10
2.5 Classifications of Types of Administrative Data
The NBS and OCGS adopted the global classification of types of statistics. On the
other hand, administrative data in Tanzania are classified into various categories
based on the sectors or domains they represent. This includes data related to social
services, such as health and education, as well as economic activities, agriculture,
and other areas.
2.6 Social and Demographic Statistics
Social and demographic statistics cover a wide range of topics related to population
characteristics, and social services such as health, education, good governance,
crime, justice and other aspects of social welfare. Demographic data includes
population size, growth rate, age distribution, fertility rates, mortality rates, migration
patterns, and urbanization trends. Social statistics encompass indicators such as
literacy rates, school enrolment rates, access to healthcare services, housing
conditions, poverty levels, and social protection coverage.
Most of the produced statistics under the social and demographic domain are
household based, whereby household is the primary source for data. Among the key
statistics collected at household level are statistics on poverty measure. Currently
there is no alternative mechanism for collecting data on poverty beside using
household-based surveys. However, in administrative data systems, statistics within
that domains are collected based on individual level records compiled by their
respective authoritative institutions.
2.7 Economic Statistics
Majority of the key macro-economic statistics indicators are compiled and produced
by NBS and OCGS using data from different sources including the administrative
data systems. Economic statistics in Tanzania focus on measuring various aspects
of economic activities including production, consumption, investment, trade, and
employment. Key economic indicators include Gross Domestic Product (GDP),
inflation rate, unemployment rate, sectoral contributions to GDP (such as agriculture,
industry, and services), balance of payments, and foreign direct investment.
Economic surveys and administrative data sources provide valuable information for
analysing economic trends, monitoring sectoral performance, and formulating
11
economic policies. Information collected from the administrative data systems serve
as potential inputs to compilation of these indicators.
2.8 Cross-Cutting Issues
Cross-cutting issues refer to themes or concerns that transcend specific sectors and
have implications across multiple domains. Efforts to address cross-cutting issues
often involve integrating data from different sources, conducting specialized surveys
or studies, and implementing policies and programs that promote inclusive and
sustainable development.
In the context of Tanzania NSS, among others, cross-cutting issues include the
following themes:
i.
Gender;
ii.
Governance;
iii.
Environment; and
iv.
Climate change.
2.9 Challenges in Administrative Data collection
Efficient management and utilization of administrative data across these domains are
essential for informing policy making, monitoring progress towards development
goals, and addressing socio-economic challenges in Tanzania. Continuous efforts to
improve data quality, enhance data governance frameworks, and strengthen
statistical capacity are vital for maximizing the value of administrative data for
evidence based decision making.
Beside efforts made Tanzania on improving the NSS, production of statistics through
administrative data systems still faces various challenges that can impact data
quality, reliability, and usability. Among the challenges in administrative data
collection include:
i.
Inconsistent data entry practices;
ii.
Lack of standardized data collection procedures;
iii.
Limited scope associated with under coverage;
iv.
Inconsistent statistics on the same indicator if reported from different sources;
v.
Insufficient training and capacity among staff;
vi.
Technological limitations;
12 vii. Limited sharing of data among key actors due to lack standard data exchanges; viii. Limited availability and use of metadata; ix. Data privacy and security concerns. Addressing these challenges requires a concerted effort from organizations to invest in quality improvement measures, adopt best practices in data collection and management, and prioritize data integrity, accuracy, and security.
13 CHAPTER THREE: GUIDELINES FOR PRODUCTION OF ADMINISTRATIVE STATISTICS 3.0 Frameworks for Statistical Production Frameworks that guide statistical production include legal framework, policies, manuals, guidelines and plans. Statistical agencies and other producers should strive to shape data collected through their statistical production frameworks in a way that satisfies the requirements of national and international standards as well as meeting user needs. Some of the key elements that are recommended and common in most of the existing statistical production frameworks are: i. Data protection: Statistical agencies should have data protection legal framework that regulate the collection, processing, and storage of personal data. These frameworks establish principles for data confidentiality, such as limiting access to authorized personnel, obtaining consent for data collection, and implementing security measures to prevent unauthorized disclosure; ii. Confidentiality: Statistical agencies and third-party entities that handle administrative data may be required to enter into confidentiality agreements guided by existing policies to protect the confidentiality of the data they access or process. These agreements typically outline the obligations of the parties to maintain confidentiality and the consequences of unauthorized disclosure. iii. Access controls: Statistical agencies should implement access controls to limit access to administrative data only to authorized personnel with a legitimate need to know. This includes user authentication mechanisms, role- based access controls, and audit trails to monitor access and detect unauthorized activities. iv. Anonymization: To mitigate the risk of re-identification and unauthorized disclosure, NBS and OCGS encourage or require the anonymization of administrative data be in undertaken before sharing or processing. Anonymization involves removing or encrypting personally identifiable information. v. Encryption: Statistical agencies are required to encrypt administrative data during storage, transmission, and processing to protect it from unauthorized
14 access or interception. Encryption techniques such as encryption-at-rest and encryption-in-transit help safeguard data confidentiality by rendering it unreadable to unauthorized parties. vi. Data minimization: Statistical agencies may promote the principle of data minimization, which involves collecting and retaining only the minimum amount of data necessary for a specific purpose. By reducing the amount of sensitive information held, data minimization helps to limit the potential impact of a data breach or unauthorized disclosure. vii. Data breach notification requirements: Policy and legal frameworks often include provisions requiring government agencies to notify affected individuals and relevant authorities in the event of a data breach involving administrative data. Prompt notification allows affected individuals to act appropriately to protect themselves from potential harm. viii. Penalties for non-compliance: Legal frameworks typically establish penalties for non-compliance with confidentiality requirements, such as fines, sanctions, or legal liability for damages resulting from unauthorized disclosure or misuse of administrative data. These penalties serve as a deterrent against negligent or intentional breaches of confidentiality. ix. Oversight and accountability mechanisms: Policy and legal frameworks may establish oversight mechanisms, such as data protection authorities or regulatory agencies, responsible for monitoring compliance with confidentiality requirements and enforcing penalties for violations. These mechanisms help to ensure accountability and promote trust in the handling of administrative data. x. Data Classification: Implement a data classification scheme to categorize administrative data based on its sensitivity and confidentiality level. This helps identify which data requires the highest level of protection and ensures that appropriate safeguards are applied accordingly. xi. Audit and Monitoring: Implement regular audits and monitoring mechanisms to ensure compliance with confidentiality policies and detect any unauthorized access or misuse of administrative data. This includes conducting periodic
15
security assessments, reviewing access logs, and investigating any
suspicious activities or breaches.
Any agency which intends to collect data for the purpose of producing official
statistics should observe and adhere to these elements. These elements are
normally documented in statistical legislation, agreements or MoU, strategic
documents, manuals and guidelines.
3.1 Generic Statistics Business Process Model
The official statistics are produced based on the international processes that are
described in Generic Statistics Business Process Model (GSBPM) to ensure the
quality of the statistics produced (Annex 3).
The GSBPM provides a standard framework and harmonised terminology to help
NSOs and other organizations which produce statistics to modernise their statistical
production processes, as well as to share methods and components. It can also be
used for integrating data and metadata standards, as a template for process
documentation, for harmonising statistical computing infrastructures, and to provide
a framework for process quality assessment and improvement. The model is
designed to be applicable regardless of the data source, so it can be used for the
description and quality assessment of processes based on surveys, censuses,
administrative data including registers, and other non-statistical or mixed sources.
3.2 Processes for production of administrative statistics
The GSBPM is a reference model for any statistical program. It is intended that the
GSBPM may be used by organisations to different degrees. An organisation may
choose to either implement the GSBPM directly or use it as the basis for developing
customised version of the model. It may be used in some cases only as a model to
which organisations refer when communicating internally or with other organisations
to clarify discussion. The various scenarios for the use of the GSBPM are all valid.
In Tanzania, it has been used as a basis for developing a customized version of the
guidelines for production of administrative statistics (Figure 3.1). Through this
customization, the following phases have been designed to be adapted during the
process.
i.
Specify needs
16
ii.
Design and build statistical program
iii.
Data collection
iv.
Data processing
v.
Data analysis
vi.
Dissemination of results
vii.
Evaluation of the statistical program
Thorough description and processes in each of these phases will be provided in the
following sections.
17 Figure 3.1: The structure of the Tanzania National Statistics Business Process Model Overarching Processes for Administrative Data
Specify Needs
Design and Build Statistical Program
Data Collection
Data Processing
Data Analysis
Dissemination of Results
Evaluation of the Statistical Program
1.1
Identify Needs
2.1 Design outputs
3.1 Set up collection
4.1 Integrate data
5.1 Prepare draft outputs
6.1 Update output systems
7.1 Gather evaluation inputs
1.2 Consultation and confirm needs
2.2 Design variable description
3.2 Conduct collection
4.2 Classify and code
5.2 Validate outputs
6.2 Produce dissemination products
7.2 Conduct evaluation
1.3 Establish output objectives
2.3 Design collection tool
3.3 Finalise collection
4.3 Review and validate
5.3 Interpret and explain outputs
6.3 Manage release of dissemination products
7.3 Agree an action plan
1.4 Identify concepts
2.4 Design processing and analysis
4.4 Edit and impute
5.4 Apply disclosure control
6.4 Promote dissemination products
1.5 Check data availability
2.5 Design dissemination component
4.5 Derive new variables and units
5.5 Finalise outputs
6.5 Manage user support
1.6 Prepare and submit business case
2.6 Test production system
4.6 Calculate aggregates
2.7 Test statistical business process
4.7 Finalise data files
2.8 Finalise production systems
18 3.2.1 Specify needs Whenever there are new statistics that need to be produced, it is critical that the exact needs of such statistics be identified and established. Also, identifying needs is crucial when some ongoing routine statistics production program need to be reviewed. In some cases, needs might come from users which will trigger changes to the ongoing statistics production process. In all these and other circumstances that may arise, the ultimate goal of any statistics production process is to ensure it meets user needs. During the design of any statistics production, consideration should be made at all levels of reporting. Beside institutional requirements and needs which may call for the need of new statistics production, there are sectoral, national, regional and global needs of such statistics. In order to identify such needs, users and stakeholders’ involvement is very critical when identifying needs. The following guidelines elaborate mechanisms through which producer of statistics can ensure inclusive of needs from users. 3.2.1.1 Identify Needs Make an initial investigation and identification of what statistics are needed and for what purpose. This initial investigation may involve internal review of institutional requirement and review of the requirements from national and other frameworks and agendas. This process will form an initial draft of requirements. 3.2.1.2 Consultation and confirm needs Undertake consultation to various intended users of the statistics to be produced in order to confirm the initial draft that was proposed. Users may involve internal and external stakeholders. 3.2.1.3 Establish output objectives Once needs have been identified and confirmed, the producer needs to set clear output objectives of producing the intended statistics. All considerations (legal and technical) should be taken on board to avoid any constraint during production process;