Problem statement

There are a number of threats that exist in the current statistical environment, and that require statistical organizations to act now:
1) rigid processes and methods
2) inflexible ageing technology environments
3) increasing cost of traditional Censuses and surveys
3) inability to quickly respond to emerging information needs
4) inability to make use of administrative data and harness alternative sources of data, such as sensor and satellite data
5) increasing challenges in attracting and retaining critically skilled staff in a competitive market

Impact of this problem

Many statistical organizations are struggling to find the resources to continue to produce robust and relevant statistics to meet the increasing demands of customers. This comes at a time when governments around the world are less likely to provide statistical organizations with large injections of funds to address critically fragile statistical infrastructure.

Over the years, through many iterations and technology changes, statistical organizations have built up their organizational structure, production process, enabling statistical infrastructure, and technology. The cost of maintaining this business model and the associated asset bases (process, statistical, technology) is becoming insurmountable and the model of delivery is not sustainable. Many statistical organizations are introducing service-oriented architecture approaches to improve the flexibility, robustness and sustainability of their technology environments.

Statistical organizations are being increasingly challenged to respond quickly to emerging information needs. Criticisms leveled at statistical organizations include;

• The inability of underlying statistical models such as classifications and frameworks to remain relevant to modern information needs
• Difficulties in producing statistics that are coherent across information domains
• Difficulty in producing richer insights into key priority areas where traditional statistical outputs are not sufficient

For most statistical organizations the underlying model for statistical production is sample survey based. Increasingly there is a need for organizations to make use of administrative or alternative source data to deliver efficiencies, reduce provider burden and make richer use of existing information sources. This requires significant new capabilities that do not exist within the majority of statistical organizations.

The skill-sets that underpin statistical organizations are becoming increasingly valuable in the wider market. It is becoming difficult for statistical organizations to compete to attract and retain these skills in this environment.

How do we judge success?

• more rapidly developed capabilities in areas of emerging need
• drawing together skilled resources from across the international statistical community
• responding to the challenges of emerging information sources such as big data
• assisting statistical organizations to modernise by providing a reference architecture
• reducing costs of production by sharing better practice across the community
• encouraging interoperability of systems and processes
• harnessing new challenges and opportunities faster
• enabling the statistical industry to build common infrastructures and services

What do we already know?

CSPA has been developed and peer reviewed by the international statistical community. CSPA:

• covers statistical production across the processes defined by the GSBPM
• provides a practical link between our conceptual standards – the Generic Statistical Information Model (GSIM) and the Generic Statistical Business Production Model (GSBPM), and statistical production
• includes application architecture and associated principles for the delivery of statistical services
• includes technology architecture and principles - limited to the delivery of statistical services
• does not prescribe technology environments of statistical organizations.

The value of collaboration and sharing

Why does the statistical community want this?

Extending Standards to shared services
As a community, statistical organizations already have shared principles, values, standards and frameworks, such as System of National Accounts and International Standard Industry Classification. We have each developed our own local implementation, duplicating effort and augmenting our local portfolio of systems. CSPA provides a vehicle for the statistical community to co-design, develop, and then implement shareable statistical services.

1. CSPA will allow for services built by the statistical industry using architectural standards to be shared.
2. CSPA gives the statistical community a framework for collaborating and sharing effectively.
3. A set of tools, templates and guidelines are available to help your organization implement CSPA
4. CSPA can be applied to the whole span of the statistical production process.
5. CSPA has a vibrant international community actively supporting and implementing the framework.

Working collaboratively, we can combine resources to develop better solutions for complex problems and opportunities. CSPA’s collaborative development approach allows the development of good systems quickly and cheaply. We can draw on the breadth and depth of skill across the community and maintain momentum over the longer term. Belonging to an international community reduces individual risk for new developments through additional scrutiny and testing. We can also harness capacity from across the community to insulate individual statistical organizations from budgetary shock.

Lifting the maturity of sharing

Visualizing the solution with CSPA Features