Questionnaire

1. What information was collected?

This section describes the content of the survey questionnaire. It presents an overview on the questions related to euroization and the sociodemographic information collected in the survey. It also provides a summary of topics beyond euroization that have been covered by the survey. With a view to maximizing comparability across countries, the main questionnaire for all countries is compiled in English by OeNB staff before each wave. Questions related to euroization must not be changed – neither in terms of English wording nor order. The survey institutes in the individual countries may only change the order of sociodemographic questions. They may also choose to collect additional sociodemographic information or to provide finer categories for the required sociodemographic questions if these can be merged into categories in the English original. All versions of the questionnaire are available for download below.

1.1 Main questionnaire 

The main questionnaire includes a core set of questions on the use of foreign currency, in particular the euro. Questions also cover the determinants of foreign currency use. These questions have been designed taking into account insights from previous research on asset and currency substitution. For research on deposit dollarization and euroization, see, for example, Ize and Levy Yeyati (2003); De Nicolo, Honohan and Ize (2005); Broda and Levy Yeyati (2006); Ranciere, Tornell and Vamvakidis (2010); Brown and Stix (2015). For research on currency substitution and saving in cash, see, for example, Engineer (2000); Craig and Waller (2004); Valev (2010); Stix (2013) For research on foreign currency borrowing, see, for example, Ize and Levy Yeyati (2003); Luca and Petrova (2008); Ranciere, Tornell and Vamvakidis (2010); Brown and Haas (2012); Fidrmuc, Hake and Stix (2013); Beckmann and Stix (2015); Fischer and Yesin (2022).

1.1.1 Foreign currency use and preferences
Questions on the use of foreign currency cover:

• whether respondents have deposits and, if so, in which currency
• whether respondents hold foreign currency cash and, if so, in what currency and what amount
• whether respondents have a loan and, if so, in which currency

Questions on the use of foreign currency also elicit preferences to:

• save in cash rather than save using a bank account
• deposit money in foreign rather than in local currency
• take out a loan in foreign rather than in local currency

1.1.2 Determinants of demand for foreign currency
Regarding the determinants of demand for foreign currency the questions cover:

• memories of past financial crises and financial loss in the wake of these crises
• current economic and political conditions (trust in policies and institutions)
• expectations regarding exchange rates and inflation
• expectations regarding the introduction of the euro as a legal tender
• network effects (commonness of savings and payments in foreign currency)
• income and remittances in foreign currency
• effect of current crises (global financial crisis, COVID-19 pandemic)
• financial literacy

1.1.3 Overview of additional topics covered by the questionnaire
The questionnaire is regularly adapted to focus on topics of particular policy relevance and current research interests. These additional research topics are usually only included in one or two waves, which means that the length of the questionnaire is approximately the same for every wave. Interested researchers can download all versions of the questionnaire below. To make it easier to find questions that are not part of the research focus on euroization, table 1 provides an overview of topics covered by further questions. The table is not comprehensive but is intended to serve as a first point of reference.

1.1.4 Time series dimension of questions
For researchers who are interested in making use of the time dimension of the survey, the GRID (see download) provides an overview which questions were included in which wave. It also includes basic tags to facilitate searching for particular topics covered in the questionnaires. For the majority of questions, the wording of the question has remained exactly the same. For some questions, wording was improved over time. Where the OeNB thinks that the change in wording might have an effect on comparability over time, the variable name is changed. Still, researchers are advised to double-check that minor changes in wording do not affect their particular research question. To do so, please make use of the different versions of the questionnaire (see download). In addition to changes in variable names that the OeNB has implemented to signal a change in comparability across time, some variable names have been changed as the order of the questions in the questionnaire was revised. For these cases, the wording of the question remains unchanged and, therefore, apart from the position in the questionnaire, it is comparable across time. Table 2 provides a list of variables where the time series dimension is longer than the different variable names would indicate at first glance.

1.2 Sociodemographic and -economic information

Sociodemographic and -economic data are “independent” variables that allow us to compare behavioral outcomes, attitudes and beliefs across different groups of respondents. For such comparisons to be meaningful in an international survey, the sociodemographic and -economic data need to be harmonized so that categories are comparable across countries. The OeNB Euro Survey applied ex post harmonization until spring 2012 and ex ante harmonization from fall 2012 onward. Ex ante harmonization has the advantage that questions are optimal for the country-specific context but has the disadvantage that the categories that are comparable across all countries covered by the survey may be very broad. For the survey waves of fall 2007 to spring 2012, each opinion poll institute asked the key sociodemographic questions in the phrasing and categories that were considered most appropriate for the respective countries. The OeNB only specified which sociodemographic characteristics had to be elicited. Thus, questions could vary across time and/or countries and sociodemographic data are harmonized ex post for the waves 2007–2012 spring at the OeNB.
Ex post harmonization aims at comparability both across countries within one wave and across survey waves. Categories for sociodemographic information that take into account these changes and are comparable across countries are sometimes rather broad. That is why the OeNB Euro Survey applies ex ante harmonization from 2012 onward.
For data users these changes mean that any analysis covering the waves prior to fall 2012 should use the ex post harmonized sociodemographic and -economic data named h_*. For pooling across the ex post (up to spring 2012) and ex ante (fall 2012 onward) harmonized survey waves, the ex post harmonized data are continued, i.e. they are available from fall 2007 to the most recent wave. For analyses using data from fall 2012 onward only, the ex ante harmonized data should be used, which are named n*.
The ex post harmonized sociodemographic and -economic data cover the following information for all waves:

• age
• gender
• number of household members, whether there are children in the household, whether the respondent is the head of household
• education
• labor market status
• household income after taxes¹

In the dataset, ex post harmonized sociodemographic and -economic data are named h_*. In addition, information is collected on:

• the region where the respondent lives
• the size of the town or village where the respondent lives

A codebook for ex post harmonized sociodemographic variables is available for download (Collection and harmonization of sociodemographic data). For the waves from 2012 onward, sociodemographic questions are harmonized ex ante by the OeNB to the extent possible. The sociodemographic and -economic questions are included in the different versions of the questionnaire, which are available for download below. These ex ante harmonized questions also elicit respondent information on:

• marital status
• religion
• the age and number of children in the household
• who is in charge of managing household finances
• personal income after taxes (since 2019)

As previous research has noted, even with ex ante harmonization, comparability across countries may still be limited, especially for data that are strongly influenced by country-specific factors, such as education. This is also the case for the OeNB Euro Survey from fall 2012 onward, where education data are still subject to some ex post harmonization.

1.3 Paradata

The OeNB Euro Survey collects auxilliary data that describe

• the field work and/or collection process: interview date, time and duration, survey mode, language of the interview
• the survey interviewers: age, gender, experience in conducting the OeNB Euro Survey, education and employment status
• observations recorded by the interviewer: the neighborhood of the respondent’s residence, the residence of the respondent and the respondent’s behavior before and after the survey

The purpose of these paradata is to improve data quality (West, 2011). They are used for internal quality checks; see also “Data consistency checks and editing”. At present, paradata are not used for nonresponse adjustments or for providing an imputed dataset to external researchers.

1.4 Translation into national languages 

For each wave, the final English questionnaire is translated into the national languages of the countries covered by the OeNB Euro Survey. Translations are done by professional translators who live in the countries covered by the OeNB Euro Survey and hold a university degree in translation studies or other country-specific proof of qualification (e.g. as state-certified translator). The selected translators are native speakers of the target languages. The translations are reviewed by the survey experts at the survey institutes in each of the countries to check that they are suitable for face-to-face surveys and the phrasing is easily understood. Translations are then checked by Gallup Vienna and its subcontractors as well as the OeNB and its
subcontractors. For questions that are repeated in several survey waves, the wording is kept unchanged across waves unless there is a strong reason to adapt it. For North Macedonia, there is an Albanian and a North Macedonian version of the questionnaire. For Bosnia and Herzegovina, the questionnaire includes alternative translations to take into account regional differences.
Despite quality checks, some questions were translated incorrectly and translation errors were only discovered at a later stage when the survey waves were already completed. Table 3 lists the questions and waves for which translation errors where discovered at a stage when data could not be recovered. We distinguish translation errors where the meaning is not comparable with the English original at all and inaccurate translations where the meaning is to some extent comparable. For the first type, data are set to “missing” for the respective variable, country and wave. For the second type of error, data are included in the final dataset. In both cases, flag variables allow users to trace the error or inaccuracy in the dataset.

1.5 Piloting and programming

The questionnaires are pre-piloted by OeNB staff in English. On this basis, pre-pilot questions are amended, in particular, new questions. Sometimes, additional questions are formulated. The pre-pilot is followed by a small pilot conducted by the institutes in all countries covered by the survey. Typically, 10–20 pilot interviews are conducted in each country. Feedback from these pilots is provided to the OeNB and may result in further changes to the questionnaire or interviewer guidelines. Frequently, these amendments reflect country-specific issues and are, therefore, not taken into account to ensure comparability across countries.
The survey institutes in each country program the questionnaire. Programming is then checked for correct filtering and also for consistency across countries. 

References

Beckmann, Elisabeth, and Helmut Stix. 2015. “Foreign currency borrowing and knowledge about exchange rate risk.” Journal of Economic Behavior & Organization, 112: 1–16.

Broda, Christian, and Eduardo Levy Yeyati. 2006. “Endogenous Deposit Dollarization.” Journal of Money, Credit and Banking, 38(4): 963–988.

Brown, Martin, and Helmut Stix. 2015. “The Euroization of Bank Deposits in Eastern Europe.” Economic Policy, 30: 95–139.

Brown, Martin, and Ralph De Haas. 2012. “Foreign banks and foreign currency lending in emerging Europe [Capital structure and financial risk: evidence from foreign debt use in East Asia].” Economic Policy, 27(69): 57–98.

Craig, Ben, and C.J.Christopher J.Waller. 2004. “Dollarization and currency exchange.” Journal of Monetary Economics, 51(4): 671–689.

De Nicolo, Gianni, Patrick Honohan, and Alain Ize. 2005. “Dollarization of bank deposits: Causes and consequences.” Journal of Banking & Finance, 29(7): 1697–1727.

Engineer, Merwan. 2000. “Currency transactions costs and competing fiat currencies.” Journal of International Economics, 52(1): 113–136.

Fidrmuc, Jarko, Mariya Hake, and Helmut Stix. 2013. “Households? foreign currency borrowing in Central and Eastern Europe.” Journal of Banking & Finance, 37(6): 1880–1897.

Fischer, Andreas M., and Pinar Yesin. 2022. “Foreign currency loan conversions and currency mismatches.” Journal of International Money and Finance, 122(C).

Ize, Alain, and Eduardo Levy Yeyati. 2003. “Financial dollarization.” Journal of International Economics, 59(2): 323–347. 

Luca, Alina, and Iva Petrova. 2008. “What drives credit dollarization in transition economies?” Journal of Banking & Finance, 32(5): 858–869. 

Ranciere, Romain, Aaron Tornell, and Athanasios Vamvakidis. 2010. “Currency mismatch, systemic risk and growth in emerging Europe [Capital structure and financial risk: evidence from foreign debt use in East Asia].” Economic Policy, 25(64): 597–658.

Stix, Helmut. 2013. “Why do people save in cash? Distrust, memories of banking crisis, weak institutions and dollarization.” Journal of Banking & Finance, 37: 4087–4106. 

Valev, Neven T. 2010. “The hysteresis of currency substitution: Currency risk vs. network externalities.” Journal of International Money and Finance, 29(2): 224–235. 

West, Brady T. 2011. “Paradata in survey research.” Survey Practice, 4(4): 1–8.