The Statistisches Bundesamt in Germany Produces 'Statistics – Figures for Everyone' with the Help of SAS^®

The experts of the Statistisches Bundesamt (the German Federal Statistics Office) create new knowledge for the public, politics and science using detailed micro data and highly dense aggregate data, tables and business statistics. They use a repertoire of high-performance standard software, supplemented with self-produced algorithms and processing procedures.

Formulating Knowledge from Data
Do the real conjuncture data correspond to the prognosis of the economics institute of the previous year? In 2000, predictions and final results were almost identical for one important value: the national product recorded a plus of 3.1 percent and rose to 3.982 billion Deutschmarks (US$2.5 billion). The Federal Statistics Office is responsible for finding this real indicator as well as other units of measurement such as the production index, population development and the total economic calculation. Highly qualified professionals have pledged themselves to the Federal Statistics Office's motto: "Statistics – Figures for Everyone."

Approximately 2,300 specialists are employed at the central office in Wiesbaden and a further 700 at the subsidiary in Bonn. The preparatory and analysis work is carried out by the IT department using special data handling procedures and by professional statisticians using dialogues or procedures that were originally run in a host environment. However, their own requirements, as well as those of people working in science and politics, have significantly increased over the past few years.

Furthermore, the technical frame conditions of data processing have changed significantly. Therefore, the conversion to a new IT infrastructure in this sector of the Federal Statistics Office was necessary. "After thorough research, we decided at the beginning of 2000 to use SAS Version 8, as well as SAS Enterprise Guide," says Ernst Schrey, group manager for IT User Services and Statistical Information Systems, about the finalization of the accurately planned procedure at the Federal Office.

Together with a special working group consisting of professional users and IT personnel, Schrey checked various standard products available on the market by measuring them against the user profiles of professional users. The conclusion: "The SAS solutions covered both our basic types of work – short-term ad hoc analyses as well as periodic evaluation processes – better than all other products available on the market." Thus, ad hoc analyses can be created easily in a few working steps for their own purpose or on request from scientists or institutes.

For periodic reports, basic patterns and working sequences must be stored, which are then used for setting values for periods of time, special questioning zones or other process parameters. "The SAS software has the extensive functionality necessary for the handling of complex tasks," was the first important argument for Schrey. Two further ones follow: "By using Enterprise Guide or the development tools like SAS/AF or SAS WebAF, the many various basic components can still be hidden from the professional user."

The deciding factor for Schrey was the possibility of integrating his own developments, in particular the large number of highly complicated algorithms that have been developed over the years at the Federal Statistics Office. If necessary, they can be tied in with the processes to be developed using the SAS solutions as a starting point.

The office creates the statistical evaluations according to a federal order that is based on the Law for Federal Statistics. The basis is made up from basic data sent directly by compulsory reporters or via the statistics of the federal countries and as business statistics from offices and ministries. They contain micro as well as aggregate data in ASCII or EBCDIC format as well as tables such as Excel files. From these, the scientists and specialists of the Federal Statistics Office create the periodic and additional one-time work materials and publications. Up until recently, they were available in print or in electronic form. The results are published on electronic data carriers and increasingly on the Internet.

The complex reasons that led to the replacement of the mainframe and the "home-grown" software are now even more obvious. The old programs were not portable and are now are widely replaced with SAS products, including customized applications. "The world of IT has progressed beyond that," Schrey emphasizes, adding that at the Federal Statistics Office commercial observations played as much a part as tight personnel resources on the IT market. Furthermore, platform-independent systems and graphic surfaces should in the future open up safe development paths.

Journey into the GENESIS Future
The implementation of a new common information system for Federal and State usage (GENESIS) was the driving force for these conclusions. Together with the analysis and evaluation functions created with SAS, it is to replace the old information system STATIS-BUND.

The data and functionality contained in the old system give an idea of the extent of this task. It contains statistical results in the form of timelines and tables with in-depth factual separation of stored structure data from all sectors of the official statistics that have been collated over a number of years. The list of the professional sectors or important factual contents contains many details of these segments under approximately 20 main headings – for example: fitness for work, traffic, monies and credit or employer's contributions. The timelines contain mainly periodic results for the Federal Territory. Structure tables, in contrast, absorb results such as election data. All this information can be subjected to a series of mathematical-statistical analyses using a number of solid as well as formulating methods via components and an integration language.

With the introduction of the SAS solutions, the IT world of the Federal Statistics Office is now extended by a highly valued component. The first step was local installation on 50 computers – meaning server and client in one. The second was the planning of a central Unix Server Sun Enterprise 3500 for flexible use by a larger number of users, the processing of large data volumes with, for example, approximately 30 million sentences, and an increase in data security. Its implementation is currently in the test phase. The NT clients are connected to this server via a Novell network. In the meantime, approximately 90 specialists have attended a two-day training course in the use of the SAS Enterprise Guide. "It fills a gap regarding simple user guidance," Schrey concludes.

Additionally, approximately half of the specialists were trained in SAS procedure development, including IML and ODS. Furthermore, the IT team was trained in the SAS development environments WebAF and SAS/AF in order to be able to create closed applications for the specialists, based on SAS. First experiences were drawn from a user platform created with SAS/AF for a procedure that calculates an index of work productivity.

Furthermore, applications were created with which monthly tables and graphics taken from the results of a timeline analysis can be placed into the Internet offer of the Federal Statistics Office as well as used for seasonal cleanups according to the known procedure X12-Arima. The advances achieved so far are the first steps in a further planned development. Thus access to data objects of the GENESIS databank should be made possible by using SAS. A further consideration is to implement thin clients for special tasks using WebAF.

Generally, digital procedures for gaining and transmitting information in a broad electronic spectrum of the Federal Statistics Office should be supported further using SAS. Here the Federal Statistics Office is also further along than some other enterprises – as stated in the second point: money values can already be reported to the statisticians in euros.

The results illustrated in this article are specific to the particular situations, business models, data input, and computing environments described herein. Each SAS customer’s experience is unique based on business and technical variables and all statements must be considered non-typical. Actual savings, results, and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software. Brand and product names are trademarks of their respective companies.

Copyright © SAS Institute Inc. All Rights Reserved.