Twice to sixty-nine times as much mail
  collection from post in DCs compared to ICs
• Totally different delivery models compared to
  industrialized countries (ICs)
• Being the last step of the postal process,
  delivery encompasses features of the whole postal
  process (backward approach)

26
A questionnaire about postal delivery

• Purpose describe delivery on the domestic
  letter-post market of developing and transition
  countries
• First with Sub-Saharan African countries
• Five sets of questions prepared with African
  public operators during a mission to Niger and
  on-going interviews of countries
• Forthcoming adaptation of the methodology and
  questionnaire for other regions (in current
  preparation for Latin America)
• All regions welcomed to participate in the
  development of the methodology for the analysis
  of their delivery

27
A questionnaire about postal delivery developed
for Sub-Saharan African countries

• Geography of the country three main cities and
  the rest defined as rural areas
• Modes of distributing mail home delivery? P.O.
  box delivery? Other modes of delivery (village
  chief, school, government agency, etc.)?
• Prices of domestic postal services for mailers
  and for addressees, and comparison with other
  basic goods (bread, rice, meat, mobile phone
  call)
• Data on the network infrastructure (human and
  physical e.g. number of employees or p.o. boxes)
  and its costs (e.g. salaries)
• Information about competitors -gt their prices
  and where they are competing with the PPO
• Information about big mailers

28
Describing postal delivery in Africa 2006
questionnaire and interviews of 30 Sub-Saharan
African countries
Access
Viability
Affordability
Reliability
29
Describing postal delivery in Africa four layers
of analysis within each country
Most heavily populated city
Second most heavily populated city
Third most heavily populated city
Rural regions
30
A first glance at Sub-Saharan African postal
models focusing on delivery

• 7 African countries analyzed so far (on going
  analysis for 23 other African countries and
  forthcoming analysis of all African countries)
• All low income countries
• 6 least developed countries out of the 7
  analyzed
• From 0.41 letters per potential user of postal
  services to 2.93 letters
• Thus a rather homogeneous cluster of countries
  with very low volumes of mail per inhabitant
• Describing their delivery in terms of access,
  affordability, reliability and viability for the
  domestic letter-post
• Question are they are applying a sound postal
  model for the domestic letter-post?

31
Geographical coverage of the delivery study
Share of the total population for each
geographical level
Remark Rural includes rural regions and other
smaller cities
32

• Access to delivery points ...
• A delivery point may be the addressees home in
  case of home delivery
• otherwise a p.o. box in most of the cases
• and sometimes other delivery points such as
  village chiefs, schools, retail shops, government
  agencies in rural areas

33
Access to the public postal operators delivery
points in our sample of Sub-Saharan African (SSA)
countries
Inhabitants per delivery point
Access to delivery is concentrated in main
cities, 85 of the population lacks very much of
it Only one country out of seven provides home
delivery
34
Access to the public postal operators p.o. boxes
in our sample of Sub-Saharan African (SSA)
countries
Inhabitants per p.o. box (countries with p.o. box
delivery only)
Anecdotic evidence tells that up to 10 persons
may share the same p.o. box the numbers above
thus suggest a large exclusion of people to
postal delivery throughout the cities and the
rural
35

• Affordability ...

36
Affordability of PPOs ordinary letter-post
services in SSA sending a letter
Price for sending a domestic item relative to the
price of other goods
Given the relative importance of the act of
sending a letter in these economies, it is quite
affordable for the mailer but the addressee
must rent a p.o. box in most of the cases
37
Affordability of PPOs ordinary letter-post
services in SSA renting a box
Annual price for renting a p.o. box relative to
the price of other goods
Renting a box is likely to be very much less
affordable for most of the people thus very
much reducing the delivery points in the network
and the value for people of renting a box
(network externalities)
38
Affordability of PPOs domestic ordinary
letter-post services in SSA overall price
Overall price of sending a letter for the mailer
and the addressee
The overall transaction cost (or revenue by
letter for the PPO) of exchanging a letter
between the mailer and the addressee is less
affordable than when only looking at the price
the sender pays (two-sided market) 2 to 3 X
less affordable!
39

• As we will all agree, no comparison with the
  postal sector in industrialized countries!!! ...

40

• Reliability ...

41
Reliability of PPOs ordinary letter-post
services in SSA
Delivery choice by large mailers in Africa
In our sample Self-delivery of the bills by
electricity distributors in all
cases! Self-delivery of the bills by water
distributors in all cases! Mix of self-delivery
of the bills by telecom operators and delivery by
the PPO Banks systematically used the PPOs to
deliver their mail
42

• Competition or Substitution ...

43

• Much strategic and economic work on postal
  reform in industrialized countries
• Little study of the reforms undertaken in
  less-developed countries
• Other network sectors (telecoms, energy,
  transport, etc.) studied in DCs
• World Bank GICT report and other isolated
  studies
• Trend to project the results of postal economics
  in industrialized countries onto developing
  countries
• This kind of extrapolation may not be
  appropriate

44

• The competitors to the PPO are most of the time
  in cities and rarely in rural areas -gt cream
  skimming
• Their volumes are very difficult to be estimated
  (only a very few surveys available as well as
  anecdotic evidence)
• but we have more information now

45
Affordability of PPOs competitors for ordinary
letter-post services in SSA
Price for sending a domestic item relative to the
price of other goods
Competitors price according to the willingness to
pay of people and to the distance, and since the
importance of sending an ordinary letter may be
very high and the distance long prices follow
and are relatively high!
46
Assessing the reliability of PPOs ordinary
letter-post services with respect to their
competitors
Proxies for assessing PPOs relative reliability
in Africa
Presence of black postal markets as well as a
much higher relative price for the competitor are
bad signals regarding the reliability of the
PPO Competitors more expensive because they are
faster than PPOs on the ordinary letter market!!!
47

• Viability ...

48
Viability of the PPOs ordinary letter-post
services in SSA
Wage in terms of letters, letters per employee
and viability ratio
The median operator with a p.o. box delivery
model is not viable enough! A real mix of home
and p.o. box delivery reaches a better viability
49
Viability of domestic ordinary letter-post
services in SSA
A balanced architecture of the network and an
appropriate pricing matters!
In our sample Higher viability is reached by
operators whose prices for sending a letter are
not too low and whose prices for renting a box
are not too high (price elasticity for sending
and receiving a mail may very much differ) Higher
viability is also reached by those operators with
a sufficient number of delivery points
distributed in a more equal way throughout their
postal territory
50

• The next question can we generalize these
  results? ...

51

• In the meantime, ongoing projects for the next
  months include
• adapt the methodology and questionnaire for the
  study of delivery in other regions of the
  developing world
• presentation of a paper about the postal sector
  in Latin America and what we can learn from this
  experience in the forthcoming Rutgers Conference
• study the impact of the postal infrastructure on
  the quality of service provided by other networks
  and infrastructures and on their development

52

• POC Postal Economics Project Group
• To join the Group, contactjose.anson_at_upu.int
• Postal Economics Project Group website with
  documents and presentationswww.upu.int/postal_e
  conomics/en/index.html