Time Use by Gender

PREFACE

The main part of the current research project between the Department of Statistics at the University of Dar es Salaam and the Norwegian Historical Data Centre at the University of Tromsø has been a nationally representative time use survey. As can be seen from this draft version of the research report, it has generated information giving a clearer idea about the time that females and males spend on various socio-economic and political activities in the course of a normal day. We believe this information is important for discourses on female welfare, on the female role in community development, and in building knowledge about the female situation more generally. Such knowledge is in great demand in Tanzania, as well as in several other African countries. The time use information will hopefully be used by planners, health officials, community-based and various non-government organizations for developing policies geared towards improving the people's welfare and especially in analyzing and confronting poverty.

We are grateful to the Norwegian Program for Development, Research and Education (NUFU) for the grant making our joint project possible, to the students in the Department of Statistics for carrying out major parts of the survey, and to the Department’s staff for monitoring them and analyzing the data, including paragraphs about the survey methodology.

Dar es Salaam and Tromsø, May 2007

INTRODUCTION

1 Background

The Department of Statistics at the University of Dar es Salaam introduced the first applied Statistics course in the academic year 1989/1990 after twenty two years of offering only highly theoretical courses from 1967 when the department started. The inclusion of an applied Statistics course in the training programme was geared towards the goal of merging theory with practice.

In the academic year 1999/2000, after ten years of offering the only applied course and considering the then changing job environment, it was decided that fieldwork, properly conceived and executed, be part and parcel of a statistics training programme in the department.

The idea of carrying out a study on Time Use by gender immediately came to mind, but could not be pursued further because of financial constraints. Fortunately, in 2002 a call for proposals was made by NUFU and the department considered this an opportune moment to promte its idea. Accordingly, the Department of Statistics at the University of Dar es Salaam, Tanzania and the Norwegian Historical Data Centre, at the University of Tromsø, Norway, entered into useful discussions and submitted a joint proposal on “Country-wide Time Use by gender and advanced analysis of census data” to NUFU.

The proposal was accepted for NUFU funding for a period of five years and the project started being implemented on 1 January 2003.

2 Time Use

The main purpose of the Time Use study was to document information on the time which females and males in Tanzania spend on various socio-economic and political activities in the course of a normal day. This information is important for discourses on female welfare, on role of women in community development, and in building knowledge about the women’s situation as well as these aspects more generally. Such knowledge is in great demand in Tanzania, as well as in several other African countries. Time Use information may be used by planners, health officials, community-based, and various non-government organizations for developing policies geared towards improving the people’s welfare and especially in analyzing and confronting poverty.

Although data on activities by gender may be obtained from census publications and other sources, these sources do not document the time spent on the relevant activities. There was an urgent need to fill this gap and we hope that the current project has initiated the process towards attaining that goal.

2.1 Methodology of the Country-wide Time Use Survey in Tanzania

The Department of Statistics of the University of Dar es Salaam in consultation with the Norwegian Historical Data Centre began preparations for the Country-wide Time Use survey in Tanzania in late 2003.

2.1.1 Pilot survey

One pilot survey on the Time Use was carried out in Morogoro and Iringa regions, from 17 October 2003 to 28 October 2003. On the basis of these experiences, a revised questionnaire was developed.

2.1.2 Study population and geographical coverage of the survey

A sample design was developed aimed at obtaining zonal and national Time Use estimates by gender.

Stratification

The stratification adopted was according to the likely economic activities and thus geographical localities were grouped into zones in which similar economic activities were deemed to be carried out. Six zones were subsequently adopted.

The adopted six zones are as detailed in Table 1 below.

Table 1: Research zones for the 2005 country-wide Time Use main survey in Tanzania

2.1.3 Survey design

The 2005 country-wide Time Use survey was implemented using a questionnaire. The questionnaire has seven main sections covering identification, list of household members, household status, community participation, empowerment, Time Use, and literacy.

Specifically, the identification details included are: Household number, date of interview, place where the interview took place, covering region, district, ward/shehia, village/street, and village/street chairperson. Also included in this section are name of interviewer and specific views of the interviewer on surroundings ad circumstances under which the interview was to take place.

A list of all household members recorded each individual’s personal particulars on relationship to the head of the household, sex, age (in completed years), marital status, educational level, religion, main activity of the previous year, and second main activity of the previous year.

Sampling Design

For the purposes of this study it was decided to adopt a three-stage stratified sampling design. The strata were formed around the six purposively defined zones.

Both the rural and urban samples were obtained by first selecting two districts followed by selecting two villages/branches or streets from each of the selected districts and subsequently by selecting the required households from each sampled village/branch or street.

Sampling frame

The sampling frame for the second stage of sampling was a list of all the villages in a zone for the rural areas and all the enumeration areas in a zone for the urban areas. The cartographic listing for the Tanzania 2002 Population and Housing Census was used for creation of this frame for each zone.

Sampling unit

In rural areas, the second stage sampling unit was a village and in urban areas it was an enumeration area. The ultimate sampling unit in both the villages and the enumeration areas was the household.

Selection of villages and enumeration areas

The villages were selected proportionally to the number of households in the zones. A systematic probability proportional to size (PPS) sampling procedure was employed in the selection of the villages/streets in every zone. Since the zones were such that similar economic activities were deemed to be undertaken within the zone, it was decided that two villages be selected in each zone to minimize the cost of traveling to various villages. The cumulative total method was used to obtain these two villages.

Sample size

In multi-indicator surveys, the sample size is usually determined by using a formula, but

it has also been argued that this textbook approach to sample size determination does not lead in practice to very clear-cut answers because analytic objectives are multiple. A simpler, and often-used, approach is to start from resource constraints. The main constraint for this particular study was funds. Other parameters taken into account to arrive at realistic estimates of sample size were the distribution of student teams between zones and locations, the number of days worked in a week and the length of the survey period.

Basing mainly on cost considerations, 150 households were to be selected from each location of research thus giving a sample size of 3,300 households for the specified 22 locations.

Respondents

It was decided right from the start that two adult members of the household would be interviewed in each household for the entire sample size and that in addition two younger members of the household would be interviewed in each household for half the sample size.

In each household where two adults were to be interviewed, the head of the household and the spouse were to be interviewed. That is, in each household where two adults were to be interviewed, a male and a female were to be interviewed. If the head of the household happened to be a female then her spouse or the oldest male in that household was also to be interviewed. Similarly, if the head of the household happened to be a male then his spouse or the oldest female in that household was also to be interviewed.

In the case where four members were to be interviewed, the two adults to be interviewed were to be chosen in the same way as outlined above and the two youths were to be a boy and a girl selected from the children in the household. Children who were to be interviewed for the Time Use survey were to be of ages five to seventeen.

2.1.4 Data collection

The survey was to be executed in 2004 but due to some unforeseeable circumstances it had to be postponed to 2005. Training of student fieldworkers took place in January 2005.

Data collection extended over the period from 24 January 2005 to 11 February 2005. Unfortunately, a few of the student interviewers could not fill in all the required number of questionnaires either because of the non-availability of the intended respondents or simply non-cooperation of the sampled respondents. Consequently, there were a number of complete non-responses.

2.1.5 Data processing and analyses

2.1.5.1 Data processing

Data processing was done, by the same students who carried out the interviews, under the supervision of the data processing coordinator. The data processing coordinator trained the students in coding and data entry.

2.1.5.2 Data analysis

Data analysis was not fully undertaken but initial frequencies were produced for data cleaning purposes only. This was simply because it was obvious that researchers had different needs and interests. Accordingly, it was decided to provide a copy of the merged raw data file to each core researcher to facilitate data analysis according to individual needs. Indeed, the papers in this volume demonstrate the different data analyses done by the paper writers.

3 Papers in this volume

A total of seven papers basing on the Time Use study results are presented in this volume.

The paper by Rugaimukamu presents a case study of the Lake and Southern Highlands Zones in Tanzania. Basically, the aim of the author is to try to establish a relationship between proper planning of human activities and attainment of a certain set goal - eradication of poverty in this case. The rationale of the paper by Katapa is that teenagers are a special group of people in transition from childhood to adulthood that faces specific problems in the society. The author presents the allocation of time by teenagers to the various activities and explains how these results can help to reveal some teenagers’ problems. Mussa compares the gender time allocation to arbitrarily categorized socio-economic and non socio-economic activities. Naimani’s paper sets out to establish the extent to which activities related to access and use of mass media feature in the peoples’ Time Use. The paper by Akarro looks at the question of whether differences in Time Use contribute to differences in poverty levels in Tanzania. The author uses the Eastern and Northern Zones as case studies for that purpose. Kamuzora introduces an interesting idea of household member activity complementarity. For this purpose the author defines four activity groups and then traces the activity distribution of household members in eight time periods. Mbago’s paper sets out to examine some factors associated with Time Use on non-productive activities as classified by the International Classification of Activities for Time-Use Statistics.