Compensating Differentials for Close Contact Services during a Pandemic

Last registered on May 23, 2022

Pre-Trial

Trial Information

General Information

Title
Compensating Differentials for Close Contact Services during a Pandemic
RCT ID
AEARCTR-0009478
Initial registration date
May 22, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
May 23, 2022, 7:22 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
Vienna University of Economics and Business

Other Primary Investigator(s)

PI Affiliation
Austrian National Bank

Additional Trial Information

Status
In development
Start date
2022-05-23
End date
2022-08-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This project estimates the risk premium individuals are willing to pay to reduce their exposure to a potentially catastrophic event: infection with a potentially lethal disease. We conduct a randomized experiment within a survey about broad economic issues, where we elicit compensating differentials for infection risk minimization during a pandemic (COVID-19) in a close physical services scenario (getting a haircut). The participants will state their additional willingness to pay (in Euro) for the service under different conditions (wearing a FFP2 mask vs. being additionally tested) and different phases (original virus vs. Omicron variant vs. new hypothetical variant). The participants are randomly selected into four treatment groups that differ in the vaccination status (vaccinated vs. unvaccinated) and conducted COVID-19 test (“Antigen” vs “Antigen + PCR”) of the person providing the close physical services for the phases “Omicron” and the new hypothetical variant. We thus obtain a quantification of the willingness to circumvent additional risk, where a broad information set about institutional trust and other socio-economic indicators allows us to identify the determinants of the compensating differential.
External Link(s)

Registration Citation

Citation
Siuda, Fabian and Thomas Zoerner. 2022. "Compensating Differentials for Close Contact Services during a Pandemic." AEA RCT Registry. May 23. https://doi.org/10.1257/rct.9478-1.0
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2022-05-23
Intervention End Date
2022-07-31

Primary Outcomes

Primary Outcomes (end points)
Reported additional willingness to pay (compensating differential) in Euro for a haircut under different conditions.
Primary Outcomes (explanation)
We ask the respondents for their additional willingness to pay for a haircut if (i) the service provider is wearing an FFP2 mask and (ii) the service provider is wearing an FFP2 mask and is tested for COVID-19 in different scenarios differing in their risk exposures.

Before the experiment starts (in the survey part), we ask respondents for the price they paid before the pandemic for a haircut. Depending on the specification the additional willingness to pay will be used in absolute values or relative to the pre-pandemic costs of a haircut. This serves two purposes: (i) test for relative vs. absolute premium and (ii) relative price may be better suited to adjust for the length of the contact (longer duration/additional services may increase the infection risk and also increase the price of the service).

Secondary Outcomes

Secondary Outcomes (end points)
Drivers that influence the additional willingness to pay (compensating differential) in Euro.
Secondary Outcomes (explanation)
The main analysis has the underlying assumption that individuals behave rational, are risk averse and have the ability to fully assess and comprehend the scenario. In this part we aim to identify drivers that cause deviation from this rational behavior. For this we have an extensive set of additional control variables regarding institutional trust and socio-economic indicators.

Experimental Design

Experimental Design
We added our experiment to an existing survey (OeNB Barometer) that is administered by IFES (https://www.ifes.at/), an independent commercial survey agency, for the Austrian National Bank (Oesterreichische Nationalbank, OeNB). IFES draws a random representative sample (N approx. 1,000) of Austrian adult (above 18 years of age) population from registry data (Melderegister). The annual survey is a repeated cross-section and is conducted as computer-assisted personal interviewing (CAPI). The main blocks of the survey concern broad economic topics, such as asset holdings, saving motives, inflation expectations, and trust in institutions. The field phase will start on May 23, 2022 and continue until approximately July 31, 2022. The survey will be in German only. Data will be made available to the researchers in August.

In the experiment, the participants will state their additional willingness to pay (in Euro) for a close contact service (haircut) for different scenarios under different conditions. We ask for three different scenarios: (i) the original virus of early 2020, (ii) the Omicron variant of winter 2021/2022 and (iii) a hypothetical new variant in the future. Before answering the questions for each scenario, we provide scenario cards for all respondents that remind them qualitatively in neutral language about the situation at that particular point in time (text see below, translated from German original).

For all three scenarios, we separate participants into four groups that are being constructed by two independent splits. The first random split where we separate between being tested for COVID-19 with an “Antigen-test” (70%, A) or an “Antigen-test and a PCR-test” (30%, B). The second random split is by vaccination status of the service provider where we separate between being vaccinated (50%, X) and not being vaccinated (50%, Y). As a result, the four groups should approximately have the following sizes: “AX” = 350, “AY” = 350, “BX” = 150, “BY” = 150.

For each scenario, all individuals are being asked for their additional (relative to price of haircut before the pandemic) willingness to pay for the haircut if (i) the service provider is wearing an FFP2 mask and (ii) the service provider is wearing an FFP2 mask and is tested for COVID-19 (according to treatment, “A” vs. “B”). For scenario two and three, the vaccination status (“X” vs. “Y”) applies to both questions (i) and (ii). For scenario one, we do not state the vaccination status. The respondents can report Euro values in integer steps.

We expect the following results:
• We expect no significant differences between the two testing treatments (“Antigen”, A vs. “Antigen + PCR”, B). The reason for this is that both test provide reliable results, when performed adequately. The treatment is mainly intended as a robustness check as the Austrian Covid-protection laws where switching between official acceptance of Antigen and PCR tests throughout the pandemic (also in the close service sector). Thus, for most parts of the analysis we pool those observations (after ensuring that there are no statistically significant differences).
• Ceteris paribus, a higher infection risk should thus increase willingness to pay for risk reduction.
• Ceteris paribus, a higher mortality, conditional in infection, should thus increase willingness to pay for risk reduction.
• Ceteris paribus, vaccination should increase the willingness to pay compared to unvaccinated.
• Within the vaccination treatment, the additional benefit from testing is lower, which should result in a smaller difference between wearing a FFP2 mask and being additionally tested within the vaccination treatment group.


For the experiment, we expect differential results, based on observed and unobserved characteristics. Thus, we will employ a heterogeneity analysis by either splitting the sample or controlling for characteristics.

Sample Slits / Dropping Observations (in some OLS regressions):
• Separating by/controlling for individuals who state positive willingness to pay (vs. negative values or zero)
• Separating by/controlling for sex, education, region, occupation, (household) income (all measured in the survey part)
• Separating by/controlling for risk attitude (as measured in the survey part, e.g. portfolio choice, nature of savings)
• Separating by/controlling for financial literacy, numeracy
• Separating by/controlling for forward looking behavior (as measured in the survey part, e.g. precautionary measures against climate change, inflation expectations along various horizons)
• Separating by/controlling for trust in institutions (as measured in the survey part) or traditional western medicine and pharmaceutical companies (measured after the participants finished the experiment in the survey)
• Separating by/controlling for vaccination status (as stated by participant, measured following the experiment in the survey; and instrumented/proxied using an instrumental/proxy variable approach)


Text scenario cards: (English transcript; participants receive these in German language only)
(i) Original virus of early 2020
• Think back to the first Corona wave of early 2020.
• At that time there is no vaccination against the Corona virus available.
• Uncertainty about potential long-term consequences as well as the mortality, in particular conditional on age and other unknown risk factors, is very high.
• The mortality rate, conditional infection, is considered very high at that moment.
• The number of cases is relatively small, but intensive care units at hospitals are reaching capacity limits due to the large share of severe cases.
(ii) Omicron variant of winter 2021/22
• Think back to the Omicron Corona wave of winter 2021/22.
• At that time there is a vaccination against the Corona virus available that reliably protects at risk people individuals from severe progression.
• The mortality rate, conditional infection, is very low for vaccinated individuals.
• The number of new infections is very high, but the share of severe progression is very low.
(iii) Hypothetical new variant in the future
• Now imagine the following hypothetical scenario.
• There is a new variant of the Corona virus against which the vaccination potentially may not work anymore.
• The mortality rate, conditional infection, is again at the level of the first Corona wave of early 2020.
• The infection rate is very high and at the level of the Omicron wave of winter 2021/22.
Experimental Design Details
Randomization Method
randomization done in office by a computer
Randomization Unit
individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
1,000
Sample size: planned number of observations
1,000
Sample size (or number of clusters) by treatment arms
The four treatment arms should approximately have the following sizes: “Antigen Test & Vaccinated” = 350, “Antigen + PCR Test & Vaccinated” = 350, “Antigen Test & not vaccinated” = 150, “Antigen Test + PCR Test & not vaccinated” = 150.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
WU BEIRAT FÜR ETHISCHE FRAGEN | WU ETHICS BOARD
IRB Approval Date
2022-05-20
IRB Approval Number
WU-RP-2022-014

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials