x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
The Political Economy of University Tuition: Representative Experiments on How Information and Design Affect Public Preferences for Tuition
Last registered on February 12, 2019

Pre-Trial

Trial Information
General Information
Title
The Political Economy of University Tuition: Representative Experiments on How Information and Design Affect Public Preferences for Tuition
RCT ID
AEARCTR-0003873
Initial registration date
February 11, 2019
Last updated
February 12, 2019 5:00 PM EST
Location(s)
Region
Primary Investigator
Affiliation
ifo Institute - Leibniz Institute for Economic Research at the University of Munich
Other Primary Investigator(s)
PI Affiliation
University of Munich and ifo Institute
Additional Trial Information
Status
Completed
Start date
2014-04-25
End date
2018-06-26
Secondary IDs
Abstract
Public preferences for charging tuition are crucial for determining higher education finance. To test whether public support for tuition depends on available information and design features, we devise several survey experiments in representative samples of the German electorate (N>19,500). The electorate is divided, with a slight plurality against charging tuition. Providing information on the university earnings premium raises support for tuition by 7 percentage points. The opposition-reducing effect persists two weeks after treatment. Information on fiscal costs and unequal access does not affect public preferences. Designing tuition as deferred income-contingent payments raises support by 16 percentage points, creating a strong majority favoring tuition. The same effect emerges when framed as loan payments. Support for tuition decreases with higher levels and increases when targeted at non-EU students.
External Link(s)
Registration Citation
Citation
Lergetporer, Philipp and Ludger Woessmann. 2019. "The Political Economy of University Tuition: Representative Experiments on How Information and Design Affect Public Preferences for Tuition." AEA RCT Registry. February 12. https://doi.org/10.1257/rct.3873-1.0.
Former Citation
Lergetporer, Philipp and Ludger Woessmann. 2019. "The Political Economy of University Tuition: Representative Experiments on How Information and Design Affect Public Preferences for Tuition." AEA RCT Registry. February 12. https://www.socialscienceregistry.org/trials/3873/history/41440.
Experimental Details
Interventions
Intervention(s)
We investigate how information provision about economic aspects of the higher edu-cation system and design aspects of the tuition payment scheme affect public preferences for tuition. The first set of experiments focuses on randomized information provision about (i) the earnings differential between persons with and without a uni-versity degree, (ii) public cost per university student, and (iii) unequal access to uni-versity education by parental education. The second set of experiments investigates the effects of the following design features on public preferences for university tuition: (i) deferred, income contingent payments; (ii) level of tuition; (iii) tuition targeted towards non-EU citizens. In all experiments, we randomized respondents between a control group who stated preferences for regular tuition without information provision and specification of design features, and different treatment groups.
Intervention Start Date
2014-04-25
Intervention End Date
2018-06-26
Primary Outcomes
Primary Outcomes (end points)
Preference for University Tution
Primary Outcomes (explanation)
“Do you favor or oppose that students at German universities or universities of applied sciences cover a part of the costs of their studies themselves by tuition?” Respondents could pick one of the following five answer categories: “strongly favor”, “somewhat favor”, “neither favor nor oppose”, “somewhat oppose”, and “strongly oppose".
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We implement our experiments in five waves of the ifo Education Survey (2014 to 2018), an annual opinion survey on education policy that we conducted in Germany. Each wave was carried out between April and July of the respective year and covered a nationally representative sample of the German voting-age population (18 years and older). In addition, in the 2018 wave, we sampled adolescents aged between 14 and 17 years. Polling and randomization was carried out by the polling firm KANTAR Public. Within each survey wave, we randomized respondents into different treat-ment arms as follows:

Wave 2014:
Control group: N=1,032
Treatment group 1 (“Earnings information”): N=1,030
Treatment group 2 (“Cost Information”): N=1,056
Treatment group 3 (“Access information”): N=1,053

Wave 2015:
Control group: N=1,390
Treatment group 1 (“Earnings information”): N=1,355
Treatment group 2 (“Income contingency”): N=1,360

Wave 2016:
Control group: N=781
Treatment group 1 (“Income contingency”): N=852
Treatment group 2 (“Level 500”): N=804
Treatment group 3 (“Level 1500”): N=865

Wave 2017:
Control group: N=2,075
Treatment group 1 (“Earnings information”): N=2,003
(Part of the respondents were also re-surveyed two weeks after main experiment to investigate treatment-effect persistence).

Wave 2018:
Control group: N=1,036
Treatment group 1 (“Income contingency”): N=1,005
Treatment group 2 (“Student loans”): N=970
Treatment group 3 (“Non-EU students”): N=1,035
(In addition to adult respondents, we sampled 1,085 adolescents and randomized them between control group (N=525) and treatment “Income contingency” (N=560)).

Description of treatments:
“Earnings information”: Information about earnings differential between persons with and without a university degree provided directly before eliciting preferences for university tuition.
“Cost information”: Information about overall public cost per university student pro-vided directly before eliciting preferences for university tuition.
“Access information”: Information about unequal access to university education by parental education provided directly before eliciting preferences for university tui-tion.
“Income contingency”: Tuition specified as income-contingent tuition that is due after graduation and has to be paid only if income exceeds a certain threshold.
“Level 500”: Tuition level specified at 500 Euros per semester.
“Level 1500”: Tuition level specified at 1500 Euros per semester.
“Student loans”: Treatment “Income contingency” framed as student loans.
“Non-EU students”: Tuition specified to be targeted towards non-EU students.

Importantly, we report treatment effects for all survey participants in the representative adult samples, and the adolescent sample who have non-missing stated preferences for tuition. Sample sizes were pre-specified with our polling firm. Thus, we do not engage in selective reporting of results on parts of our sample.
Experimental Design Details
Randomization Method
Randomization is carried out by the survey company KANTAR Public, using a computer.
Randomization Unit
at the individual level
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
19,702 responses from adults (19,577 non-missing responses from 15,013 individual respondents)
1,085 responses from adolescents (1,085 individual respondents)
Sample size: planned number of observations
19,577 responses from adults (15,013 individual respondents) 1,085 responses from adolescents (1,085 individual respondents)
Sample size (or number of clusters) by treatment arms
Wave 2014:
Control group: N=1,032
Treatment group 1 (“Earnings information”): N=1,030
Treatment group 2 (“Cost Information”): N=1,056
Treatment group 3 (“Access information”): N=1,053

Wave 2015:
Control group: N=1,390
Treatment group 1 (“Earnings information”): N=1,355
Treatment group 2 (“Income contingency”): N=1,360

Wave 2016:
Control group: N=781
Treatment group 1 (“Income contingency”): N=852
Treatment group 2 (“Level 500”): N=804
Treatment group 3 (“Level 1500”): N=865

Wave 2017:
Control group: N=2,075
Treatment group 1 (“Earnings information”): N=2,003
(Part of the respondents were also re-surveyed two weeks after main experiment to investigate treatment-effect persistence).

Wave 2018:
Control group: N=1,036
Treatment group 1 (“Income contingency”): N=1,005
Treatment group 2 (“Student loans”): N=970
Treatment group 3 (“Non-EU students”): N=1,035
(In addition to adult respondents, we sampled 1,085 adolescents and randomized them between control group (N=525) and treatment “Income contingency” (N=560)).

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
Yes
Intervention Completion Date
June 26, 2018, 12:00 AM +00:00
Is data collection complete?
Yes
Data Collection Completion Date
June 26, 2018, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
19,702 responses from adults (19,577 non-missing responses from 15,013 individual respondents)
1,085 responses from adolescents (1,085 individual respondents)
Was attrition correlated with treatment status?
No
Final Sample Size: Total Number of Observations
19,702 responses from adults (19,577 non-missing responses from 15,013 individual respondents)
1,085 responses from adolescents (1,085 individual respondents)
Final Sample Size (or Number of Clusters) by Treatment Arms
Wave 2014: Control group: N=1,032 Treatment group 1 (“Earnings information”): N=1,030 Treatment group 2 (“Cost Information”): N=1,056 Treatment group 3 (“Access information”): N=1,053 Wave 2015: Control group: N=1,390 Treatment group 1 (“Earnings information”): N=1,355 Treatment group 2 (“Income contingency”): N=1,360 Wave 2016: Control group: N=781 Treatment group 1 (“Income contingency”): N=852 Treatment group 2 (“Level 500”): N=804 Treatment group 3 (“Level 1500”): N=865 Wave 2017: Control group: N=2,075 Treatment group 1 (“Earnings information”): N=2,003 (Part of the respondents were also re-surveyed two weeks after main experiment to investigate treatment-effect persistence). Wave 2018: Control group: N=1,036 Treatment group 1 (“Income contingency”): N=1,005 Treatment group 2 (“Student loans”): N=970 Treatment group 3 (“Non-EU students”): N=1,035 (In addition to adult respondents, we sampled 1,085 adolescents and randomized them between control group (N=525) and treatment “Income contingency” (N=560)).
Data Publication
Data Publication
Is public data available?
No

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers