Meetings Programme on Public Statisttics of the Inter-American Statistical Institute (Iasi)

SECOND MEETING ON PUBLIC STATISTICS OF THE INTER-AMERICAN STATISTICAL INSTITUTE (IASI)

WORKSHOP ON HOUSEHOLD EXPENDITURE AND INCOME MEASUREMENT IN A SYSTEM OF HOUSEHOLD SURVEYS

Organized by

NATIONAL INSTITUTE OF STATISTICS AND CENSUSES (INDEC)

INTER-AMERICAN STATISTICAL INSTITUTE (IASI)

I Introduction

II Workshop objectives

Organization

III Organizing and sponsoring institutions

PART 1

I.- Introduction

II.- Discussion Topics

III.1996/1997 National Survey on Household Expenditures

1. Purposes of the survey

2. Population and units under research

3. Main variables under research

4. Measurement of variables and data gathering methods

5. Sample design

6. Survey organization

7. Validation and Editing

8. Data processing

ANNEX A: Reference periods used for data gathering

ANNEX B: Questionnaires contents

ANNEX C:Non response rates

PART 2

Report of Proceedings

Opening of the sessions

Topic 1. Present situation and prospects of Latin American household expenditure and income surveys

Topic 2. Present situation and prospects of household expenditure and income surveys in the European Union

Topic 3. The Argentine National Household Expenditure Survey

Topic 4. Methods for reducing the respondents’ burden

Topic 5. Data-gathering instruments

Topic 6. Frequency of the expenditure and income survey

Topic 7. Reference periods

Topic 8. Quality control and imputation

Topic 9. Non response

Topic 10. Income measurement

Conclusions and recommendations

Closing session

Annexes

1. Workshop aGENDA

2. Invited eXPERST

3. Organizer and special guests

IASI’s Executive Committee

IASI’s Secretary

INDEC Work team

Special Guests from INDEC

Other foreing guests

4. Proceedings team

I. INTRODUCTION

The Inter-American Statistical Institute (IASI) started in 1998 a series of technical meetings dealing with subjects of public statistics, sponsored by the International Statistical Institute (ISI).

As part of this program, IASI organized, jointly with INDEC, a workshop on household expenditure and income measurement in a system of household surveys.

Even when obtaining the weigthing of the Consumer Price Index remains as one of the main goals of household expenditure surveys, there is at present a greater demand, both public and private, of information for research and studies on diferent economic and social aspects of households, specially those related with living conditions.

The structural transformation process of the Argentine economy and society – which reached a higher pace in the 90s due to the implementation of stabilization, privatization, deregulation and opening-up policies – has substantially changed the consumption patterns, income levels and distribution and purchasing habits of households.

These circumstances created the need of reviewing topics, methodology and frequency of household surveys to obtain relevant, reliable and timely data that would allow to monitor living standards evolution, as well as help on measuring of the impact of public policies applied by government, and redesign new ways of action.

II. WORKSHOP OBJECTIVES

v To contribute to a better setting of surveys objectives, methodology planification, and timeliness of household expenditure and income surveys as part of a system of household surveys.

v To contribute to the improvement of technical capabilities of the national statistical officers in Argentina for the design and implementation of household expenditure and income surveys.

v To share experiences with official statisticians from other countries participating in the workshop, and to assess with them possible improvements and changes.

v To facilitate the communication with users of official statitstics.

Issues to be covered during the Workshop:

v Frequency of household expenditure and income surveys

v Methods for reducing respondents’ burden

v Reference periods

v DATA GATHERING DATA GATHERING instruments

v Quality control

v Non-response

v Income measurement

ORGANIZATION

The preparation of the Workshop was under the responsability of an Organizing Committee coordinated by Jacob Ryten, former Assistant Chief Statistician of Statistics Canada, and former President of IASI, other members of the Committee were Evelio Fabbroni (Technical Secretary of IASI), Hugo Ramos (Assistant Technical Secretary of IASI), Carmen Dopico (INDEC), Alfredo Monza (MECOVI), Clyde Trabuchi (INDEC), Adriana Semorile (INDEC), Norma Pizarro (INDEC), and Heber Camelo (ECLAC – INDEC).

Experts were invited to take part in the Workshop to debate different aspects of a technical document prepared by INDEC as a basis for discussion. The panel was coordinated by Jacob Ryten, assisted by Pilar Martín Guzmán, President of IAOS and former President of the National Statistical Institute (INE) of Spain. The invited experts were Mary Mc Carthy (U.S. Bureau of Labor Statistics), Alison Hale (Statistics Canada), Johanne Tremblay (Statistics Canada), Reginald Gatenby (Great Britain’s ONS), Céline Rouquette (INSEE of France), Paloma Seoane (INE of Spain), and Juan Carlos Feres (ECLAC).

III. ORGANIZING AND SPONSORING INSTITUTIONS

National Institute of Statistics and Censuses:

The National Institute of Statistics and Censuses (INDEC) is the official institution that designs and coordinates the production of statistical information on public interest matters. Under normative centralization and executive decentralization principles, established by law 17.622, INDEC carries out the Annual Program of Statistics and Census and developes methodologies and standards that make possible the comparability of the information coming from different sources for its later consolidation and publication. INDEC also conducts population and housing and economic censuses, and elaborates statistics on prices, employment and unemployment, family consumption, productive activities, foreign trade, population projections, etc.

International Statistical Institute:

The International Statistical Institute (ISI) is one of the oldest international scientific associations operating in the world. Established in 1885, the ISI is an autonomous society that seeks the development and improvement of statistical methods and their application throught the promotion of international cooperation.

Inter- American Statistical Institute:

The Inter-American Statistical Institute (IASI) is a professional organization established in 1940 with the purpose of promoting statistical development in the American region. It seeks the following objectives: a) Development and strengthening of the statistical profession; b) Promotion and dissemination of advances in statistical theory and methods; c) Improvement in the methodology for production of both governmental and non-governmental statistics; d) Promotion of measures aimed at improving the comparability and availability of economic and social statistics among the nations of the region; e) Cooperation with national and international organizations in activities aimed at statistical improvement in the region.

IASI has individual and institutional members. INDEC is one of the institutional members, with the status of Affiliated member. IASI is an Affiliated member of the International Statistical Institute.

PART 1

I.- INTRODUCTION

Since 1924, several urban households expenditure and income surveys have been carried out in Argentina. Nevertheless, the frequency has not been regular, ranging between 5 and 15 years, whereas the geographical coverage has been generally focused in the Greater Buenos Aires. In 1996/1997, the National Institute of Statistics and Censuses (INDEC) carried out the first National Survey of Household Expenditures and Income (ENGH) based on a representative sample of the localities of 5,000 inhabitants and more from all over the country (86 per cent of the overall country population).ENGH not only provides national and regional estimations on the expenditure patterns and income levels of urban households, but also the basic information needed for the estimation of several essential indicators, such as a national consumer price index, urban poverty lines of the various regions of the country, and the economic activity levels of certain services, both for the estimation of the Gross Domestic Product and the Gross Geographical Product.

In Argentina, recent experience shows a higher demand from the statistical system and from society, for the type of data collected by household expenditure and income surveys. The structural transformation process of the Argentine economy and society – which reached a higher pace in the 90s due to the implementation of stabilization, privatization, deregulation and opening-up policies – has substantially changed the consumption patterns, income levels and distribution and the purchase habits of households. Concerning this, it should be mentioned, for instance, the effect of price stability on the access of households to loans for the purchase of durable goods (home appliances and cars) and to mortgages for the purchase of housings; the access, as a consequence of privatizations, to services previously rationed (telecommunications); the greater diversity of available consumption goods derived from the opening-up process; and the changes in the purchasing habits of households as a result of a greater concentration of retail outlets.

At the same time, the significant changes that took place in the work market, such as the increase in participation rates, particularly for women, and the increase of unemployment and underemployment rates, had a great impact on the levels and distribution of households welfare, stirring a strong interest on ENGH microdata.

Within this context of accelerated changes in the living conditions of households and, therefore, of larger data demands on the former, INDEC intends to carry out a deep review of the ENGH methodology, in order to improve the quality of the information, to regularize and increase the frequency of household income and expenditure measurements, and to extend its coverage to rural areas, while optimizing the use of budgetary resources. On that regard, INDEC considers that learning from experiences of household expenditure and income surveys around world is an essential factor in the methodological revision process. For this reason, the INDEC, together with the Inter-American Statistical Institute, decided to organize a Workshop on Household Expenditures and Income Measurement, so as to create a favorable context for the international exchange of these experiences, in a way that every participant can profit from it.

This document is submitted in order to motivate the discussion within the Workshop. It is divided into two parts: the first oneoutlines the Workshop discussion topics, and the second one describes the essencial features of the 1996/97 ENGH.

DISCUSSION TOPICS[1]

1. FREQUENCY OF EXPENDITURE AND INCOME SURVEYS

The expenditure survey in Argentina collects data on all household expenditures and income. This survey is in field during a whole year. The last survey was carried out in 1996/97 (ENGH) and, for the first time, had a national coverage, aimed at the urban population of localities with 5000 inhabitants and more.

At present, the need of having more frequent information is being verified as to: 1) the expenditure structure of consumer units for the revision of the coefficients of the consumer price index; 2) the impact of the acceleration of socio-economic changes on household consumption patterns; 3) the modification of the composition of consumption service expenditures aimed at the estimation of the activities related with the former in national accounts.

International Experiences:

· In Canada, the Food Expenditure Survey is carried out during a whole year, every four years, by direct recording of the household during two consecutive weeks. The Survey of Household Spending, in turn, records an overall amount of food expenditures, is conducted during three months, every year, by means of an interview, and collects data on yearly expenditures.

· The United States has a continuous Consumption Expenditure Survey since 1980. It comprises two surveys with different DATA GATHERING DATA GATHERING techniques and different samples. The first one (Quarterly Interview Survey) is a panel interview survey which collects disaggregated information on nonfood expenditures and only overall food expenditure amounts. Each consumption unit of the sample is interviewed every three months during five consecutive quarters. The second one (Diary Survey) is a survey intended to obtain a record of high-frequency and small-amount expenditures, which usually are difficult to recall. Such items include a detailed record of home and away-from-home expenditures on food and beverages, cleaning products, over-the-counter medicines and personal care products. Nonetheless, this survey (Diary Survey) not only records these types of expenses but also any expenditure made during the weeks of the survey.

· In Spain, two family budgets surveys coexisted until they merged in 1997: the Continuous Survey on Family Budgets (ECPF), and the Basic Survey on Family Budgets (EBPF), which was performed every 10 years (the last one, in 1990-91). Taking into account the need to harmonize the Family Budgets Surveys (EPF) under the coordination of EUROSTAT, a new project of a Continuous Survey on Family Budgets was fostered. The new ECPF tries to achieve the goals of the two surveys it replaced.

· France conducts an income and expenditure survey (Family Budget Survey), every 5 years, since 1984.

· In other member countries of the European Union, there are generally two typical patterns: annual surveys and surveys that are carried out at five-year or longer intervals. The survey period (period in which all data is collected) is usually of one year and in general it coincides with the calendar year. All annual surveys are continuous in the sense that the fieldwork takes place on a continuous basis throughout a year.

Questions:

1.1 Which are the advantages and disadvantages of conducting a single continuous survey?

1.2 Which are the advantages and disadvantages of combining a continuous survey with a noncontinuous one with different thematic coverage and samples?

1.3 Are the estimations of continuous surveys better than the other ones?

1.4 Which is the recommended frequency in the case of noncontinuous surveys?

1.5 Can surveys with different scopes and disaggregation level of expenditure recording be combined for results estimations? Can we simultaneously use data from both surveys in order to make consistency analysis or to impute missing data?

2. METHODS FOR REDUCING RESPONDENT´S BURDEN

The ENGH made a detailed research of all income and expenditure items. Each household in the survey was under study for a complete week, during which three interviews of an estimated length of three hours were performed. The daily questionnaire was completed directly by the respondent. Questionnaires on Household Characteristics, Miscellaneous Expenditures and Income (one for each income earner) were completed by the interviewer.

Respondents showed a gradual fatigue due to the length of the interviews; some households had difficulties to complete the daily questionnaire and showed reluctance to give information about their income in a disaggregated way.

International Experiences:

· Canada mainly uses global questions, with the exception of food expenditures, that are recorded in detail by the household.

· United States uses global questions combined with questions that inquire about characteristics of the purchased good. The daily questionnaire was recently modified, incorporating more disaggregated expenditures and more detailed orientations for the respondents.

· In Spain, households do not take part in the survey during all quarters with the same intensity. During four quarters they cooperate intensively and during the other four they do it in a “weak” way. “Intense Cooperation” means that during the quarter, households must take note of expenditures of all periods of reference, whereas “weak cooperation” means that during such quarter, households must only take note of expenditures with a period of reference larger than a week.