Citizen Attitudes about Policy Reforms Involving Change in Administrative Requirements

The intervention consist of a 2x2 factorial experimental design embedded in an electronic survey among a sample of about 2,000 Danish residents (age 18+).

Survey respondents are expose to three separate Scenarios. Each scenarios starts by the following text:
“Imagine that there are plans for a reform in the [policy area].”

Each Scenario specify one of three distinct policy areas (text in brackets):

1) “area of employment policy, which includes, e.g., allocation of unemployment benefits and early retirement rights”
2) “area of specialized social policy, which includes, e.g., allocation of public subsidies to cover additional expenses for adults with disabilities”
3) “area of business subsidy policy, which includes, e.g., allocation of subsidies for business projects promoting green transition or making energy consumption more efficient”

Each policy areas appear once across the three Scenarios. The order of the three Scenarios (and, thus, the distinct policy areas) is counter-balanced.

Each Scenario will include the same 2x2 factorial experimental design outlined below ("Experimental Design").

The intervention consist of a 2x2 factorial experimental design embedded in an electronic survey among a sample of about 2,000 Danish residents (age 18+).

Survey respondents are expose to three separate Scenarios. Each scenarios starts by the following text:
“Imagine that there are plans for a reform in the [policy area].”

Each Scenario specify one of three distinct policy areas (text in brackets):

1) “area of employment policy, which includes, e.g., allocation of unemployment benefits and early retirement rights”
2) “area of specialized social policy, which includes, e.g., allocation of public subsidies to cover additional expenses for adults with disabilities”
3) “area of business subsidy policy, which includes, e.g., allocation of subsidies for business projects promoting green transition or making energy consumption more efficient”

Each policy areas appear once across the three Scenarios. The order of the three Scenarios (and, thus, the distinct policy areas) is counter-balanced.

Each Scenario will include the same 2x2 factorial experimental design outlined below ("Experimental Design").

2022-12-09

2022-12-16

A measure of citizen attitudes toward bureaucratic reform involving change in administrative requirements.

We measure our primary outcome by a composite index measure based on responses to three survey items:

a) “Implementing the reform is a good idea”
b) “The reform is a step in the wrong direction” (reversed)
c) “The benefits of the reform are far greater than its disadvantages”

All items are Likert scale items (anchored at 1 = ‘Fully disagree’ and at 7 = ‘Fully agree’). Item order is counter-balanced. We generate the index measure using the item row sums. The final index measure is z-transformed (mean = 0, standard deviation = 1).

Each of the three Scenarios in the survey involve the same 2x2 factorial experimental design.

After the introductory Scenario text (cf. "Intervention"), further text describes the planned reform.
Our experimental design involves manipulation of the exact reform description text.

Treatment 1: By random assignment (50/50), the reform is presented as either:

A. Introducing certain registration- and documentation requirements with the aim of avoiding [fraud/waste/abuse], but at the expense of the administrative work becoming heavier
B. Removing certain registration- and documentation requirements with the aim of easing the administrative work, but at the expense of increasing risk of [fraud/waste/abuse]

We manipulate at random whether the stated gain/cost in program integrity relates to consideration of fraud, waste, or abuse.

Treatment 2: By random assignment (50/50), the reform is presented as having the recipients of introduction/removal of administrative requirements being either:

A. Citizens
B. Public employees

Exact reform text
“The reform involves [T1.1] certain registration- and documentation requirements for [T2] within the policy area. The reform aims to [T1.2]. However, the reform also entails [T1.3].”

The text in brackets is as follows:

T1.1
A = introducing
B = removing

T2
A = citizens
B = public employees

T1.2
[If T1A:]
(1) avoid fraud, including that citizens are denied access to a public benefit for which they are not entitled
(2) avoid waste, including efficient use of government resources; spending only where there is actual need for it
(3) avoid misadministration, including that all citizens' rights and entitlements are respected

[If T1B & T2A:] ease the administrative work for the citizens, so that getting through the administrative process becomes less onerous

[If T1B & T2B:] ease the administrative work for the public employees, so that getting through the administrative process becomes less onerous

T1.3
[If T1A & T2A:] that the administrative work becomes heavier for the citizens, so that getting through the administrative process becomes more onerous

[If T1A & T2B:] that the administrative work becomes heavier for the public employees, so that getting through the administrative process becomes more onerous

[If T1B:]
(1) increased risk of fraud, including that citizens may gain access to a public benefit to which they are not entitled
(2) increased risk of waste, including inefficient use of government resources; spending where there is no actual need for it
(3) increased risk of misadministration, including that not all citizens’ rights and entitlements are always respected

Example of the reform text (received by recipients of T1A (fraud) x T2A):

The reform involves introducing certain registration- and documentation requirements for citizens within the policy area. The reform aims to avoid fraud, including that citizens are denied access to a public benefit for which they are not entitled. However, the reform also entails that the administrative work becomes heavier for the citizens, so that getting through the administrative process becomes more onerous.

Example of the reform text (received by recipients of T1B (waste) x T2B):

The reform involves removing certain registration- and documentation requirements for public employees within the policy area. The reform aims to ease the administrative work for the public employees, so that getting through the administrative process becomes less onerous. However, the reform also entails increased risk of waste, including inefficient use of government resources; spending where there is no actual need for it.

Using this research design, we test the following five hypotheses:

H1: Citizens have similar attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) and reforms that remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity).

H2: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on whether the change in requirements befalls on the public employees or citizens.

H2a: Citizen will have preferences for bureaucratic reforms that add additional administrative requirements (vs. remove existing administrative requirements) when the additional requirements befall on the public employees relative to citizens (and vice versa*).

*Citizen will have preferences for bureaucratic reforms that remove existing administrative requirements (vs. add additional administrative requirements) when the additional requirements befall on the citizens relative to public employees.

H2b: Citizen will have preferences for bureaucratic reforms that remove existing administrative requirements for citizens vs. add additional administrative requirements for citizens.

H2c: Citizen will have preferences for bureaucratic reforms that add additional administrative requirements for public employees vs. remove existing administrative requirements for public employees.

H3: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on political ideology.

- Left-wing-oriented citizen will have preferences for bureaucratic reforms that add additional administrative requirements (vs. remove existing administrative requirements).
- Right-wing-oriented citizen will have preferences for bureaucratic reforms that remove additional administrative requirements (vs. add additional administrative requirements).

H4: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on policy area experience.

- Experience as service recipient will increase preferences for bureaucratic reforms that remove existing administrative requirements for citizens vs. add additional administrative requirements for citizens (H2b).

H5: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on employment experience.

- Experience as public employee will reduce or negate preferences for bureaucratic reforms that add additional administrative requirements for public employees vs. remove existing administrative requirements for public employees (H2c).

Moreover, we test for heterogeneous effects across gender, age, educational level, policy area and dimension of program integrity (for H1 and H2).

Each of the three Scenarios in the survey involve the same 2x2 factorial experimental design.

After the introductory Scenario text (cf. "Intervention"), further text describes the planned reform.
Our experimental design involves manipulation of the exact reform description text.

Treatment 1: By random assignment (50/50), the reform is presented as either:

A. Introducing certain registration- and documentation requirements with the aim of avoiding [fraud/waste/abuse], but at the expense of the administrative work becoming heavier
B. Removing certain registration- and documentation requirements with the aim of easing the administrative work, but at the expense of increasing risk of [fraud/waste/abuse]

We manipulate at random whether the stated gain/cost in program integrity relates to consideration of fraud, waste, or abuse.

Treatment 2: By random assignment (50/50), the reform is presented as having the recipients of introduction/removal of administrative requirements being either:

A. Citizens
B. Public employees

Exact reform text
“The reform involves [T1.1] certain registration- and documentation requirements for [T2] within the policy area. The reform aims to [T1.2]. However, the reform also entails [T1.3].”

The text in brackets is as follows:

T1.1
A = introducing
B = removing

T2
A = citizens
B = public employees

T1.2
[If T1A:]
(1) avoid fraud, including that citizens are denied access to a public benefit for which they are not entitled
(2) avoid waste, including efficient use of government resources; spending only where there is actual need for it
(3) avoid misadministration, including that all citizens' rights and entitlements are respected

[If T1B & T2A:] ease the administrative work for the citizens, so that getting through the administrative process becomes less onerous

[If T1B & T2B:] ease the administrative work for the public employees, so that getting through the administrative process becomes less onerous

T1.3
[If T1A & T2A:] that the administrative work becomes heavier for the citizens, so that getting through the administrative process becomes more onerous

[If T1A & T2B:] that the administrative work becomes heavier for the public employees, so that getting through the administrative process becomes more onerous

[If T1B:]
(1) increased risk of fraud, including that citizens may gain access to a public benefit to which they are not entitled
(2) increased risk of waste, including inefficient use of government resources; spending where there is no actual need for it
(3) increased risk of misadministration, including that not all citizens’ rights and entitlements are always respected

Example of the reform text (received by recipients of T1A (fraud) x T2A):

The reform involves introducing certain registration- and documentation requirements for citizens within the policy area. The reform aims to avoid fraud, including that citizens are denied access to a public benefit for which they are not entitled. However, the reform also entails that the administrative work becomes heavier for the citizens, so that getting through the administrative process becomes more onerous.

Example of the reform text (received by recipients of T1B (waste) x T2B):

The reform involves removing certain registration- and documentation requirements for public employees within the policy area. The reform aims to ease the administrative work for the public employees, so that getting through the administrative process becomes less onerous. However, the reform also entails increased risk of waste, including inefficient use of government resources; spending where there is no actual need for it.

Using this research design, we test the following five hypotheses:

H1: Citizens have similar attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) and reforms that remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity).

H2: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on whether the change in requirements befalls on the public employees or citizens.

H2a: Citizen will have preferences for bureaucratic reforms that add additional administrative requirements (vs. remove existing administrative requirements) when the additional requirements befall on the public employees relative to citizens (and vice versa*).

*Citizen will have preferences for bureaucratic reforms that remove existing administrative requirements (vs. add additional administrative requirements) when the additional requirements befall on the citizens relative to public employees.

H2b: Citizen will have preferences for bureaucratic reforms that remove existing administrative requirements for citizens vs. add additional administrative requirements for citizens.

H2c: Citizen will have preferences for bureaucratic reforms that add additional administrative requirements for public employees vs. remove existing administrative requirements for public employees.

H3: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on political ideology.

- Left-wing-oriented citizen will have preferences for bureaucratic reforms that add additional administrative requirements (vs. remove existing administrative requirements).
- Right-wing-oriented citizen will have preferences for bureaucratic reforms that remove additional administrative requirements (vs. add additional administrative requirements).

H4: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on policy area experience.

- Experience as service recipient will increase preferences for bureaucratic reforms that remove existing administrative requirements for citizens vs. add additional administrative requirements for citizens (H2b).

H5: Citizens’ attitudes toward bureaucratic reforms that add additional administrative requirements (for promoting program integrity at the expense of heavier administrative work) vs. remove existing administrative requirements (for easing administrative work at the expense of protection of program integrity) depends on employment experience.

- Experience as public employee will reduce or negate preferences for bureaucratic reforms that add additional administrative requirements for public employees vs. remove existing administrative requirements for public employees (H2c).

Moreover, we test for heterogeneous effects across gender, age, educational level, policy area and dimension of program integrity (for H1 and H2).

Randomization is carried out by simple randomization by computer (randomization based on a single sequence of random assignments).

The individual survey respondent.

No

2,000 survey respondents

2,000 survey respondents

Respondents are equally distributed across the four cells in the 2x2 factorial experimental design:

T1A & T2A (“introducing” and “citizens”) = 500
T1B & T2A (“removing” and “citizens”) = 500
T1A & T2B (“introducing” and “public employees”) = 500
T1B & T2B (“removing” and “public employees”) = 500

We are powered to detect “small” effects (Cohen 1969, p. 348). Power analysis (G*Power) shows that our sample size (n = 2,000) allows for identification of effects of size f = .0725 (at alpha = .05; power = .90; numerator df = 1 [(2-1)(2-1)]; no. of groups = 4 [2x2]). A test that considers responses only for one of our three Scenarios. Serving as an indicator for our three-way interaction analyses (H4 & H5), a sample size of 500 would allow for identification of effects of size f = .1452.

Analysis Plan.txt