Skip to content

Accuracy and reliability

Contact info

Labour Market, Department of Social Statistics
Torben Lundsvig
+45 39 17 34 21

tlu@dst.dk

Get as PDF

Benefits During Sickness or in Connection with Childbirth

The statistics summarize the reports of illness, birth or adoption that have triggered the payment of unemployment benefit. The expectation is that all sickness benefit issues with payment will be reported. Similarly, the expectation is that all cases of payment due to maternity leave, maternity leave or leave due to adoption are reported. Therefore, the statistics can be expected to be reliable. However, there are a number of cases that will only be reported long after the end of the year to which the case relates, why the last year is not fully updated.

Overall accuracy

There are no sampling errors as the statistics are a census of administrative records. There may rarely occur malfunction in the sense that not all the alerts on the data delivery time is in the supplying IT system is actually sent to Statistics Denmark. It is not always possible to detect such errors. The magnitude of such errors can not be quantified. In addition, at the time of calculation there are always some unfinished process, whose contribution to the total number of per diem days depends on how the end time is calculated by the count. For disease it is about 3-4 per cent and for maternity 15-18 per cent of all reports.

Sampling error

Not relevant for these statistics.

Non-sampling error

A statistic that has an administrative data source is dependent on

  1. The practice that has arisen in connection with the building of the administrative organization and culture together with the attached IT system.
  2. The programs and the business logic the data supplier provides.

Examples of uncertainties related to 1). - Reports come in very late compared to the time period the reports cover. - The meaning of negative amounts. In the case of reverse entries, a negative number is always a corresponding positive number, or is it for reimbursement of benefits paid in error?

  • The number of reports of pregnancy sickness is low compared with the number of reports of brief periods of pregnancy leave early pregnancy.
  • The frequency with which different fields of responsibility occur is changed in connection with administrative changes.

Examples of uncertainties related to 2). - One year to date data delivery can be wrong initiated, so it still is not all reports that are delivered to Statistics Denmark.

But how much such uncertain called means for the overall results is difficult to quantify.

Quality management

Statistics Denmark follows the recommendations on organisation and management of quality given in the Code of Practice for European Statistics (CoP) and the implementation guidelines given in the Quality Assurance Framework of the European Statistical System (QAF). A Working Group on Quality and a central quality assurance function have been established to continuously carry through control of products and processes.

Quality assurance

Statistics Denmark follows the principles in the Code of Practice for European Statistics (CoP) and uses the Quality Assurance Framework of the European Statistical System (QAF) for the implementation of the principles. This involves continuous decentralized and central control of products and processes based on documentation following international standards. The central quality assurance function reports to the Working Group on Quality. Reports include suggestions for improvement that are assessed, decided and subsequently implemented.

Quality assessment

The statistics adds up the reports on sickness, childbirth or adoption which have released deposits of benefits. There can not be paid any benefits without there being a record in either the municipal public administrative system or a record in Payment Denmark's maternity system. The coverage is in this sense 100 per cent. When the individual annual statements can still be tainted with a bit of uncertainty, it is because not all reports for the year has come into the administrative systems when the statistics are published in March. From a user perspective it is probably more crucial to be aware that if the statistics are used to show the extent of disease, respectively, maternity, especially for maternity case, to clarify the division between men and women, then there is information that is not reported. For disease is, unless the citizen is chronically ill or working at an insured employer, the employer's obligation to pay the first 30 days of sick leave. Progress of disease shorter than 30 days never reaches the register that is the basis of sickness statistics. One can not exclude that some illness with a relatively small excess of the limit of 30 days does not reach the register either. If an employer do not wants the trouble in a busy day having to apply for a refund for a few days in view of that he may already have paid full salary for a sick employee in a month. When it comes to maternity leave 20 percent of new mothers do not keep leave on maternity leave benefits. Either because these women do not want to take the leave or more likely because they are not entitled to maternity leave benefits. The corresponding figure for men is 35 per cent. The difference of 15 percentage points between men and women may be because the men are less entitled to leave on benefits or just do not want to take parental leave. But the bottom line is that the statistics utility is adversely affected by the fact that most illness in under 30 days are not reported, and the leave plan at zero days (e.g. The father could have kept maternity leave, but did not do) is not reported.

Data revision - policy

Statistics Denmark revises published figures in accordance with the Revision Policy for Statistics Denmark. The common procedures and principles of the Revision Policy are for some statistics supplemented by a specific revision practice.

Data revision practice

There are some delays in the reporting of absence because of birth or illness. For this reason, published annual statistics early March so that Statistics Denmark can manage to get some of the backlog of reports covering a yearly mileage. The past year is recalculated half a year after the first publication and again one year after the first publication, after which the figures are considered final.