Accuracy and reliability

Overall accuracy

The goal of the statistics is a value for the enterprise's sales that can be used as an estimate for the enterprise's turnover. The fact that statistics contains data from all VAT-liable enterprises strengthens the precision. Furthermore, the information on the three variables domestic purchases, total purchase and domestic sales, is important for the amount of VAT that the enterprises are obliged to pay to the Tax Agency. It has a direct impact on the enterprise's finances, it therefore motivates the enterprises to fill in data on the three variables correctly. The enterprises do not have the same motivation when it comes to reporting exports and non-VAT sales. The field in the report is called "section C". The value of sales from recycled goods in category C is considered to be understated.

There are no calculations on uncertainty.

Sampling error

The study is based on register information. The sampling uncertainty is zero, since it is a total count.

Non-sampling error

Coverage error: Two cases have been identified where the frame population covers the target population, and the actual level is underestimated: 1. Enterprises with an annual turnover below 50,000 DKK are not part of the framework population. 2. Enterprises that report VAT only using One Stop Shop, are not part of the frame population.

Over-coverage rate: Over-coverage occurs when enterprise becomes inactive and the CVR register is not updated. At the specific point in time, the company continues to be part of the frame population but not part of the target population. When an enterprise becomes inactive, its values are imputed until the enterprise's status is updated in the CVR register. The statics are then adjusted. Imputations are made for a maximum of three periods, which are either months, quarters or six months.

Measurement error: Four types of measurement error have been identified: 1. Errors in reporting are either typos or a conscious or unconscious attempt to cheat. The extent of this type of error is not known. 2. Another type of measurement error is a result of how we divide the sales into months from enterprises that only report VAT quarterly and half-yearly. We have made the assumption that the enterprises that report VAT quarterly and half-yearly have the same behaviour as enterprises that report VAT monthly. 3. A third type of measurement error is because of the way we divide sales for enterprises that reports VAT on a joint SE number (“fællesindberetter”). The split of values depends on how many employees each CVR has. The size of this measurement error is unknown. 4. A fourth measurement is when we place all sales from CVR in the industry group where the enterprise generates the largest added values.

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

Examples of lack of quality when enterprises report VAT:

Uncertainty about the end user. Errors can occur if the enterprise is not aware of 1) Only report VAT when the goods or services reach the end user 2) Goods not subject to VAT must be registered in “Rubrik C”. The extent of any errors from this source is unknown.

Misreporting that may affect the detailed industries without being detected in the month of release but will most often be corrected in quarterly releases.

Below is a list of possible quality issues in relation to statistics processing:
The definition of split of joint VAT reporter. Enterprises that report VAT jointly have one SE number, which covers several CVR numbers. We do not have knowledge of how the values in a joint VAT report are divided into the individual CVR numbers. For a joint report, purchases and sales are distributed among the individual enterprises in relation to the number of employees and the industry's turnover per employee. A change of industry for a CVR number included in a joint report can affect all the industries included in the joint report.

Comparison with Danish statistics: From 2010, the statistics are based on register data. A register of the VAT returns that companies report to the Danish Tax Agency. This means that the data in the statistics is comparable over time from the year 2010, as the statistics include all companies that report VAT. The statistics contain the variable "Salg i alt", which can be used as an estimate of the enterprises' net turnover and can be compared with the net turnover in other statistics such as General enterprise statistics. When comparing, you need to take into account differences such as which types of sales or turnover are included, whether excise duties are included, and whether smaller companies are included in the statistics.

Comparison with foreign statistics: Note that the statistics on "total sales" can be used as an estimate of the net turnover of firms or variable 140301. Note that according to the definition of variable 140301 (net turnover) the following must not be included
- all taxes and duties directly related to revenue. - amounts collected on behalf of a principal if the statistical unit acts as an agent in relation to that principal any income that cannot be linked to the ordinary operations of the statistical unit.

The statistics have the following shortcoming to the above definition. When calculating "total sales", indirect taxes are included e.g. excise duties. Furthermore, as it is not possible to break down sales to invoice level, sales collected via authorisation (reverse charge) and sales that cannot be linked to the statistical unit's ordinary operations cannot be deducted.

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

The first publication of a period is provisional. The reason why the first period is provisional is firstly that enterprises have the right to revise data for that period and secondly, it is a period that contains imputed data for missing reports. The imputed data covers reports from enterprises that report for a quarter or half-year and have not yet reached the deadline. The imputed data also covers data from enterprises that report on a monthly basis but have not reported despite the deadline for reporting having passed. Enterprises can revise a period that goes back three years.