Page 452 - Invited Paper Session (IPS) - Volume 2

IPS320 Tine Cordes et al.
Summing up on these different tests, it seems that no single indicator alone
captures dependency. Thus, a number of indicators will have to be present to
qualify as a dependent contractor, while other indicators may disqualify.
Indicators regarding pricing seem unable to capture dependent contractors
effectively, while the other indicators may have potential. It is likely that those
who are identified as dependent contractors are not more exposed than
others in the (already relatively exposed) test group.

4. Discussion and Conclusion
4.1. Conclusions with discussion
We may conclude the following regarding dependent contractors in an
economic setting like Denmark:
1. Any single criterion used for identifying a dependent contractor, seem
to include many workers who are not likely to be dependent
contractors – probably more than half. Therefore workers should most
likely meet more than one criterion to qualify as a dependent
contractor. The composition of such criteria needs more work.
2. From the results it seems that using the price control approach is not
a particularly viable way of identifying dependent contractors. In the
2017 test, the groups that were identified were deemed not sufficiently
relevant, and in the new test, the manual check showed that the share
of potential dependent contractors was very low. The failure of these
criteria might be due to the fact that many self-employed in the test
group (no incorporated company, no employees) – even if they do not
depend on another economic unit – have little influence on the market
price. Even if the self-employed could in theory set the price of the
goods and services provided, this may not be possible in reality due to
market factors.
3. Criteria about having a single dominant client may, according to the
results, work in some cases, but like any other indicator, not by itself
(cf. conclusion No. 1).
a. The concept of client is unfortunately ambiguous in many cases
regarding capturing potential dependent contractors. The
presumed client may in some cases consider persons as his or her
own clients when they are in fact the clients of the employing
company. This could lead to misclassification.
4. It seems like a good idea to focus on working arrangements and
conditions that are similar to those of employees, a point that is
also mentioned in ICSE-18. Initially, it is arguably a good theoretical
starting point that could also serve as an umbrella for other
characteristics of dependent contractors. Moreover, the results show
that indicators of working arrangements and conditions that are similar

439 | I S I W S C 2 0 1 9