17(34)
Computing the Threshold Value
Imposing a regulatory framework on public procurement implies both gains
and costs. Gains can be assumed to be roughly proportional to the value of
a given procurement, whereas costs consist of a start-up cost and a cost
term that increases more slowly with procurement value. The optimization
problem consists of finding a reasonable trade-off between the gains from
public procurement – increased competition, reduced risk of corruption –
and the administrative costs of following procurement rules. The problem
of choosing a threshold value is equivalent to selecting a subset of all
procurements for which the likelihood that the framework yields a positive
pay-off to the procuring entity is sufficiently large.
Quality Aspects
Quality is notoriously difficult to come to grips with in the public
procurement context. In certain sectors, such as architecture, form and
design, or care for children or the elderly, it is difficult or impossible to
formulate all desiderata as mandatory requirements. On the other hand,
quality judgments open the door for arbitrariness and corruption.
The academic literature is relatively sparse; for a few examples, see Che
(1993), Naegelen (2002), and Asker and Cantillon (2008). The
Handbook
on Public Procurement
(Dimitri et al. 2006) devotes one rather short
chapter to quality issues. A common solution is to use some form of
weighted sum of the quality and the price variable, where the relative
weights reflect the importance assigned to each dimension. Procedures for
trading quality off against price not only lack transparency to observers but
also often seem to be incomprehensible to users themselves (Bergman and
Lundberg (2013), Keeney (2002), Mateus et al. (2010)). The result of the
evaluation can depend on bids that are uninteresting to the outcome, which
is referred to as ―dependence on irrelevant alternatives‖—a common
phenomenon according to Bergman and Lundberg (op. cit.). In some
procurement operations, criteria are used that would indicate that the
procuring entity is prepared to pay an infinitely high price–an obvious
absurdity. The source of these problems is often the relative scales of
evaluation normally applied when price and quality are to be weighed
together. For the quality dimension, an absolute scale is most often used,
but for the price scale, the standard approach relates the prices to the least
expensive alternative. If the weight assigned to quality is sufficiently high,
infinitely high prices will be accepted.
Given the wide variety of methods used for trading quality off against price,
the question arises of how quality dimensions should be handled in the
present analysis. If the consumption functions of procuring entities were
known, different tenders could be ranked according to the welfare scale
implied, but they are not.
