studies.jpg (1974 bytes)studies.jpg (1974 bytes)studies.jpg (1974 bytes)


Programmes

ATAC

Partners

Resolutions

Activities

NICI Plans

Events

Contact Us

Links

 

aisilogo2home.jpg (1966 bytes)

Search | What's New?| Site Map | Discussion Lists

CONNECTIVITY IN AFRICA: USE, BENEFITS AND CONSTRAINTS OF ELECTRONIC COMMUNICATIONS

Synthesis Report - Part 2 : Overview of the project findings

Study carried out under the CABECA project of PADIS (UN-ECA) as part of

the IDRC sponsored research program on the Impact of information on development

Michel J. Menou

 

DRAFT

1. Introduction

The project which is being reported here was originally called "Impact of electronic communications on development in Africa", consonant with the overall theme of the program of the International Development Research Centre (IDRC), Canada through which it was supported. The research team felt that this title sounded too broad and ambitious with respect to the actual scope and outcome of the project. It thus decided to use the above main title for the final reports.

The final report of the project is made of a series of self-contained documents with a view to facilitate their further dissemination and use. The series include:

  • Extended abstracts of the synthesis report and national reports;
  • A synthesis report in two parts, one dealing with methodological issues, the other one with the overall findings and conclusions of the project;
  • The reports of phase one and phase two of the national studies in Senegal, Uganda and Zambia and a consolidated report for phase one and two in Ethiopia, all in a revised version.

The reports will be available in print or electronic form from Mrs. Nancy Hafkin of DISD, ECA. They are also posted on the Web site of the African Information Society Initiative, whose URL is: http://www.uneca.org/aisi.

This part of the synthesis report is devoted to the overall findings and conclusions of the project. Readers are referred to the individual reports of the national investigators for more details. The present synthesis report has no intention to substitute them but tries to complement them by offering a common perspective. The background of the study, its methodology and the problems it offered are discussed in the first part of the synthesis report.

We wish to gratefully acknowledge the contribution and commitment of the other members of the international team which carried out the project: Mrs. Nancy Hafkin, (Officer in charge, PADIS, UN Economic Commission for Africa), Ms. Jane F. Asaba, (Coordinator, Documentation and Information, National AIDS Documentation and Information Centre (NADIC), Uganda AIDS Commission), Mr. Bernard Bazirake Bamuhiiga, (Senior systems Librarian, Makerere University, Kampala, Uganda), Mr. Vitalicy Chifwepa, (Head, Department of Library and Information Studies, School of Education, University of Zambia), Mr. Abebe Rorissa, (Automation Librarian, Thomas Mofolo Library, National University of Lesotho), and Mr. Omar Diop, (Consultant, Dakar, Senegal).

top.gif (903 bytes)

2. Implementation of the study

The study was due to proceed as follows:

- 1st Phase

  • Identification of the population and sampling;
  • Baseline survey using a questionnaire distributed by Electronic mail (Appendix A);
  • Interviews of a sample of 50 subscribers among the respondents to the baseline survey.

- 2nd Phase

  • Second series of interviews with the same respondents;
  • Interviews or group discussions with non users and decision-makers.

As discussed in the first part of the synthesis report, circumstances prevented the second phase to be the complete repeat of the first phase as initially planned. The coherence between the population surveyed and interviewed across the four countries, which was to some extent achieved in the first phase could not either be kept in the second phase. The second phase thus encompassed the following activities:

  • Restoring the lists of subscribers to be surveyed and/or interviewed;
  • Complementing the baseline survey using a simplified questionnaire (Appendix B);
  • Conducting a survey of all subscribers about the use of the Internet with a questionnaire distributed by Electronic mail(Appendix C);
  • Second series of interviews with the same respondents as in the first phase.

Table 1 below summarizes the main steps in the implementation of the study. The discrepancies in the periods when the data were collected in each country are not such that they might affect the overall coherence of the results in this respect. As a matter of fact each country is rather living in a particular time according to its history in the area of electronic communications with regard to both the extension of connectivity and regulations.

The baseline survey was due to be distributed to all active users of the Cabeca or Fido nodes in a reference period (January - February / March 1995) which was itself considered as fairly representative or the normal traffic. The subscribers' migration to other ISPs caused discrepancies between the two phases. Tables 2 and 3 below shows the response rates to the various surveys in both phases.

Table 1 Time table of the project activities

Activities/dates

ETHIOPIA

UGANDA

ZAMBIA

SENEGAL

Project preparation: May 1994 - Jan. 1995
Project approval: Nov. 1994
1st Workshop: Feb. 27 - Mar. 1, 1995
Preparation of 1st phase survey Mar. - Sep. 1995 Mar. 1995 Mar. 1995 Mar. 1995
1st phase survey Aug. - Dec. 1995 Apr. - Dec. 1995 Apr. - Dec. 1995 Aug. 1995 - Jan. 1996
1st phase interviews Dec. 1995 - Mar. 1996 Jan. - Feb. 1996 Apr. - May 1996 Jul. 1995 - Mar. 1996
Progress reports: Feb. 1996
2nd Workshop: Feb. 12 - 15, 1996
1st phase reports May 1996 Feb.1996 Jun. 1996 Jun. 1996
Data collection for 2nd phase   Apr.- Oct. 1996    
Complement of 1st phase Oct. 1996 - May 1997   Feb. - Mar. 1996  
2nd phase survey   Apr. - Dec.1996 Feb. - Apr. 1997 Mar. - Dec. 1996
2nd phase interviews Mar. 1997 Jul. - Dec. 1996 Jun. - Aug. 1997 Mar. - Dec. 1996
2nd phase reports Aug. 1997 May 1997 Aug. 1997 Sep.1997
3rd Workshop: March 19-21, 1998
Preparation of final reports: March April 1998
Preparation of synthesis report: April - June 1998

 

Table 2. Response rates for the baseline survey

 

ETHIOPIA

UGANDA

ZAMBIA

SENEGAL

Reference population

206

164

90/256 (2nd ph.)

106

1st phase number

101

48

23

98

1st phase percent.

49.03

29.27

25.56

92.45

2nd phase number

n.a.

57

50

100

2nd phase percent.

n.a.

34.76

19.53

94.34

Table 3. Response rates for the survey of Internet use

 

ETHIOPIA

UGANDA

ZAMBIA

SENEGAL

Reference population

206

164

256

100

Number

58

26

50

100

percent.

28.15

15.85

19.53

100

The structure of the tentative sample for the interviews was shown in the first part of the synthesis report. The tentative target number of 50 interviews in each country was decided on the basis of the estimates by the national investigators of the time required and available in order to carry out the task. The interviews were supposed to be conducted using either an analytical or "historical" schedule (Appendix D). We wanted to secure both representativeness and homogeneity across countries of the persons to be interviewed.. Due to the circumstances, the size and structure of the groups which were actually interviewed were quite different. All those interviewed did in principle fill the baseline questionnaire. Note that in Uganda 20 persons were interviewed during the first phase and 18 testimonies were later on collected by electronic mail using a simplified list of questions in order to complement the survey. In Senegal, especially in the second phase, some of the interviews have been reported as being group discussions or testimonies obtained by electronic mail. The population interviewed in the second phase was different from the one interviewed in the first phase, except perhaps in Senegal. Further more the schedules were adapted in each country in order to possibly take into account the effects of access to the Internet. One has thus rather a single series of interviews, stretched over the time and with variations in scope, than the expected repetition. Table 4 below shows the number of interviews achieved in each phase.

Table 4. Implementation of the interviews

 

ETH

IOPIA

UGA

NDA

1

UGA

NDA

2

ZAM

BIA

1

ZAM

BIA

2

SEN

EGAL

1

SEN

EGAL

2

Target

50

50

50

50

50

50

50

N

33

38

26

20

50

50

50

ACAD

8

*

11

7

12

*

*

PRIV

11

*

3

2

9

*

*

GOV

2

*

6

4

15

*

*

NGO

7

*

3

5

9

*

*

IGO

5

*

3

2

5

*

*

* Data not available

Key: ACA = Academic & Research; GOV = Government organizations and parastatals;

PRIV = Private sector companies &individuals; NGO = Non governmental organizations; IGO = International and foreign governmental organizations

top.gif (903 bytes)

3. Characteristics of the population

Through the baseline survey, the study tried to check how a number of characteristics were represented whose role is often regarded as influencing the use of electronic communications.

Gender

As shown in table 5, the gender distribution is predominantly male. It would be interesting to verify if this distribution is similar to the one found in the respective categories of staff, as it seems, or if there are signs of relative advantage or prejudice for women in relation to the use of information technologies. The HealthNet evaluation (Evaluation unit, 1994) noted that professional women were under represented among the users while the proportion of clerical staff was high.

Table 5. Gender distribution of respondents (in percent)

 

ETHIOPIA

UGANDA

ZAMBIA

SENEGAL

N

101

105

24

100

Women

13.9

31.5

37.5

12

Men

86.1

68.5

62.5

88

Age

Collecting data on this item proved even more difficult than expected. Those obtained are too scarce and dubious for drawing any reliable indication. The rate of no response on this item was as high as 28% in Uganda. It seems that the majority of users is likely to be found in the age group between 30 and 45. It would be worth investigating the relationship between hierarchical level, seniority, age and the use of information technologies. The HealthNet evaluation (Evaluation unit, 1994) noted that electronic communications were said to be more available to the higher levels of the hierarchy than to line staff. If confirmed this fact may cause some difficulty as the more senior and older staff, the more likely it is to have not been trained in the use of computers and telecommunications. Conversely, management which is itself fluent in computer usage seems to be more concerned with the application of the technology.

Nationality

The special issues raised by the proportion of nationals and expatriates within the users' population were discussed in the first part of the report. Table 6 below shows the data obtained; in Zambia, expatriates and non-responses were counted together in the second phase. In order to get a more precise view of the links between the expatriates and the development of the countries, we tried to obtain the proportion of permanent residents (5 years or more) versus temporary residents (below 5 years). It is shown in table 7.

Table 6. Distribution of respondents by nationality (in percent)

 

N

Nationals

Expatriates

Ethiopia

101

58.4

41.6

Uganda

105

60

40

Zambia (1st phase)

23

43.5

56.5

Zambia (2nd phase)

50

70

n.a.

Senegal

100

74

26

Table 7. Proportion of permanent versus temporary residents (in percent)

 

Temporary

Permanent

Ethiopia

69

31

Uganda

69

31

Senegal

42

58

The examination of their other characteristics, patterns of use, benefits and difficulties did not reveal any significant difference with the national users. The size of the groups is however too small for drawing any firm conclusion. Furthermore, it is the demonstration and training function of expatriates who are experimented users of electronic communications in relation to their local colleagues which would deserve particular attention.

Educational level

As one may expect the educational level of the electronic mail users is quite high. The data obtained suggest that few are below undergraduate university education. The educational levels of national and expatriate users did not show much difference. Caution should be exerted in the interpretation of the data. The indications given by the respondents may not be fully reliable or precise enough. It was also difficult to align the degrees in the French style educational system with those of the anglo-american one without knowing where the degrees were obtained.

As a matter of fact the place and date of the university education, and possibly the discipline, may be a more decisive factor in relation to the exposure to electronic communications, or computer use, than the level of the degree. It will nevertheless be interesting to watch the evolution of the proportion of users with the lowest levels of education which may point to the generalization of the facility.

Table 8. Educational level of respondents (in percent)

ETHIOPIA

UGANDA

ZAMBIA (1st P)

Zambia (2nd P)

SENEGAL

N

101

57

23

50

100

High school

1

|

Diploma

3

|

| 20

BA/ BS / Maîtrise

35

| 26.3

33.3

|

35

MA/ MS / DEA/ DESS

29

36.8

44.4

| 80

49

PhD/MD

32

26.3

22.2

|

16

Language

Our attempt at assessing the language barrier was not very successful either. Most authors (e.g. Brown, 1994, p. 63; Sagna and Denis, 1996, p. 144) highlighted the domination of English as an obstacle to the international use of electronic communications, but little is known in practice about its effects. The observation of international discussion groups shows that participation from countries where English is not the communication language tend to be quite low. In Senegal 40% of the respondents indicated that they speak English and 45% that they can read it. In Ethiopia, 91% of the national users indicated that they knew English. In Uganda 67% of respondents whose mother tongue was another one said they were fluent in English and about one third that they also knew another language; interestingly, USA and UK respondents did not indicate any other language.

One should probably better fine tune the approach of language as an ability or obstacle to the use of electronic communications by considering the mother tongue, the language in which higher education was taken, the working language and the communication language in the country. The spread of English as a working language, or second language in countries which use another communication language will be interesting to observe. This is especially true for Portuguese and Spanish speaking countries of Africa, which may suffer more from isolation. Observation rather than a questionnaire survey will provide insights in this complex issue.

Computer skills and electronic mail experience

Significant national differences seem to exist in relation to the level of computer skills, as shown in Table 9. They are partially reflected in the replies regarding the way electronic mail is being sent. The high proportion of Senegalese respondents who consider themselves as beginners would be worth validating both within this country and among other non English speaking countries. Computer skills were found, as one may expect, associated with the intensity of use.

Table 9. Computer skills (in percent)

 

N

Experienced

Intermediate

Beginner

Ethiopia

101

60

37

3

Uganda (2nd. P)

57

72

23

5

Zambia (1st P)

23

78

22

---

Zambia (2nd P)

50

60

27

13

Senegal (2nd P)

100

20

39

41

The previous use of electronic communications was tentatively analyzed taking 1993 (more than 3 years) as the break point, since it is the date when electronic mail became generally accessible in the countries under study. Only fragmentary data, shown in table 10, were obtained. With more data points, it would be worth considering the difference in the duration of experience between national and expatriate users in order to ascertain the role eventually played by studies and work abroad in familiarizing users with electronic communications, and therefore pre-determining their perceptions and expectations.

Table 10. Duration of use of electronic mail (in percent)

 

Ethiopia

Uganda

Zambia

Senegal

N

85

59

50

100

Less than 1 year

31.8

23.7

30

54

Less than 2 years

37.6

10.2

|70

10

Less than 3 years

27.1

30.5

|

25

More than 3 years

3.5

35.6

0

11

Professional affiliation

The professional affiliation of the respondents (R) is shown in Table 11. It presents, as one may expect, noteworthy discrepancies with the distribution found among all subscribers in the reference population (S). The scope of each category is itself subject to varying interpretations, e.g. academic institutions or public research centers being listed as Government agencies as was often the case in Uganda. It further proved impossible to distinguish private individual subscribers from business ones, thus the single category. The structure of the population studies in Zambia during the second phase could not be established for lack of data; it may present marked differences with the one found in the first phase. It seems that government organizations and private sector organizations have subscribed to electronic communications services at a later stage than academic and research ones, especially when full Internet access became available. Their share in the total population is likely to have grown significantly in all countries compared to the "Fido" era. The private sector users appeared also to be more suspicious about the survey; they are certainly under-represented in the present findings.

Table 11. Distribution of respondents' and subscribers' professional affiliation (in percent)

 

ETHIOPIA

UGANDA

ZAMBIA

(1st P)

ZAMBIA

(2nd P)

SENEGAL
 

R

S

R

S

R

S

R

S

R

S

N

101

206

48

164

23

159

50

256

100

106

ACA

39.6

14.6

6.2

19.5

43

20

24

---

66

30

GOV

3

6.3

46

9.8

4.3

25

30

---

12

14

PRIV

17.8

32

12.5

49.4

8.7

16

18

---

6

16

NGO

15.8

30.6

20.8

14.6

21.7

13.8

18

---

6

32

IGO

23.8

16.5

14.5

6.7

21.7

25.1

10

---

10

8

Key: ACA = Academic & Research; GOV = Government organizations and parastatals;

PRIV = Private sector companies &individuals; NGO = Non governmental organizations; IGO = International and foreign governmental organizations

The indications provided by the respondents regarding their particular profession or specialties, principal activities and responsibilities are too general and inconsistent from one country to the other to allow for any meaningful comparison or analysis. This one more type of data which can only be obtained through interviews and observation.

[end of first part]

CABECA Synthesis report Part 2 Findings DRAFT

Part 2 / 4

4. Patterns of use

Number or users per account

The convenience of using electronic communication facilities, thus the intensity of the use by individual users, is dependent on the availability of a connected computer. When a single machine is serving a large number of users, one may expect that the individuals' use will be constrained. Without direct observation it is difficult to have an accurate perception of this aspect. The indications obtained from the questionnaires which are summarized in table 12 are nevertheless interesting. One may wonder if it is the exposure to electronic communications or the availability of more effective services which is prompting the development of equipment as the data from the second phase in Uganda and Zambia seem to suggest. Single use is also likely to be linked with resources; in Ethiopia, 60% of the sole users are expatriates

Table 12. Number of users per account

 

ETH

IOPIA

UGA

NDA 1

UGA

NDA 2

ZAM

BIA1

ZAM

BIA 2

SEN

EGAL 1

Total N

101

48

57

23

50

96

Multiple

75.25%

83.3%

78.9%

 

36%

 

Single

24.75%

16.7%

21.1%

 

64%

 

Average number

 

4

5

8

4

16

Maximum

   

20

36

20

100

Use of multiple networks

In countries where full Internet has been recently introduced it seems that, for the time being, a number of subscribers retain their account with the Fidonet node and use it in parallel with their Internet account, though to a much lesser extent. As Robinson (1996, p. 208) noted the continuation of Fidonet use may be linked with delays in upgrading the physical infrastructure and training needs. When several networks are available, the proportion of organizations running multiple accounts seems to be relatively low (10 to 20%).

Payment for electronic communications

In all countries the cost of electronic communications is predominantly born by the organizations to which respondents belong, or through projects implemented by these organizations as shown in table 13. The latter source of funding is of concern as it often causes an interruption of the service at the end of the project. Among the respondents who pay themselves is a high proportion, e.g. 60% in Ethiopia of private or individual users. Informal sources have also pointed to a negative effect of the privatization of public telephone operators. The budget allotments is often delivered with noteworthy delays to government organizations, thus preventing them to pay their bills in time. With public monopolies this usually had limited consequences, but with commercial operators the telephone lines are cut after a short time with all the consequences one can imagine upon electronic communications.

Table 13. Payment for electronic communications

 

ETH

IOPIA

UGA

NDA 2

ZAM

BIA1

ZAM

BIA 2

SEN

EGAL 1

SEN

EGAL 2

Total N

100

57

23

50

98

100

Oneself

14.0

22.8

 

12

 

14

Organization

46.0

54.4

82.6

82

 

32

A Project

40.0

21

 

6

97

54

Other

 

1.8

       

Information and telecommunications facilities in the organizations

Except in Ethiopia, the number of respondents who feel that the information facilities available in their organizations are adequate is quite high, as shown in table X. below. But the number of those holding the opposite view is also significant. The data does not allow to make any link between the information facilities available and the use of electronic communications. One may hypothesize that the richness of the information environment has a unique effect. If it is high, it raises the desire for more resources and facilitates the identification of potential sources. If it is low, the need for overcoming the drawback is more pressing. Thus the role of electronic communications as a means to strengthen information resources would be secondary compared to personal networking. The data on the purposes of use below seem to support the hypothesis. It will be interesting to watch how these factors evolve over time.

Table 14. Information facilities

 

ETH

IOPIA

UGA

NDA 2

ZAM

BIA1

ZAM

BIA 2

SEN

EGAL 1

SEN

EGAL 2

Total N

99

54

20

50

98

100

Adequate

31.3

42.6

50

50

70

64

Inadequate

60.6

38.9

50

50

30

35

Not Available

8.1

18.5

--

--

--

1

The survey did not provide the data we would have needed regarding the telecommunications facilities available. It seems that all organizations have at least one telephone line. Fax is not available in a small number of cases (e.g. 20% in Ethiopia, 4% in Zambia). Given their wide availability, it might be more appropriate to consider the frequency of use and failure of both facilities. Some indication about the way they are used can be drawn from the figures presented in table 15. Data about other facilities (telex and radio) were provided only in Ethiopia. Respondents in Zambia and Uganda said the communications facilities available in their organizations are appropriate in 66% and 61.4% of the cases respectively.

Table 15. Use of other facilities for communication (in percent; multiple choices)

 

ETHIOPIA

UGANDA 1

Total N

345

151

Mail

34.2

27.8

Telephone

20.6

27.8

Telex

8.1

4

Fax

18.5

26.5

Travel

18.5

13.9

Some indications about the substitution effect of electronic communications were obtained from the first survey in Uganda. They replace Fax in 28%of the cases, Mail in 26.5%, Telephone in 25%, Travel in 12.1% and Telex in 8.3%.

Mode of using electronic communications

We did not succeed in obtaining detailed information about the use of computers for sending and receiving electronic mail. It would be useful to know for instance if the users have their own computer connected or should use a computer dedicated to electronic communications and where the latter is located. Other studies have shown that the farther the machine, the lower the use. . Network accessibility was for instance identified by Abels and co-workers (1996, p. 156) as the primary factor in the decision to use electronic communications by the faculty of small US colleges. Fragmentary data from Ethiopia and Zambia tend to indicate that less than 20% of respondents have access to electronic communications through the computer on their own desk. The connection of the computers to a LAN which is itself connected to the Internet is another situation which may make electronic communications easier and thus wider used.

Table 16 shows the distribution between office and home use. The use of computer at home, though limited, is significant in Uganda and Zambia, and to a lesser extent in Ethiopia. It is much lower in Senegal, but this is possibly due to the high proportion of respondents employed in government organizations. Cost and technical reasons are obvious limitations to an utilization at home, which may also be hindered by socio-cultural factors.

Table16. Places where electronic communications are used

 

ETH

IOPIA

UGA

NDA 2

ZAM

BIA1

ZAM

BIA 2

SEN

EGAL 1

SEN

EGAL 2

Total N

99

57

20

50

98

100

Office

77.2

66.7

70

76

97

92

Home

7.0

8.8

     

4

Both Office & Home

15.8

24.6

30

24

3

4

We were also interested in knowing how the facility was used, one the assumption that personal and direct manipulation was manifesting a higher level of integration. Intermediaries seem to play an important role which does not match the declared computer literacy and may be associated with status and other socio-cultural conditions. We tried to obtain details about the process, e.g. whether messages were dictated to an intermediary, or passed to him/her as a written draft, or on a floppy disk, but too few responses were received. We can only see from table 17 the rough distribution between personal and indirect input.

          Table 17. Mode of sending messages

 

ETH

IOPIA

UGA

NDA 1

UGA

NDA 2

ZAM

BIA1

ZAM

BIA 2

SEN

EGAL 1

SEN

EGAL 2

Total N

100

48

56

23

50

98

100

Oneself

90

55.1

55.3

95.6

88

42

85

Through someone else

10

44.9

30.3

4.4

12

 

15

Both

   

14.3

       

Intensity of use

The intensity of use was first assessed by analyzing as shown in table 18 below the number of messages sent using the traffic data for the reference period (January-February/March 1995) according to a standard formula. Data collected in the second phase are also available for Uganda and Senegal. In Uganda they are based on the average of traffic data for the three months of April, August and October 1996, using the same definitions of intensity as in the firs phase. The intensity was further observed from the responses obtained in the baseline questionnaire, which are presented in the following tables.

The respondents in the survey seem to belong to the more active group of users. It is in a way quite natural that they feel more concerned. The picture resulting from the survey may be quite different from the one which the users' population as a whole could offer. A study in which all types of users would be taken into account could only be made with the full cooperation of the access providers.

Table 18. Intensity of outgoing traffic (in percent)

 

ETH

IOPIA

UGA

NDA 1

UGA

NDA 2

ZAM

BIA 1

SEN

EGAL 1

SEN

EGAL 2

N

206

164

116

159

106

100

Very high

13.1

1.8

0.8

12.6

13.9

15