National Statistical Agencies Research Articles

Integration of geography and statistics

The rapid development in mapping technologies has visibly upgraded the power of statistical analyses and their dissemination. Focused on spatial dimension and using diverse remote sensing methods, geographic information systems (GIS) make the integration of statistics borderless and statistical analysis more value-added, permitting the evaluation of statistical data across traditional subject-matter fields. Management of statistical offices is increasingly aware that the integration of socio-economic and geographical information, when combined with GIS, assists in decision-making across subject-matter fields affecting education, health, environmental protection, municipal services, economic development and location of essential infrastructure. Internet and Intranet accelerate users’ access to these systems and transparency is a key word for any geographically related information. In addition to these strong points, however, many challenges still remain to be explored by national and international statistical agencies to make GIS applications efficient and safe. The Statistical Division of the United Nations Economic Commission for Europe (UN/ECE) is conducting a region-wide transfer of know-how on the use of mapping technologies for the collection, production and dissemination of statistical data. As well as regular meetings in this field, the Statistical Division is organising the international work on the development-related methodological materials and guidelines. All activities are undertaken in close cooperation with Eurostat, OECD, IMF, UN Statistics Division and other international organisations involved as a part of the programme of work of the Conference of European Statisticians. The aim of this paper is to report about the major objectives and activities going on in this field and to draw attention to some national and international projects on the use of mapping technologies for statistical purposes.

The Role of Behavioral Survey Methodologists in National Statistical

SummaryThe expertise of behavioral survey methodologists is needed in national statistical agencies because of the necessity of using theory and research from the social sciences to reduce survey error. In this paper various social science based explanations for measurement error and nonresponse error are described in order to illustrate the conceptual foundations of such error reduction efforts. Three roles for behavioral survey methodologists in national statistical agencies are then discussed. They include: 1) bring an error reduction perspective to bear in an influential way for all aspects of designing and implementing agency surveys; 2) bring theoretical efficiency and effectiveness to experimental tests of alternative questionnaire designs and implementation procedures through the use of theories, concepts and pretests and findings of past behavioral science research; and 3) contribute to the expanding science of survey methodology.

The Role of Behavioral Survey Methodologists in National Statistical Agencies

The Role of Behavioral Survey Methodologists in National Statistical Agencies

External Statistical Data: Understanding Users and Improving Access

Every year, government agencies, financial institutions, and other businesses spend millions of dollars gathering and publishing statistical data, presumably intended for use by other organizations in planning. From the point of view of the end user, these data are external statistical data (ESD) because they are provided by external sources over which the end user has little control. Typical sources of ESD include national statistical agencies, such as the Australian Bureau of Statistics (ABS), regional information services, and various state and national government departments.

To Cut or Not to Cut

This article provides an empirical assessment of the social psychological underpinnings of arguments that link labor market security to wage flexibility. We employ data from two surveys of blue-collar manufacturing workers, one from Canada and the other from Sweden, that were conducted by the respective national statistics agencies. The dependent variable is a dichotomy tapping respondents' willingness to take a wage cut to save their jobs. Independent variables index a variety of objective and subjective labor market conditions and experiences. We test two hypotheses. The first, that Canadians should be more willing than Swedes to accept a pay cut, is amply supported. The second, that observed differences between Swedes and Canadians should be due to higher levels of anxiety about labor market prospects among the latter, found no support whatsoever in our logistic regression analyses. We conclude with speculative alternative explanations and suggestions for additional comparative research.

Models of complex human screening and correcting of social data

National statistical agencies annually spend great budgets in the continuous collection of a huge amount of data on individual persons, households, business activities, etc., to serve the information needs of a number of national and international government and private users. Substantial parts of their budgets are consumed in checking and improving the quality of the data collected. Because of their complexities, these tasks have depended on the handling of specialists. To save both processing time and resources and to improve the processing involved in solving and servicing data requests, these tasks have high priority. The present paper outlines research carried out in Norway on using the neural network paradigm to improve the data quality checking and improvement in large-scale data masses.

Socio-Economic Statistics and Public Policy: A New Role for Microsimulation Modeling

Users of socio-economic statistics typically want more and better information. Often, these needs can be met simply by more extensive data collections, subject to usual concerns over financial costs and survey respondent burdens. Users, particularly for public policy purposes, have also expressed a continuing, and as yet unfilled, demand for an integrated and coherent system of socio-economic statistics. In this case, additional data will not be sufficient; the more important constraint is the absence of an agreed conceptual approach. In this paper, we briefly review the state of frameworks for social and economic statistics, including the kinds of socio-economic indicators users may want. These indicators are motivated first in general terms from basic principles and intuitive concepts, leaving aside for the moment the practicalities of their construction. We then show how a coherent structure of such indicators might be assembled. A key implication is that this structure requires a coordinated network of surveys and data collection processes, and higher data quality standards. This in turn implies a breaking down of the stovepipe systems that typify much of the survey work in national statistical agencies (i.e. parallel but generally unrelated data--production lines). Moreover, the data flowing from the network of surveys must be integrated. Since the data of interest are dynamic, the proposed method goes beyond statistical matching to microsimulation modeling. Finally, these ideas are illustrated with preliminary results from the LifePaths model currently under development in Statistics Canada.

Network Models for Complementary Cell Suppression

Abstract Complementary cell suppression is a method for protecting data pertaining to individual respondents from statistical disclosure when the data are presented in tabular form. Several mathematical methods for complementary suppression have been proposed in the statistical literature; some have been implemented in large-scale data processing environments by national statistical agencies. Each method has either theoretical or computational limitations. This article presents solutions to the complementary cell suppression problem based on linear optimization over a mathematical network. These methods are shown to be optimal for certain problems and to offer theoretical and practical advantages, including comprehensiveness, comprehensibleness, and computational efficiency.

Network Models for Complementary Cell Suppression

Abstract Complementary cell suppression is a method for protecting data pertaining to individual respondents from statistical disclosure when the data are presented in tabular form. Several mathematical methods for complementary suppression have been proposed in the statistical literature; some have been implemented in large-scale data processing environments by national statistical agencies. Each method has either theoretical or computational limitations. This article presents solutions to the complementary cell suppression problem based on linear optimization over a mathematical network. These methods are shown to be optimal for certain problems and to offer theoretical and practical advantages, including comprehensiveness, comprehensibleness, and computational efficiency.

Issues relating to the UN system of national accounts and developing countries

The recent revision of the United Nations System of National Accounts (SNA) represents improved harmonization of the international trade, balance of payments, government finance and monetary data collected by national and international statistical agencies. Although such harmonization is welcomed by the developing countries, a number of other major unresolved issues remain. Specifically, the SNA has not solved problems relating to (1) delineation of the formal and informal sectors of an economy, (2) appropriate treatment of market and non-market production, and (3) satisfactory analysis of sector saving and capital formation and (4) harmonization of household sector macrodata with household survey microdata.

Quality Management: Development of a Framework for a Statistical Agency

Over the last 20 years there has been an increasing focus on quality management (QM) as a distinct and necessary element of any competitive business organization. Although quality has always been a concern of national statistical agencies, quality improvements within such organizations have been achieved through a variety of different independent initiatives rather than through concerted effort under the umbrella of a formal “QM program.” This article addresses the application of QM to a national statistical agency. The essential elements of QM are reviewed from the perspective of a statistical agency. Ongoing quality initiatives at Statistics Canada and at other statistical agencies are described. The article concludes by identifying some of the major considerations in developing a QM strategy for Statistics Canada.

Improving Migration Statistics: Policy and Conceptual Issues

Abstract. The availability and international comparability of migration statistics pose a severe handicap to analysis and policy debate. This paper explores a few of the underlying policy and conceptual issues at stake and identifies some of the institutional factors contributing to the slow and uncertain in progress in this field.Three main issues are discussed: the statistical definitions of migration; the administrative sources of data; and the international initiatives to improve the comparability of statistics.At the beginning of the 1990s the context appears promising for a renewed effort of national and international organizations and statistical agencies. New opportunities and challenges for the improvement of migration statistics are being developed.

Modelling Job‐Search Duration from the Italian Labour Force Data

ABSTRACT: Micro (and possibly panel) data from the Labour Force Surveys are a valuable source for analysing the labour supply behaviour. Among the information collected by the quarterly Italian Labour Force Survey conducted by ISTAT (the Italian national statistical agency), there are data on the duration of job‐search.From a single survey, one can only know the duration of search up to the time of the interview. Nonetheless, by appropriately exploiting the longitudinal structure of the survey (the sample scheme is a rotated one, with a 50|X% overlapping of the sample units in two subsequent quarters) it is possible to measure, although imperfectly, some completed spells of job‐search.In this paper, we try to model the duration of job‐search for a sample of young unemployed. Considerable attention is devoted to the measure of duration of search, and to the evaluation of its accuracy. A one state proportional hazard model is specified and estimated, accounting also for unobserved heterogeneity.The model is estimated on a sample of young (15 to 29 years old) unemployed in January 1986 reinterviewed in April, in an Italian region (Lombardy).The major finding is a fairly clear evidence, both for males and females, of negative duration dependence, even after accounting for the unobserved heterogeneity that turns out to affect the female population.

ADP strategic planning in a national statistical agency – An overview

ADP strategic planning in a national statistical agency – An overview

NEW DEVELOPMENTS TOWARDS RESOLVING THE COMPANY‐ESTABLISHMENT PROBLEM*

The paper begins by stating various aspects of the national economic accountant's “company‐establishment problem.” Six possible approaches to the problem are briefly outlined. The paper concentrates on one approach based on new developments in business accounting theory and practice, namely divisional‐reporting procedures. The division represents the smallest operating entity capable of reporting both a complete set of production (income) statistics and a set of related financial (balance‐sheet) statistics. When companies are owned and controlled by the same interests, namely the enterprise, each division reports on an enterprise‐wide basis. In this important case, the traditional company‐establishment problem has an enterprise‐division‐establishment resolution.There is considerable emphasis on clarifying the issues needed for systematic development of divisional‐reporting to meet the requirements of a national statistical agency. Key aspects are the provision of appropriate conceptual distinctions relating to statistical structure of corporate organizations and patterns of intercorporate ownership consolidation. Practical experience gained by the U.S. Federal Trade Commission's line of business reporting program is also highlighted. Two tables show details with respect to a proposed divisional income statement and balance‐sheet statement that a systematically developed division‐reporting unit can provide. The tables are related to existing statistics yielded by traditional company‐ and establishment‐reporting units. In effect the paper is part of a movement giving national economic accounting more microdata dimensions. Future research must integrate the proposed new statistical reporting unit within systems of national accounts presently constructed on the basis of a dual sectoring classification.

Probabilistic forecasts. The case of population projections

Probabilistic models for assessing the random error and the bias of population forecasts are discussed. The predictions of the models reflect the “expert opinion” about the future vital rates, and the widths of the prediction intervals are calibrated using the performance of past forecasts. An application with the U.S. data is presented and comparisons with the situation in Nordic countries are made. It is found that the interval between the high and the low birth forecasts of the national statistical agencies in the United States, Finland, and Norway are narrower than our 67% prediction intervals. The U.S. intervals are by far the narrowest.

Response Burden, Sample Rotation, and Classification Renewal in Economic Surveys

This paper, which is addressed to the business and institutional survey operations of national statistical agencies, is a proposal for a sampling scheme and a system of coordinated samples selected according to this scheme. By presenting an analogy between the taxation system (or the duty of governments to collect taxes and the obligation of businesses to pay them) and the statistical system (or the necessity of governments to collect information and the obligation of businesses to provide it) the paper provides a basis for a definition of response burden and the management of a system to control its equitable distribution. Sample rotation is shown to be simple to manage within the proposed system. Since a consequence of maximal distribution of response burden is maximal coverage of the survey population in any given period the system provides an opportunity for continuous updating of the survey frame with respect to industrial classification and other aspects subject to deterioration over time. Finally, the paper suggests that a statistical agency that can demonstrate the fairness of its distribution of response burden is in a better position to insist on the legal obligation of response to its surveys.

Improving the accessibility of data from the census of Canada

Improving access to its data resources can be considered one of the major objectives for the Census of Canada. The samc could be said for the wide range of statistical activities undertaken by Statistics Canada, the National Statistical agency in Canada. Obviously, accessibility is only one consideration, and it must be viewec jointly with the needs of users and the actual packaging of the data products The Census data resource, aside from its statutory purpose for establishing Parliamentary Representation, has a very-wide and varied audience. This factor creates substantial challenges in trying to establish some (sub) optimal mix for the contents of the preplanned outputs. With respect to outputs, we have to be concernec in the public interest with the cost aspects to provide rational substantiation of the funding, and with the practical limitations on the organization's ability to actually produce the materials in acceptable time frames. We have also become exposed in the current time frame (with the 1971 Census of Canada) to the full force of significant changes and technological expansion in the area of automated information processing. These changes, as many people realize, have not been easy to keer up with nor have they been the solution tc all our problems. All otf a sudden we are in a new ball game, and the comfortable stability gained from experience with past errors and conventional products and methods of producing the latter is now overshadowed by the need to gain at least an appreciation for the new technical operating environment and the development of some competency in assessing the impact of such technological changes, not only for internal operations but in the user environment. This paper deals with the variety of steps taken to address the problems of accessibility. Some of these were by design and planned; others, however, were developed by dynamic adjustment to the situations and environment. The development of technologically current products and innovative approaches to the process of dissemination of information are essential if there is to be any degree of compatibility between the users' application needs and the nature of products and services available. Improvements in accessibility will be determined, in general terms, by the degree to which one can expand and communicate the collective knowledge about the information readily or potentially available and the degree to which the potential of the data base(s) can be exploited. Data availability and the facility to

Production planning and control in a national statistical agency

The collection, analysis and dissemination of national statistical information can be likened to a manufacturing process. Despite obvious differences, the strength and extent of the analogy is such that it provides a powerful argument for the organisation of the production of statistics along industrial lines. This paper describes the implementation of this approach in Statistics Canada, the Canadian national statistical agency. It discusses the problems which indicated the need for action and the proposed activities of the centralised Production Planning and Control Branch (now in operation).

Measurement Error and Choice of Econometric Estimation Method: Some Empirical Findings

The question of what procedure to choose for estimating a given simultaneous-equation model is a basic one in applied econometrics. Considerations of computing cost, small-sample properties of estimators, and the sensitivity of different estimators to specification error are all relevant. Also relevant is the sensitivity of estimators to errors of observation or measurement. Do measurement errors in the data tend to distort some estimators more than others? How important in practice is the problem of measurement error in relation to the more widely discussed problem of choosing among alternative estimators ? The purpose of this paper is to present some results bearing on both of these questions. There are two ways of approaching the questions. One is to carry out Monte Carlo experiments with artificial data.' The other and this is the one followed here is to seek and make use of real-world data for which there is some information about actual errors of measurement.2 Both approaches have advantages and disadvantages. The particular advantage of the second approach is obvious: it produces results based on actual errors rather than artificial ones for which distributions must be assumed. On the other hand, a major disadvantage is the scarcity of data for which anything is known about errors. The presence of measurement error in economic statistics is widely recognized, of course, but specific information is seldom available. One case for which some information is available is the case of preliminary national accounting data. The United Nations Statistical Office compiles and publishes, in standardized form, annual national accounts estimates for a large number of countries. Each year the estimates for previous years may be modified by the national statistical agencies reporting to the U.N. as new or improved information becomes available. Undoubtedly, even the most thoroughly revised figures are still less than perfect but presumably they are better than the preliminary ones. In this sense, the differences between preliminary and revised data may be viewed as measurement errors of a kind. What we have done in this study is to specify a small model, estimate this model using preliminary national accounting data for each of 21 countries, and then re-estimate the same model for the same time period using revised data. The characteristics of the data revisions themselves and their effects on parameter estimates obtained by two-stage least squares have been reported in detail in Denton and Oksanen [4]. In the present study, we have estimated the model by ordinary least squares (OLS) and three-stage least squares (3SLS), as well as two-stage least squares (2SLS). The focus here is on (a) the relative sensitivity to