European Central Bank - eurosystem
European Central Bank - eurosystem
Search
Search Options
Home Media Explainers Research & Publications Statistics Monetary Policy The €uro Payments & Markets Careers
Suggestions
Sort by

Surveys on the Access to Finance of Small and Medium-sized Enterprises (SAFE)

Last updated: 21.04.2010

Questions and Answers

Below are the questions submitted by potential bidders together with the respective answers provided by the ECB. These questions and answers are deemed relevant to all potential bidders.

Question 1:

"We have one main question, before we can decide if - and how to proceed: on page 6 (of 49) under table 3, one of the requirements is phrased as follows : "... a single network partner may not cover more than four countries".

Are we allowed to tender for this project using a centralised setup/fieldwork entity for all countries involved, or are we required to use semi-local partners who don’t cover more than four countries?"

"We can confirm that you are entitled to tender using a centralised setup/fieldwork entity for all countries involved."

Questions 2 & 3: existing panel

Q2:In Annex 1, chapter 2.3 first paragraph (page 9) the following statement is made: “the ECB shall provide the contractor with the contact details of the firms already participating in the panel before the beginning of the first wave.” Assuming these panel members were recruited during the first two waves can you disclose the current panel size per market or the percentage of respondents willing to participate in the panel after the first two waves?

Q3: Would you be able to provide us with a detailed description of the contact information contained in the panel sample? In addition, we would like to know how the incumbent is managing the panel component. How is the panel being interviewed currently? What proportion is being interviewed by telephone, by e-mail and by fax. What is the response rate for each methodology?

The current panel members were recruited during the first and second waves. In the first wave, 78% of respondents agreed to be interviewed a second time. 91% of those provided their name. In the second wave, 42% of the attempted contacts with respondents of the first wave were successful. During the second wave, 82% of respondents agreed to be interviewed the next time. By country, this percentage varies from 61% (the Netherlands) to 94% (Portugal), with 86% (Germany), 90% (Spain), 79% (France) and 74% (Italy)

Almost all interviews were conducted by cold-call CATI, including those of the panel (out of 5320 interviews in the second wave, 27 were CATI by email facilitation, 17 web and 7 by fax). Telephone was always the first contact, and the choice of a different mode was offered if the interviewer thought it would allow the interview. 2.8% of the attempts to switch interviewing mode lead to a successful interview.

The information available for firms in the panel is: name of company, name of contact, address, phone number, Dun & Bradstreet identifier, and email. Some of these fields may be empty.

We have no information on how the contractor for the first two waves managed the panel.

Question 4: interview length

Since already 2 waves were conducted can you disclose the average interview length achieved for the current ECB wave (questionnaire as stated in chapter 5.1.1 in Annex 1 page 25) and the current Common wave (questionnaire as stated in chapter 5.1.2 of Annex 1 page 32).

The length of the ECB wave (with common and ECB questions) is on average 12 minutes and 50 seconds (39 minutes maximum, 4 minutes minimum). The contact rate for the second wave is approximately 74% and the response rate 23% (AAPOR methodology, response rate 4 and contact rate 2, adapted from the random digit dialing definition).

The length of the common wave depends on the questionnaire used: ECB, Commission, or both together. We have no information on the length of the Commission questions, although it could be approximately derived from the number of question units.

Question 5 & 6: definition of employee

Q5: It is stated that the universe consists of companies having at least one employee. The question is whether indeed the word employee is intended or whether the more general term ‘person employed’ is meant (i.e. including owner(s) and family workers).

Q6: Universe definition: we can see that we will be including enterprises one or more employees. It is also mentioned that we will be covering different sectors (mining, manufacturing, arts…). We fear that the “employee” definition will vary across sectors. Therefore, we would like to know if there is a more precise description for employees. Are we talking about employees who are part time/ full time, working as external consultants or part of the payroll, should family members working in the enterprise be included or not, etc.

The explanation in question D1 of the current questionnaire states the following:

“THE COMPANY MUST HAVE AT LEAST 1 EMPLOYEE BEYOND THE FOUNDER(S), IF THE FOUNDER IS THE ONLY EMPLOYEE – WE STILL CONSIDER THAT TO BE A 0 EMPLOYEE FIRM. FULL-TIME AND PART-TIME EMPLOYEES SHOULD EACH COUNT AS ONE EMPLOYEE. UNPAID FAMILY WORKERS AND EMPLOYEES WORKING LESS THAN 12 HOURS PER WEEK ARE TO BE EXCLUDED.”

This implies that both full-time and part-time employees, as well as seasonal workers, should be counted as one employee, as long as they work 12 hours or more per week. Family members should be excluded if they do not receive a salary (“unpaid”), but should be treated as employees if they receive a regular salary. External consultants who are not part of the payroll should not be counted as employee.

Question 7, 8 & 9: use of NACE rev 1.1

Q7: Reference is made to NACE rev 2, however Structural Business Statistics from Eurostat (being the basis for the population of enterprises database) using the NACE Rev. 2 classification is not yet available. According to Eurostat results will become available after July 2010 only. Is it therefore permitted to start the project using population data based on NACE rev 1.1?

Q8: If the answer to above question is no: Is it permitted to start the project using population data based on NACE rev 1.1 converted to NACE 2 classification using a conversion table?

The use of NACE rev 1.1 is appropriate at the start of the project, and until comparable results on the universe of firms are available in all countries covered.

Q9: After Eurostat has completed the transformation to NACE rev. 2 and published all Structural Business Statistics using NACE rev. 2, we might build a new statistical picture of the universe using such data and elaborating it where necessary. Is it allowed to include in the price table to be included in the tender (Annex 1, section 5.5.3 other items to be priced; p. 48/49) such an activity as option to be decided upon by ECB/EC when needed somewhere along the project duration of 4 years?

This is certainly allowed. Please note that it is the responsibility of the contractor to ensure that the sample is representative of the universe, irrespective of the creation of a “new statistical picture of the universe”.

Question 10 & 11: source for the weighting

Q10: Regarding the Universe figures used for weighting, in the Technical Reports provided by the incumbent contractor there is no mention of what has been used so far. Would you be able to disclose this information so we take it into account when preparing our proposal?

According to the information at our disposal, the weighting for the first two waves uses the number of firms by size (1 to 9, 10 to 49, 50 to 249, 250+, where available), sector (NACE rev 1.1 codes C, D, E, F, G, H, I, K, M, N, O) and country, available either through national official statistics or through commercial registers of firms, or a combination of both. Comparisons with the Structural Business Statistics (Eurostat) were performed by the ECB for the sectors covered by the SBS.

The following lists the source for the weighting of the euro area countries (countries not marked use information from the Dun & Bradstreet register):

  • HV der Sozialversicherungsträger, Statistische Daten aus der Sozialversicherung: Beschäftigte in Österreich. 2007. (Statistical data from social insurance institutions)
  • Statistics Austria, Arbeitsstättenzählung 2001 (Business demography statistics 2001)
  • Statistics Belgium (Algemene Directie Statistiek en Economische Informatie) (2004)
  • 2005 Sencus of Establishements Statistical Services of the Government of Cyprus
  • Statistisches Bundesamt, Statistical Yearbook 2008
  • National Statistics - Enterprises whole of Greece 2002
  • DIRCE 2008, INE (Central Directory of companies 2008, byt National Institute for Statistics)
  • Tilastokeskus. Finnish statistical centre 2007
  • ISTAT Census 2001
  • STATEC (Luxembourg) and KOMPASS Benelux
  • ETC company lists 2007
  • Statistics Netherlands
  • Central Statistical Office (Poland)
  • FUE - Ficheiro de Unidade Estatística (Dados Económicos - December/2006 ) - Instituto Nacional de Estatística – INE
  • IPIS – Slovenia
  • Statistical Office of the Slovak Republic: Business Reregister of the Slovak Republic, 2009

Q11: Sample frames: In the Technical Reports, the incumbent contractor mentions that they used the Dun & Bradstreet sample frames. Are these frames property of the ECB and EU or they are property of the incumbent contractor? If a property of the ECB and EU, will these be made available to the company winning the contract?

These frames are not property of the ECB. The contractor is expected to procure himself an appropriate sample frame.

Question 12:

We would like to clarify the following matter regarding the survey on the access to finance of SMEs 2010/S 54-079183, with regard to your specification on the usage of a panel in the ECB samples (point 2.3 in the Technical Specifications).

In the first sentence the specification says "... the ECB waves shall have a rotating panel of firms."

Elsewhere the specification mentions: "In case a firm drops out of the panel before its specified number of waves, for each country in the euro area, tenderers shall indicate how they will ensure that the sample drawn will remain representative of the universe."

Which interpretation of the task is closer to the aim of ECB:

A. to carry out cross-sectional studies that carry over willing respondents to participate from wave to wave as much as possible as a 'panel component' (applying the given rules of dropping the units after a certain number of waves) without the aim of controlling the number of the respondents who are willing to participate in a follow-up wave, and controlling the representativity of this ‘panel component’ itself.

B. to build up and maintain a representative enterprise panel in each country specifically for the SAFE study, and use that for the ECB waves?

Interpretation A is the correct one. Firms should not be asked if they want to participate in the panel before they are interviewed for the first time, but towards the end of the first interview.

Question 13

Can CATI fieldwork be conducted by one centralised but multi-national and multi-lingual interviewing team based in one country (which is a common approach on many large international studies and has potential benefits in terms of cost and consistency across countries)?

We interpret the response to a recent tenderer's question as confirming this is the case but we would just like to double check because this is slightly different to the ITT wording "a single network partner may not cover more than four countries".

In line with the response to a recent tenderer's question, we confirm that it would be fine to conduct the CATI fieldwork by one centralised but multi-national and multi-lingual interviewing team based in one country. As long as the company can prove capable to meet the conditions of having sufficient interviewers being (1) native speakers and (2) being sufficiently knowledgeable on the surveyed countries, there is no impediment for sending an offer.

Question 14:

For the ECB sample, do you require us to cost for achieving the same full sample sizes each wave (i.e. panel plus top-up to address attrition in full) or to just interview existing panel members + those refreshed according to your requirements to replace micro firms every four waves, SMEs every 6 waves and large firms every 12 waves?

With regard to sample (section 2.2 of annex 1), the tender document states that the samples need to be representative for each country's universe (with the exceptions cited in the document) for each survey wave. It comes from there that drop-outs/attrition of panel units (as well as possible additional sample refreshments derived from changes in the universe of SMEs) should indeed be catered for.

Question 15:

We are currently preparing our proposal document, which we plan to finalise on Friday and send to you via courier then so that it arrives before the deadline of Monday 26th, 5pm. However, we have a few concerns related to the volcanic ash issue and our courier being able to be delivered on time should air traffic be disturbed over the weekend. Would you be able to let us know what would happen in case there were more (volcanic) problems at the weekend? Will there be any flexibility regarding the deadline?

Considering the ongoing problems caused by the volcanic ashes, there will be no problem in your offers reaching the ECB after the deadline of 26 April, as long as you can prove that the tender was finalised and posted before the deadline indicated in the ITT. In addition, we kindly ask you all to confirm it to us by email as soon as you have submitted your tenders. If needed, we will consider the extension of the 10 calendar days period indicated in our ITT document, section III.1.2 which reads: "...However, the ECB will not consider tenders which reach the ECB 10 calendar days after expiry of the above time-limit."