Statistics systems are either centralised or decentralised. In a centralised system, there is a single authority with a legal mandate for collecting statistics and this institution collaborates with others by formal arrangements on the collation of statistics. The Handbook of Statistical Organisation defines such organisation as follows “A system…of one department within the government to organise and operate a scheme of co-ordinated social and economic statistics pertaining to the whole country. This department collects, compiles and publishes statistical information …and, in addition collaborates with other departments of government in the compilation of administrative and specialised statistics.”3
Examples of countries with a centralised system are Australia, Canada and South Africa. Those known to have a decentralised system are Japan, America and the United Kingdom (UK), the latter until recently and is moving towards a centralised system.
What are official statistics:
Official statistics are defined by the United Nations Statistical Commission as “an indispensable element in the information system of a democratic society, serving the Government, the economy and the public with data about the economic, demographic, social and environmental situation. To this end, official statistics that meet the test of practical utility are to be compiled and made available on an impartial basis by official statistical agencies to honour citizens’ entitlement to public information.”4
The White Paper on official statistics in the UK define them as those statistics that “are collected by government to inform debate, decision making and research both within government and by the wider community. They provide an objective perspective of the changes taking place in national life and allow comparisons between periods of time and geographical areas.”5
“Open access to official statistics provides the citizen with more than a picture of society. It offers a window on the work and performance of government itself, showing the scale of government activity in every area of public policy and allowing the impact of public policies and actions to be assessed. Reliable social and economic statistics are fundamental to ...open government (and) it is the responsibility of government to provide them and to maintain public confidence in them.”6
An important aspect of official statistics is the trust that the participants accord the system. To retain trust, the agency should decide on the basis of professional considerations, scientific principles and professional ethics, what methods and procedures should be followed for the collection, processing, storage and presentation of statistical data.
Statistical development: an international perspective
In this part of the paper we explore statistical developments in a number of countries in particular, the Americas, Oceania, Europe and Africa. Finally we focus on South Africa.
It must be noted that the evolution of official statistics in each country is mainly a product of its history. A recurring feature across the country studies is the dominance of one of the three patterns in the development of their statistics. Firstly, we observe a pattern in a group of countries where statistical development has generally evolved over time without any direct intervention by the political leadership. In these countries the statistics systems succeeded or declined over time. Success stories of statistics systems where this pattern was predominant are Australia, Sweden and Canada. A decline in the statistics system has been observed largely in countries on the African continent. The UK has also had a tumultuous era in the development of its statistics system which was precipitated during the reign of Thatcher. A second pattern is that of countries where statistical development had advanced to impressive levels but collapsed dramatically towards the end of the 1980’s. In this category are the Eastern Bloc countries and China, which through their centralised planning approach developed a sophisticated battery of statistics systems for centralised economies but faced an unprecedented onslaught with the demise of communism and the emergence of the market economy. The third pattern of statistical development features in most of the third world countries. This came about with the advocacy for writing off debt in respect of the highly indebted and poor countries. With the movement towards addressing debt relief, statistical organisations are beginning to adopt strategic plans as the means of managing their affairs. In this countries we observe political leadership taking the lead to address issues of information and statistics. In this regard PARIS21 continues to play a positive catalytic role. Zambia, Mozambique, Uganda and Malawi are some of the countries that have adopted this approach. The fourth pattern of statistical development consists of a situation where countries without addressing matters of debt relief, discover that it cannot be business as usual; democracy, informatisation and globalisation demand of the political leadership to adapt strategic plans and follow them through. As a consequence the leadership put statistics at the top of the agenda. South Africa, Korea and some of the Latin American countries fall in this category.
The 1990’s witnessed relentless efforts towards reform in government systems, and since 1995 the Government of Chile had been committed to a process of state modernization. The management of the National Institute for Statistics (INE), like the Australian Bureau of Statistics, as we will see later responded to the historic importance of this initiative.
An important initiative was recognising that for INE to succeed in its modernising programme, it was essential to ensure the full commitment and involvement of staff. As a result, the staff of the institute was brought in to form joint task teams carrying out an organizational diagnosis. These teams were supported by the specialist work of a team of professionals assigned this purpose in 1998. The diagnostic exercise focused on the assessment of critical areas of management, namely: Human Resources Quality, Management Styles and Strategic Planning.
In addressing these four areas INE, ascertained that if it is to achieve quality statistics, essential skills for driving statistics had to be acquired, mastered and assimilated in the Institute. This was achieved through implementing vocation for statistical work at INE.
A key management area for analysis was Human Resources Quality. The main effort here went into ascertaining the degree to which three essential skills, regarded as the critical ones were to be addressed. Firstly, effort was put on the quality of human resources to ensure that they have been assimilated and mastered in the Institute. Secondly, focusing on responsibility and commitment towards society, which in the main emphasises the importance of statistical work as an aspect that requires to be nurtured. Thirdly, vocation for public service which emphasises user orientation in the production of statistics has been implemented. As a result of addressing this area of work a common vision between management and staff was forged.
Management Style was the second critical area analysed within INE. As regards this aspect, the diagnosis pointed out to a practice that focused on procedures and oversight of tasks and activities than towards processes and empowerment. A new management style was implemented which emphasised participation. The third critical management area analysed was Strategic Planning and the diagnosis indicated complete paucity of use of modern tools required for planning. The new design brought about a flatter structure which applied “values of quality, transparency, teamwork and respect for individuals in each and every project undertaken. It is a style that values creativity, responsibility, innovation and criticism in a context where users and their requirements are accorded the highest priority.”7
The changes in structure were followed through with changes in the budgeting procedures and adopting new technological platforms such as the incorporation of the intranet in the work process and extending the facility to the regional offices. The benefits have been in notable improvements in the quality of decisions being made. By automating processes the quality of data from field has also improved dramatically.
In Mocambique, the National Statistics Institute (INE) was created by Presidential Decree 9/96 of August 28th. INE is the central executive body of the National Statistics System (SEN) and it is charged with the production and publication of the country’s official statistics. The INE is an autonomous institution that reports to the Council of Ministers. The SEN’s constituent bodies include the Senior Statistics Board (CSE), the General Population and Housing Census Co-ordination Board, the National Statistics Institute (INE) and the Bank of Mozambique.
The creation of the SEN and the INE came as a sequel of the economic, social and political transformations that started, in particular, with the introduction of the Economic and Social Rehabilitation Plan in 1997. The structure, functions and results of official statistics activity came as a direct response to the new era of multi-party democracy, peace and market economy that emerged in Mozambique.
Within this short space of time INE conducted the 2nd General Population and Housing Census in 1997, the National Household Survey in 1996-97, the production of the Consumer Price Index, the Demography and Health Survey in 1997 and the re-launch of general and specialised statistical publications.
It is observed that INE increased its human resources, recruiting technical personnel with university degrees and promoting several short- and long-term training courses in order to meet its new challenges. The status of INE was administratively promoted to a deputy minister position, similar to the situation in Korea.
A recurring theme in country experiences is training of staff in order to meet emerging challenges. INE carries out a prospective appraisal of the human resource requirements for the National Statistics System. This is particularly laying emphasis for recruiting and retaining better-qualified and specialised staff. They promote the progressive creation of working conditions – organisational, functional and operational – that will help recruit and select personnel. They also promote co-operation with other divisions in the development and implementation of training programmes in priority areas.
As regards infrastructure, INE promotes the progressive establishment of conditions that will modernise the NSS and this is done through a strategic information systems development plan.
Furthermore, INE aims at increasing the geographic decentralisation of the INE (deconcentration) by adequately equipping the Provincial Delegations. INE intends to set up information use committees within the divisions of the appropriate central services and carry out studies and hold seminars concerning user requirements.
To date the Zambian statistical system is thirty eight years old. It came into being by an Act of -Parliament Cap 425 of the Laws of Zambia in 1964. Under this Act the Director of the Central Statistics Office (CSO) is mandated to conduct all censuses and surveys and to organise a coordinated scheme of social and economic statistics relating to Zambia. CSO like many other institutes is a corporate and legal body. Furthermore it is also an autonomous body except for matters concerning human resource. Under the current set up, CSO falls under the Ministry of Finance and Economic Development (MOFED) with Budget and Economic Affairs Division.
The statistical development in Zambia followed a centralised trajectory which was consolidated in 1970, when the unified statistical information system came into being. This process involved the secondment of CSO staff to various line ministries. A similar approach was followed in Malawi until the time when budgetary constraints undermined this approach leading to abandoning of this approach. Currently Malawi is reviving its system and is captured in the 2002-2006 Strategic Plan for statistics in Malawi.
The current status of the statistical system in Zambia consists of the CSO, the statistical units in the line Ministries and parastatals, the University Research Institutes, the Bank of Zambia and the Statistical units in the private sector organizations and NGOs.
In their recent User producer meeting the delegates concluded as follows: “A Standing committee of CSO, BOZ and MOFED should be in place to work on macro-economic data required for developmental policies.
Data analysis should be undertaken by the users according to their needs but in cases of low capacity there should be collaboration with the producers to carry out necessary analysis. CSO should give leadership in this task.
There is need for relevant Institutions to meet regularly to discuss their data needs as well as means of collection and analysis. Committees could be established made up of all stakeholders and these committees should be led by the CSO. This should lead to the establishment of better and sustainable dialogue between users and producers.
There should be coordination between the producers of data and donors in funding the establishment of databases into which all sectors could tap for policy formulation and monitoring.
There is need to integrate both the methods and data in the measurement of Poverty. In this regard, Quantitative and Qualitative approaches should be examined for Integration.
Data should be made more accessible by “going beyond talking to like-minds” through the presentation of complex data in a simple manner to ordinary Zambian.
Need to re-establish a coordinated national statistical system.
Need to re-establish the System of Administrative Statistics (SAS) by re-building statistical teams in the ministries with close links with the CSO.
Establish an Inventory Library at CSO as well as electronic database containing abstracts of what is available (and their data sets).
Necessary to develop capacity building programmes in the areas of data analysis, dissemination, coordination, data management for producers and users of data and top policy makers.
There is need to revise the current statistical Act of 1964 so as to provide an appropriate legal framework.
It is recommended that a strategic plan to implement a statistical framework be developed.
CSO and MOFED should initiate establishment National Statistical Coordinating Committee.
It is recommended that Poverty desks or focal point officers in all Institutions be established.
There is need to evidence based policies that are technically sound, growth oriented, poverty reduction focused and politically achievable and implementable.
A national forum on poverty should be held.
There is the need to strictly monitor the expenditure on Poverty Reduction Programmes of the debt relief fund. To ensure transparency of the monitoring system, Private Sector and the Civil Society should be encouraged to participate in the monitoring.
Carry out a comprehensive statistical needs assessment exercise.”8
The National Statistics Office of Malawi is directly responsible for all statistics produced by government ("official statistics"). Despite good co-operation and collaboration amongst those involved in the production process, this is largely ad–hoc and mechanisms for preventing overlaps, gaps, or incompatibilities in the provision of statistics need to be formalised and implemented. Although the 1967 Statistics Act determines the legal responsibilities of the Commissioner of Census and Statistics, the provisions are outdated and largely not enforced.
Part of the reform process in Malawi for statistics is legislative reform, defining roles and responsibilities of the National Statistics Committee, compiling an inventory of statistics that are required, develop and publish a framework for national statistics and co-ordinate standards and classifications and promote the use of statistics. Furthermore promote training and publicise the training programme.
It is further believed that there is need for the NSO to become a semi-autonomous agency, governed by a revised National Statistical Act and accountable to a Management Board. This suggests a closer organisational pattern to that of Uganda where the statistics office is managed by a board of directors.
Australia is one of the top three countries noted for their good statistics systems. At one point in time the outgoing Commonwealth Statistician reformed the statistics system of the UK. The Australian experience was initially dominated by its colonial history beginning in 1787 whereby it was demanded of the system to submit regular reports on the number of people and animals by the UK, her colonial master. Over time the Governor of the colonies began to use the data for managing the colony and this became important when the self-government was granted in 1855.The UK required regular reports and the colonies obliged, often by conducting “musters” to count the people and the animals. Thus began the centralised collection of data. However, the Governors of the colonies gradually began to use the data to help with the management of the colonies, and this became particularly important from 1855 when self-government was granted. Fifty years later in 1905, the Census and Statistics Act created a Federal Bureau charged predominately with conducting population and agriculture censuses, and coordinating issues with the State Bureaux. As a consequence of this, there was rapid expansion of the system into areas of macro economic statistics following the great depression in the 1920s.
It took another fifty years, in 1970, for the Australian Government to establish an independent central statistical authority, known as the Australian Bureau of Statistics (ABS). This was to be a user oriented institution that would serve the governments and the community as a whole. Key features of this institution would be being policy neutral, and promote appropriate balance between different fields of economic and social statistics.
Government reforms on improving public sector management spurred another major development in the ABS, and the ABS developed and implemented a strategic planning and an overall corporate management approach. Simultaneously it adopted matrix management with statistical and services program managers in Canberra Office being responsible for planning and strategic guidance for all activities in their programs across all offices, while the day-to-day operational responsibilities lay with the senior managers in each State office. With these changes for the first time the integrated statistical service operated as a whole rather than the sum of parts. It is said that the ABS success lies in part on this style of management.
For it to retain a competitive edge, the ABS acknowledged that its success could not only be judged by its outputs, but rather by how well these products are used by government and the community at large. The ABS has therefore, had to market its services extensively.
It should be pointed out that a considerable amount of statistics in the ABS is handled by the Federal and State Government departments, although in general, the ABS features prominently across most of this activity spectrum. For instance, there is an Australian Institute of Health and Welfare that has a significant statistical role. The Australian Statistician is on its board of management, and under law the ABS must agree to them undertaking the collection of data.
Another critical aspect that has led to the ABS being a successful institution is the long-term investment that has been made by the ABS in recruiting, training and developing staff. Although there are many facets to this, only three are worth mention. First, the ABS employs young promising undergraduates and pay them while they continue with completion of degrees before commencing work as full time employees in the ABS. This has been a successful internship programme for long-term tenure. Second, the ABS spends about 8% of the budget per annum on staff development and this has paid dividends in terms having a stream of knowledgeable statisticians and managers available for deployment. Third, by adopting a policy of staff rotation ABS, ensures that all senior staff acquire a broad-range of work experiences, across various subject-matter and service areas and in regional and central offices.
What is important to note is that the development of an effective statistics system is not a simple or straightforward process. In the Australian case, it has taken over 200 years to evolve. South Africa does not have 200 years.
6.7 The United Kingdom
As part of the developed world the UK experience is an interesting one. All power was centralised in Whitehall. But in addition to this the system was decentralised, with each department having a separate and independent statistical unit. This legacy started in 1832. This arrangement was to for approximately a century when the depression that preceded the war and the post World War II reconstruction necessitated the compiling of the national accounts as macro-economic statistics became important for governing the country. This additional role actually enhanced the co-ordination function.
It took yet another thirty years, in 1970, to realise that there were significant efficiency gains and respondent management gains to be exploited by bringing under the same governing body however separately, into two separate units, the collection of data from business and households. The Business Statistics Office (BSO) and the Office for
Population Censuses and Surveys (OPCS) came into being. But the implementation was such that these were not located at the centre. While the Central Statistics Office was given the overall responsibility the decentralised statistical structure remained.
As was the case in Australia in 1970 there was a rapid expansion of the statistical system to match the higher profile of statistics in government decision making. However, in the Thatcher years, the gains were pushed back, with cuts in staff and the only statistics to be collected were those that would meet only the needs of government. It was not long before quality issues came to the fore both within and outside government. By the close of the decade, twenty years from 1970, the office was expanded and included responsibility for the collection of business statistics (including managerial responsibility for the BSO) for compilation of trade and financial statistics and for the retail prices index and family expenditure survey. The CSO was launched as an Executive Agency, and was empowered to run extra collections, and increase the sample sizes in many others. The result was a dramatic and significant improvement in the reliability of UK's macro economic statistics.
Five years later in 1995, the CSO were to merge with the Office for Population Censuses and Surveys (OPCS) to form a new "Executive" Agency hat would be known as the Office for National Statistics (ONS). The merger brought together in a logical way the long overdue merger. As part of the logic, the Employment Department was abolished, and responsibility for labour market statistics was transferred to the CSO. The ONS, has now assumed responsibility of the economic and household based statistics produced in the UK statistical system. The overall employment stands at 2,500 staff out of about 5,000 overall in the Government Statistics System. With this approach the following was achieved:
“greater integration between social and economic statistics;
improved access to official statistics, especially through the database of key statistics
combining the advantages of a decentralised system with a strong and independent co-ordinating agency; and
providing benefits for all users, both government and non-government.”9
As the century drew to a close further advocacy work has continued in the statistics system of the UK, even drawing more on political leadership. The Labour Government elected to power in 1997 had on its agenda, the establishment of an independent statistical service. As a step to getting close to this in early 1998, it issued a Green Paper “Statistics: A Matter of Trust” and sought the views of the public on the matter. The outcome of this process was published in a White Paper “Building trust in statistics.” It is said that the White Paper did not however, go far enough to revolutionarise the UK statistics system and the opportunity for a quantum leap was lost and parts of the influential decentralised components survived intact. However, a new post of National Statistician responsible for national statistics was established with responsibilities as the Director of the ONS and head of the Government Statistical Service. Furthermore a Statistical Commission was established and advises the Government on issues of statistical integrity and quality assurance as well as commenting on the program of work for national statistics. The Commission reports to Parliament annually and Parliament will have some direct involvement in over sighting national statistics. The trajectory in the UK has been much more difficult compared to that of Australia and in both cases the development was over an extensive period of time.