Statistical Journal of the IAOS

For more than half a century, the United Nations (UN) Statistical Commission has been the highest body of the global statistical system. Bringing together Chief Statisticians, the UN Statistical Commission considers and decides upon statistical standards and the development of concepts and methods, and their implementation at the national and international level. In contrast, the UN Committee of Experts on Global Geospatial Information Management has existed for only the last 5 years. Established by the UN Economic and Social Council (ECOSOC) in 2011, the Committee brings together senior national government geospatial experts to develop strategies to build and strengthen national capacity on geospatial information, as well as disseminate best practices and experiences of national, regional and international bodies on geospatial information.

The 60^th International Statistical Institute World Statistics Congress, held in Brazil in July 2015, recognised the importance of the integration of statistical and geospatial information. The then President of the International Association for Official Statistics, Professor Kawasaki, used his Presidential Address to shine a light on this important area for official statistics.

Driven by a growing demand for better data for more informed decision-making, there is an increasing body of material available about the technical aspects and benefits of bringing together statistical and geospatial information, but little has been written about the cultural aspects of two diverse professional communities working together. This paper focuses on some of these cultural aspects, and highlights cooperation underway globally, and nationally within Australia, and provides some suggestions for further improvement.

As the age of open data and data visualization arrives, the Statistics Bureau of Japan and the National Statistics Center of Japan have released a new geostatistics web service. Called jSTAT MAP, the web service can aggregate data and produce choropleth maps based on small area statistics, whether they may be census data or users' own data. jSTAT MAP has expanded horizons for the integration of statistical and geospatial information and acquired a new statistical user base in Japan. This paper describes the key elements of jSTAT MAP and reports on the impact it has brought about on statistical users.

The United Nations initiative on Global Geospatial Information Management(UN-GGIM) aims at playing a leading role in setting the agenda for the development of global geospatial information and to promote its use to address key global challenges like the UN Sustainable Development Goals. The regional committee UN-GGIM: Europe was established in October 2014 and by October 2015, at its second plenary meeting, recommendations were adopted on better integration of geospatial and statistical information in order to foster further usage. The recommendations can be found in a report titled ``Definition of the priority user needs for combinations of data''. This article reflects the main content of the report: identifying relevant policies, collecting use cases, describing and responding to user needs and finally making recommendations on how to improve the integration of statistical and geospatial information. The most strategic recommendation is that Europe should aim for a European Spatial Data Strategy. Other recommendations deal with priority data for a Statistical Geospatial Framework and support the improvement of workflows with geospatial technology.

Statistics has been traditionally based on administrative boundaries(country, states, etc.) and, in few cases, on operational boundaries(enumeration areas, output areas). Climate hazards, however, disregard these boundaries. Statistics based on the geography of climate hazards need to be considered. The level of granularity of statistical data is quite low in most cases but a reduction in the size of aggregation units is necessary to meet the geography needs of climate hazards. Nowadays, these tasks are easier to perform due to the technological advances in GIS and its growing use in national statistical offices.

This paper discusses a case study in Brazil that uses aggregate Population Census microdata in small and regular geographical units to characterize the population settled in areas susceptible to geological hazards and flooding on the northern coast of São Paulo.

The paper finds that Census data, combined with data on the geography of climate hazards, can significantly improve our understanding of hazards and vulnerability, but many steps are necessary before this result can be achieved. The existence of georeferenced statistical or environmental data is not enough. A common spatial unit or basis for both data types is required so that data integration can be simple, quick, and effective.

Statistics Finland published datasets via the Finnish National Geoportal, Paikkatietoikkuna, for the first time in 2013. The Geoportal provides spatial data discovery and viewing services based on standard Web Map Service interfaces (WMS). In addition, the Geoportal enables standard Web Feature Services (WFS) and browsing of features in an integrated view where data are visualised as maps and tables of attributes. Today, Statistics Finland provides more than 200 WMS and WFS datasets for browsing and integrating on the Finnish National Geoportal together with about 2,000 other national geospatial datasets.

Spatial Statistics on Web was a joint project by Statistics Finland and the National Land Survey of Finland to improve the national geoportal services from the viewpoint of statistical data users. The project followed the European Union's INSPIRE Directive (2007) obligations and received a grant from Eurostat. The project implemented an open source web application(Oskari) to tailor spatial analysis tools to statistical data and focused on developing tools for analysing grid-based data (although most of the developed analysis methods can be used for other kinds of spatial data as well). The application, including the user guide, is in three languages: Finnish, Swedish and English.

The paper describes how the joint project between the National Statistics Institute and National Mapping Agency in Finland was inspired by the expansion of open data and open source development.

Statistics New Zealand produces a lot of data about places. Place puts location at the core of statistical production, whether its data about a neighbourhood, a regional government area or the entire country. Modern geospatial technology is providing new ways to collect, analyse and present statistical data. It can also help reduce the cost of collecting and processing the statistics. The efforts of the international community are helping change the way New Zealand delivers statistics and helps Statistics NZ lead national programmes to improve the data ecosystem. This will help Statistics NZ achieve its vision to ``unleash the power of data to change lives'' and spatially enabling statistical data is key. This paper outlines the story so far, the challenges and where to from here.

Comparable regional statistical data for Germany are becoming increasingly important. Regional classifications and standards are fundamental to such data. The European Union (EU) in particular, but also the Organisation for Economic Co-operation and Development (OECD), have carried out vital preliminary work in this area. For example, they have created a hierarchical and non-overlapping territorial classification and laid down criteria for the classification of regions. These classifications are essential in order to be able to conduct harmonised, transboundary sub-national analyses, as required for `open government' or for the EU's funding programmes and environmental and sustainability policy. An overview of the current regional statistical data for Germany as provided by federal statistics completes the analyses.

Recently, many statistical institutes have been moving from traditional estimation approaches based on sample survey data to new approaches that try to exploit the increased availability of administrative data, due to the need of reducing the response burden and providing users with more reliable statistical information. In this context, problems concerning the use of multiple sources for estimation purposes have been receiving an increasing attention in Official Statistics. A commonly adopted strategy is to rely on a ``hierarchy'' of the sources, based on preliminary analyses of the data quality of each source. In this work, we propose an alternative approach based on the concept of latent variables, where one takes advantage of the simultaneous availability of information from different sources. The true values of the target variable are viewed as realizations from a latent(unobserved) variable and the distinct (possibly coinciding) observed values from different sources are considered as imperfect measurements of this latent variable. According to this approach, all the available information is used and ``weighted'' according to its reliability, and a prediction of ``true'' values of some numeric variable of interest is obtained conditional on all the available information.

Cross-national comparability of census data as uniformity was intensely debated from the First International Statistical Congress(ISC), held in Brussels at Quetelet's initiative in 1853, until an agreement was reached at the Eighth ISC, in St. Petersburg, in 1872. However, not much progress was made until the last half of the twentieth century, when the Statistical Commission of the United Nations issued the first set of principles and recommendations for the national population censuses in 1958. In this paper, the progress of statistical internationalism is investigated from Quetelet's vision of census data uniformity and the first international decision to conduct decennial censuses directed at the actual population to Kish's definition of cross-national sample survey comparability. The presentation is based on the detailed documentation of the Integrated Public Use Mictodata Series (IPUMS)-International. As this database is comprised of samples drawn from censuses, it relates to this first international decision on census data comparability.

Effective management of microdata acknowledges that there is no `single fix' that will adequately address all of the potential risks associated with releasing unit record information. Different methods used to maximise data security have different risks and all data confidentialisation methods limit the usefulness of the underlying data [1]. We have developed an approach to data management and release that takes a balanced view of potential re-identification or disclosure threats and prioritises data utility while actively addressing more probable risk scenarios.

This balanced approach to risk management will be discussed in relation to the Australian Early Development Census (AEDC) which is conducted every three years to measure children's development as they enter their first year of formal schooling. AEDC data is collected through a teacher-completed, online checklist that measures five areas of early childhood development. Issues relating to risk and the research community, data management risk mitigation strategies, microdata access modalities and the implications of sharing the risk between academics and the data custodians will be explored.

In the previous census, Estonia provided the people the opportunity of self-enumeration in an electronic e-census environment. The goal of the new solutions was to ensure a bigger coverage for the Census, a better quality and faster publication of the results. The e-census was a success: 2/3 of the persons enumerated took advantage of the new possibility and participated in the census online. By the way the e-census reduced census costs. The present article describes the experience of Statistics Estonia in preparing and conducting the e-census.

For National Statistical Institutes, administrative data sources are an attractive alternative to survey sampling to produce official statistics. The question may arise whether a candidate administrative data source can replace existing survey sampling, for a range of publication domains. We combined an analysis of conceptual differences with indicators from a robust linear regression analysis to assign the domains to four groups: a Control group without conceptual differences, an Accept group with conceptual differences but only small numerical differences, an Adjust group with conceptual differences and substantial, systematic, numerical differences and a Reject group with conceptual differences and substantial, non-systematic, numerical differences. We applied our approach to Value Added Tax data at Statistics Netherlands from which quarterly and yearly turnover statistics are derived. Using the range of values for the indicators of the Control group, we derived thresholds to sort the remaining domains into the groups Accept, Adjust and Reject. Our approach can be used as a first means to sort the domains, but additional analyses and sometimes additional collection of data is needed when conceptual and numerical results are not in line with each other.

In the last years, the Brazilian Institute of Geography and Statistics(IBGE) has been focusing on different aspects of the quality of statistical production, while taking into consideration best practices and principles of official statistics. The aim of this paper is to describe some of the statistical quality management practices implemented recently. Although a lot of improvements have been made, much still remains to be done. A set of procedures to enhance the statistical quality management system is in progress. They will be also presented in this paper.

National Statistical Institutes are currently facing significant changes: innovative approaches that promote integration and standardization are replacing the traditional statistical production chain, based on the vertical integration of survey-specific tasks. Given the significant impact of such innovative approaches, it is necessary to manage and harness the changing process. The adoption of an Enterprise Architecture (EA) helps in this direction by identifying business needs, improving collaboration across an organization and ensuring that the technology is aligned to the strategic vision.

In this scenario, CSPA offers a ``reference architecture'' for a wide spread of the ``plug and play'' approach in designing, implementing, sharing and reusing statistical software solutions, largely based both on statistical standards such as GSBPM and GSIM, and on the ''service-oriented architecture'' model.

The implementation of modernization programs is a current issue, requiring several National and Supra-national investments. Istat is currently investing on CORE as the principal solution for SOA-based process industrialization.

This paper presents CORE main features and highlights how CORE meets important goals, such as process automation, software sharing and support for collaborative work.

Good strategies start with diagnosis of the challenge. The paper presents indications that the High-Level Group for the Modernisation of Statistical production and Services took lightly on that task. The argument of the paper is that in order to detect the challenges that the industry is facing, the venues where it operates should be considered as markets. This is also a precondition for meeting the challenges with the tools for strategic management. Four venues are analysed: The Treasury, which serves as the industry's capital market, the product markets of sample surveys and administrative records statistics, and the public sphere or marketplace of ideas. The venues have different market characteristics. The capital market is a monopsony. The market for sample surveys statistics is subject to free competition. Administrative records are natural monopolies. In the public sphere statistics may be subject to monopolistic competition. Depending on the market the members of the official statistics industry can take four positions, as competitor, collaborator, coordinator and controller. Some are compatible, but impossible to reconcile are the positions as competitor and coordinator or controller. The development in the increasingly important market for administrative records statistics suggests that the industry should emphasize the controller option.

Ranked set sampling (RSS) is a cost-effective sampling technique that induces stratification on the population through rank orders of samples. This, in turn, provides a more structured sample than a simple random sample does with the same sample size. This yields more efficient estimators of some parameters of interest. In this method a fairly large number of randomly identified sampling units are portioned into small subsets of the same size. The units of each subset are ranked separately with respect to the characteristic of interest without using their actual measurements. The measurements of the units with some specified ranks constitute a ranked set sample. In this paper we wish to discuss theory, methods and some recently reported applications of RSS to highlight its advantages. This method could be of some particular interest to those who look for a cost-effective and more efficient data collection technique for sampling and monitoring situations.

In a world of ever increasing data availability and user expectations, National Statistical Offices face mounting challenges to produce relevant and timely statistics. They need to transform their business practice to take advantage of big data - especially administrative data - by integrating non-traditional and survey data sources to maximise value, and utilising new technology to enable enhanced analysis. An example of a response to these challenges is the prototype GLIDE (Graphically Linked Information Discovery Environment) the Australian Bureau of Statistics (ABS) is currently developing using semantic web technology. This environment includes as a test case a prototype semantic linked employer-employee database (LEED) which integrates administrative tax data and ABS business register data to enable detailed microeconomic analysis. However, as data structures become more complex and multi-dimensional, data integration and exploration encounters challenges within traditional relational databases, prompting the exploration of alternatives. Semantic web technology allows for a flexible data structure, machine reasoning and inference on the dataset, a shared understanding of the data's meaning, reusable classifications and standards, easy exploration of many dimensions, and network analysis. The possible advantages of such an approach for official statistics are demonstrated through two practical examples, showing how the prototype GLIDE supports effective data exploration and visualisation, and enables network analysis, to solve real business problems.

Use of internet based tools of data collection is a global trend that is rapidly gaining popularity over the traditional paper and pencil method. A typical web survey will be self-administered by a literate community, with access to internet. Even though technology is one of the fastest growing sectors in Kenya, with a particularly rapid increase in access to the internet and mobile phones, web surveys are yet to gain popularity as the data collection method of choice. This is mostly attributable to low literacy levels and poor access to internet connectivity in the rural areas, where the biggest proportion of population lives. Motivated by the inherent benefits associated with use of internet based surveys, one organization in Kenya, TARDA, successfully developed and adapted web based tools to collect data from 3,400 households most of who were offline. The mobile phone was the basic tool of data collection. The survey yielded a response rate of 96.5%, was completed within one week, reduced cost of data collection by 60% and eliminated errors resulting from data entry. The paper looks at the design and execution of the survey, the success story and lessons learned from the challenges encountered.

Study design, from objective to implementation, determines the effectiveness of the data. This design problem is exacerbated by the specific needs of an organization and lack of information about the area. The organization and SwB members must work closely together to ensure that the study fits the purpose and meets the ultimate objective.

Pre-natal care plays a critical role in maternal and infant healthcare. The present work seeks to assess a maternal and infant care program administered by Global Community Service Foundation, (GCSF), in the Inle lake area of Myanmar (formerly called Burma), and identify ways to expand it. Such expansions includes both the identification of additional villages to service by GCSF, as well as the identification of maternal and infant care knowledge gaps of women and health care workers with the objective of training. Statistics without Borders (SwB) members worked closely with the organization to design and implement the surveys, as well as process and analyze the data. This paper discusses the key results.

In the modern world, statistics has been playing an important role in policy and decision making processes and in monitoring the results. It has strengthened its statute and position particularly since the industrialization trends in Europe, and has contributed to the positivist and rationalist understanding of the world. National Statistical Offices (NSOs) of all states are under intense pressures from changing needs and user profiles in addition to the growing demands from traditional users. Thus, they need to build capacity to cope with these new challenges. Interestingly, many NSOs in different countries are aware of this fact and have already begun to do something. Within this context, a general framework of capacity building in statistics is outlined in this article. While the importance of capacity building in statistics and steps in this context are dealt with theoretically, the capacity building efforts of TurkStat are also examined during the period between 2000 and 2014 with reference to international projects and national efforts.

This paper discusses a way to document the statistical editing process with process tables. Work on process tables for official economic statistics has so far focussed on high level aggregates like Gross National Income (GNI), but here they are presented for every item of the Financial Accounts statistics, i.e. the financial part of the National Accounts. A more widespread use of process tables for official statistics may have been hampered by difficulties with their interpretation and by the costs of producing them. This paper first tries to give process tables a theoretical foundation by generalizing their concept to all statistics. Then it is shown how we have integrated them at the Luxembourg Central Bank (BCL) into our compilation system for the Financial Accounts statistics. However, some issues remain such as a certain arbitrariness in the definition of the starting point and the resulting limited comparability of process tables. It is argued that process tables provide both an important quality management tool for statistics compilers and a quality assessment tool for end-users. At present, more experience with process tables and feedback from users is needed to answer the question to what extent they could be disseminated to other potential users.

The need of new indicators that cover cross-cutting information on household economic well-being is among the current priorities of the National Statistical Institutes as well as a major goal at European level. The purpose of this paper is to apply statistical matching techniques on two different data sources to provide joint information on household income and consumption in Italy. We use data observed on the EU Statistics on Income and Living Condition and the Household Budget Survey. This paper focuses on the role of the available information in improving the matching outputs obtainable from traditional statistical matching methods. More precisely, rough information concerning household income, derived from the Household Budget Survey through a suitable method, is considered. The statistical matching methods will use this additional variable as a matching variable that is highly correlated with one of the target variables, thereby justifying the use of the usually neglected conditional independence assumption. In this paper, important insights on the application of the Renssen's weight calibration approach when matching data from complex sample surveys are also provided. Finally an ex-ante collection of information in SILC could enhance the application of matching techniques and improve the accuracy of the final estimates.

This paper discusses the experience of the Palestinian Central Bureau of Statistics (PCBS) [1] in designing a user-friendly framework for metadata and microdata documentation. The PCBS uses two metadata specifications: the Data Documentation Initiative (DDI) [2] and the Dublin Core Metadata Initiative (DCMI) [3]. Both are defined in the Extensible Mark-up Language (XML) and the Resource Description Framework (RDF). This paper focuses also on the DDI and DCMI as well as its relationship to other relevant metadata standards (e.g., The Statistical Data and Metadata Exchange (SDMX)) [4] and the semantic web technologies. We address the features of these standards as Richer content, Coverage, On-line analytical capability, Search capability and Interoperability since these standards are defined in the Extensible Mark-up Language (XML).

Micro-integration brings together records from different micro-datasets and subsequently resolves data inconsistencies. From a National Statistical Institute's point of view, micro-integration is a means of improving the quality, compatibility and scope of its datasets. Statistics Netherlands has developed various micro-integration processes. This paper describes the micro-integration process, and development thereof, which underlies a dataset representing jobs which were terminated due to lay-off. The input consists of a dataset representing layoff permits issued by the Employee Insurance Agency. That dataset is not compatible with available datasets on jobs because the statistical unit is person instead of job. Moreover, it is flawed by a substantial representation error in the sense that eight percent of the employees for whom a lay-off permit is obtained are actually not laid off. The micro-integration process transforms the dataset on lay-off permits into a dataset on terminated jobs. The latter is compatible with other datasets containing job characteristics. The debate on flexibilisation of the Dutch labour market as well as the huge increase in lay-offs during the financial crisis have underlined the need for data on involuntary job termination. As such, the dataset on terminated jobs unquestionably adds value to Statistics Netherlands' stock of information.

Since 2008, when the U.S. subprime mortgage crisis triggered the financial crisis, financial stability analysis has been increasingly interested in the leverage and indebtedness of households along with the vulnerability of different household groups. The reason for this interest is that the household balance sheet and thus, also their risks, are typically counterparts of those of financial institutions. Moreover, several reports, for example the IMF/FSB report to the G-20 Finance Ministers and Central Bank Governors concerning data gaps, emphasise the need for household data which is broken down by different household types. However, none of the reports specify how, in practice, the accounts should be used.

This article uses the micro-macro linkage of wealth and income accounts and thus creates a set of macroeconomic wealth accounts broken down by household groups, using micro data available at the national level. The aim of the project is to derive a framework where indebtedness indicators can be optimally estimated in a timely manner and at a quarterly frequency. This article makes the first attempt to estimate annual time series by using historic micro-macro linkage and highlights problems related to time series estimation and suggests how these estimations could be developed further.

In many situations, official statistics institutions (OSI) and researchers try to produce statistics with the assisted-model methodology based on incomplete or poor quality data from an unknown data generating system that may display power-law (PL). The ubiquity of PL - measured for high frequency series phenomena and then Big Data - in natural phenomena or manmade complex systems is supported by a vast, recent literature. When, additionally, the number of parameters to be estimated is higher than the related observed points, we are dealing with a non-ergodic inverse problem. This article treats the particular case of aggregated time series characterized by PL and explains how to solve this kind of inverse problem through non-extensive cross-entropy econometrics (NCEE). This is a model-assisted methodology which can be seen as a coincident junction of three scientific disciplines: non-additive statistics, the Kullback-Leibler statistical information theoretic, and the traditional econometrics from the Cowles Commission. This paper provides links to comparative, empirical literature on this technique, which include an application to national production modelling through a constant elasticity of technical substitution (CETS) function.

At the invitation of the University of Minnesota Population Center (MPC) the author carried out an assessment of the IPUMS International integrated census microdata programme during January-March 2016. The terms of reference included the assessment of the measures taken by the MPC to safe guard the security of the microdata, the quality and adequacy of services provided, characteristics of users and satisfaction with IPUMS, use of available microdata, support to participating developing country National Statistical Offices (NSOs) and adequacy of a proposed Remote Data Center(RDC).

The conclusions of the review are that IPUMS International is a unique, flexible, successful and secure programme for managing access to anonymized, harmonised and integrated microdata to academic users and policy makers. While currently the user base is predominantly in developed countries, steps are being taken to expand usage by researchers world-wide. The physical, methodological and technical arrangements for safeguarding the security and confidentiality of the data files are excellent; the possibilities of breaches are minimal. Data users have very positive opinions of the quality of the data, scope of services and expertise of staff but desire more detailed, up-to-date microdata. NSOs rate IPUMS International and its services positively but request advanced methodological training for staff and regular information on the use of their country's data. IPUMS International planned activities are presented and their contributions to census methodology are highlighted.

Explicit bias reflects our perceptions at a conscious level. In contrast, implicit bias is unintentional and operates at a level below our conscious awareness. Implicit stereotypes shaping implicit biases are widely studied in criminal justice, medicine, CEO selection at Fortune 500 companies, etc. However, the problem of unconscious bias remains. E.g., while women constitute an increasing proportion of all STEM undergraduates, they still make up only a small proportion of faculty members at research universities, and they are substantially under-represented in organizational leadership and as recipients of professional awards and prizes. Can we afford to have unintentional perceptions continue to hinder the success and advancement of women and other underrepresented groups? Can we afford to continue to underuse human capital in science? This session at the 2015 Joint Statistical Meetings (JSM) aimed to illuminate what statisticians need to know and do to break the glass ceiling of implicit bias.

In January 2014 the General Assembly of the United Nations endorsed the Fundamental Principles of Official Statistics, which was adopted by the Statistical Commission of the United Nations in April 1994, following an initiative of the Conference of European Statisticians. Valid and reliable information is essential for the management of the affairs of a democratic society aiming at generalised wellbeing and prosperity. It is important that users and stakeholders of official statistics and the citizens at large have total confidence in statistics. To produce valid and reliable statistics it is necessary that Governments provide the legal framework and resources to the statistical system of their countries to allow statisticians to produce the required statistical information, without interference using the best available methodology and techniques from the best suited sources of information. Respondents, be they individual, enterprises or organisations, have to provide the required information truthfully and as completely as possible. Official statistics have to guarantee that such individual information will be used for statistical purposes only. Moreover the results of statistical enquiries have to be made available to all users without distinction. Such basic requirements of official statistics were not respected in the centrally planned economies before 1989 and even in some of the countries with market economies. During the transition process toward democracies and market economies of the countries from Eastern and Central Europe it was recognized that official statistics plays an essential role for preserving democracy and that its special and unique role should be recognized by governments and the public at large. At the request of one of the Eastern European countries the Conference of European Statisticians proposed a Charter called ``Fundamental Principles of Official Statistics'' establishing the parameters to guarantee the production of valid and reliable official statistics. As years passed, it was recognized that these ``Principles'' should have a universal validity. This was reached in 2014 with the endorsement of the ``Principles'' by the United Nations General Assembly. Consequently the Fundamental Principles of Official Statistics have universal acceptance and should be adhered to by all nations and societies. Suggestions are made ensure that the Fundamental Principles are continued to be adhered to.