Barriers to entry in public procurement

We conduct a conjoint survey targeting individuals representing firms potentially participating in Public Procurement (PP) auctions. The central focus of our study lies in the respondents' decision-making process when choosing between two specific tenders for the submission of their offers. These tenders are characterized by variations along eight distinct experimental dimensions, which are randomized for each respondent. The study contains background questions. Additionally, the survey contains randomized questions asking respondents to provide numerical answers regarding the likelihood of submitting an offer, the subjective probability of winning the tender, and the estimated offer price, given a randomized value x which is related to for example the effective working time required for offer preparation, geographical distance from the purchasing unit, a potential secondary objective of the procurement, or the number of other bidders. The survey records firm identifiers which can be used to match the responses to firm-level data.

2023-05-16

2023-12-31

The research aims to study the specific attributes of procurement tenders that are favored by individuals representing the offering side. The primary focus is on the respondent's choice between two specific tenders when deciding where to submit their offer. The tender they choose reveals valuable insights into the preferences of both the respondent and their employer. These tenders vary along eight distinct experimental dimensions that are randomized for each respondent. These dimensions measure various aspects of the procurement process, including their experience with similar tenders, the project's size relative to their company's turnover, the nature of the procurement organization, secondary objectives related to the product or provider, criteria for selecting the winning bid, the time invested in drafting and submitting the offer, the estimated likelihood of competitors filing complaints with the market court, and the expected level of competition.

The study conducts a conjoint survey experiment involving employees from companies that potentially participate in procurement tenders in Finland. We initiate contact with these employees through our collaboration partner, the Competition and Consumer Authority. Initially, we make phone calls to reach out to potential participants, and subsequently, we send them invitations to participate in an online survey. Each respondent receives a unique web link for accessing the survey, which includes the conjoint scenarios. Some respondents are contacted only by email, containing a unique web link for accessing the survey.

In the survey, all respondents are first asked background questions to assess their personal experience and their employer's experience in participating in procurement tenders, as well as the suitability of their company's product for such tenders. Subsequently, all respondents are presented with an introductory text that remains consistent across different experimental conditions. Following this, they are presented with a series of choices between two fictitious tender profiles. In each pair of tenders, respondents must indicate which one they would prefer to participate in. These tender profiles vary across eight attributes, which are randomized for each scenario.

Specifically, we ask the following question and use the subsequent attributes with randomized realizations along each of the 8 dimensions of hypothetical procurement tenders:

“Select which of the two potential procurement tenders you are more inclined to prepare and submit an offer for in each pair.”

Experience
• Your company has not previously made an offer in this procurement unit´s tender
• Your company has previously made an offer and lost in this procurement unit´s tender
• Your company has previously made an offer and won in this procurement unit´s tender

The estimated size of the project relative to your company's turnover
• 10%
• 30%
• 60%

Nature of the procurement organization
• Municipality or other organization at local level
• Other country or EU organization
• Government organization
• Joint procurement unit

Secondary goals for the product or to the provider
• No secondary goal
• Innovation goal
• Employment goal
• Environmental goal
Winner selection criteria
• Lowest price
• Best price quality - score

Efficient working time spent on drafting and filing the offer
• 6 hours
• 20 hours
• 80 hours

Estimated risk (probability) that competitor complains to the market court
• 0%
• 2%
• 8%

Anticipated competition
• Our company makes the only offer
• A bid is left by one other company
• A bid is left by six other companies
• Not possible to assess the expected level of competition

Following the conjoint survey, respondents are presented with a hypothetical secondary objective for the tender and are required to respond to the following numerical questions:

• What is the likelihood, expressed as a percentage, that your company would submit an offer in such a situation?
• What is the likelihood, expressed as a percentage, that your company would win the tender?
• As a percentage, how likely do you think that, in addition to your company, other bidders would participate in the tender?
• Would the mentioned secondary objective for procurement alter your company's offer price? If so, by what percentage and in which direction compared to a situation where there was no secondary objective.

Additionally, the survey includes the following questions with randomized attributes:

• How likely, expressed as a percentage, do you think your company would submit an offer when the distance from the purchasing unit is [XXX]? You can select any number between 0% and 100%.
• What is the likelihood, expressed as a percentage, that your company would submit the offer when it takes [XXX] units of effective working time to prepare and submit the offer?

Uniform randomization of choice scenario from the set of possible choice scenarios based on a professional survey software.

Randomization is at the individual choice scenario level

No

Number of individuals (firms) in our surveys. We field the survey to as many firms as possible in Finland. We expect a very low ex ante the response rate of about 5%, but expect about 500 completed surveys embedding our experiment. The response rate depens on our attempts to contact the firms. We have funds available to contact some firms by phone to elicit response.

The targeted number of respondents in the conjoint survey is at least 500. We anticipate a survey response rate of 5%. To ensure a robust sample size and account for potential variations, we plan to reach out to 12,000 potential respondents.

Uniform randomization of 8 possible pre-determined choice scenarios across 2 choices (16 choices per expected individual response). The number of observations for 500 respondents is therefore 8000.