Why do some communities trust more than others? Experimental evidence from the North-South divide in England

Last registered on October 28, 2024

Pre-Trial

Trial Information

General Information

Title
Why do some communities trust more than others? Experimental evidence from the North-South divide in England
RCT ID
AEARCTR-0014660
Initial registration date
October 26, 2024

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
October 28, 2024, 1:35 PM EDT

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

Locations

Primary Investigator

Affiliation
Norwegian School of Economics

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2024-10-28
End date
2024-11-22
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Trust and trustworthiness are integral components of cooperation: a behaviour which drives economic growth and prosperity. However, the levels of such vary dramatically across the world. Drivers of trust heterogeneity include institutions, socioeconomics and cultural norms. In this experiment, I investigate intra- and inter-cultural trust levels across the so-called North-South divide in England using a risk-adjusted trust game. The setting is of high policy relevance due to the national government’s increasing focus and investment in spreading economic opportunity across the country. In reproducing results from the World Value Survey, I aim to decompose the trust differential into socioeconomic factors and belief asymmetries across the divide. I then study how specific trust differs across individuals above and below the divide. Following this I test the efficacy of a collaborative intervention in bridging the gap in trust.
External Link(s)

Registration Citation

Citation
O'Leary, Ethan. 2024. "Why do some communities trust more than others? Experimental evidence from the North-South divide in England." AEA RCT Registry. October 28. https://doi.org/10.1257/rct.14660-1.0
Experimental Details

Interventions

Intervention(s)
Participants will be recruited and categorised by their self-assigned identity: Northern or Southern (if other identity is applciable then participants are not accepted). Participants observe 10 real trust and dictator games and are asked to guess the distribution of outcomes from these games. The intervention is the identity of the decision makers.
Participants are also asked to guess the distribution of trusting beliefs among 100 individuals of some identity.
All decisions are incentivised.
Intervention (Hidden)
Intervention Start Date
2024-10-28
Intervention End Date
2024-11-22

Primary Outcomes

Primary Outcomes (end points)
Trust in general English person among Northerners and Southerners.
Trust in Northerners among Northerners and Southerners.
Trust in Southerners among Northerners and Southerners.
Primary Outcomes (explanation)
The trusts are measured by taking the belief an individual holds on the distribution of trustworthiness among a sample of the target population. From this, the conditional expectation of trustworthiness is computed. Using subjective weights on the probability of the states of world, unconditional expectations can also be computed.

Secondary Outcomes

Secondary Outcomes (end points)
Beliefs of the distribution of trust among some population: General England; Northerners; Southerners.
Secondary Outcomes (explanation)
Participants are asked how many of 100 individuals of the target population agree with the statement 'Most people can be trusted'

Experimental Design

Experimental Design
The experiment is conducted using English participants on Prolific.
Experimental Design Details
1/ Pre-screening

Participants in the experiment need to be of either two identities. This pre-screening will be present at the beginning of each session for both the 'game' and the spectator setup and thus warrants its own explanation. I identify an individual as a Southerner (Northerner) if they meet the following criteria. First the individual was born in a Southern (Northern) region (Southern regions- London, South East, South West and East of England. Northern regions - North East, North West, Yorkshire and the Humber). Second the individual currently lives in a Southern (Northern) region. Third, the individual self-identifies with the corresponding identity.

The aim of this screening depends on the experiment session that is being run. In categorising individuals, I wish to identify if one has moved into or out of some region which would alter their exposure to other identities. Under the contact hyopthesis, this may affect one's trusting beliefs and pro-social behaviour. For instance, given that there is substantial net-South migration in England (ONS, 2023), Southerners will naturally have a higher exposure to Northerners than the contrary: a point I may exploit in a follow-up experiment.

In the 'game', I need to recruit at least 10 Northerners and 10 Southerners for each role. The remaining participants may be of any English origin. Therefore, I adopt a recruitment quota policy in the pre-screening.

In the spectator setting, the aim is to exclude individuals who do not identify as a Northerner or a Southerner. This is since these are my two identities of interest in the experiment.

Participants who do not meet the recruitment criteria in any stage are screened out of the study and paid a small compensation (£0.10) for their time. Participants who pass are moved on to the next stage of the experiment.

2/ The 'game'

The purpose of the game is not to measure or identify differences in trust beliefs or behaviour between identities but merely to provide real incentives for the spectator round. Participants are paid £1.00 as a base wage for participation and have the opportunity to earn up to £18.00 in bonus payment.

The game is split into two sub-experiments. Each experiment recruits 30 individuals. In the first sub-experiment, a recruitment quota system is used to ensure that there are at least 10 Northerners and 10 Southerners in the pool. All other participants may be of any English identity. The quota is implemented via the pre-screening feature. Participants are informed that they will play 2 games with another individual that is English. The players do not learn the identity of the person that they play these games with. In practice, I pair participants up across these two sub-experiments: I do not need these individuals to play simultaneously.

Players in the 'game' are told that in each game, they will be assigned a role of either person A or person B, with their partner playing the remaining role. They are informed that each token has a monetary value of £2.00 and that they will be paid their token balance in cash for one of the two games. They then proceed to play a trust game and a dictator game in a randomised order.

In the trust game they are given the following instructions: 'Person A starts with 3 tokens, and Person B starts with none. Person A can send some, all, or none of their tokens to Person B. Any tokens kept by Person A are theirs. For every token sent, Person B receives 3 tokens. For example, if Person A sends 2 tokens, Person B will get 6 tokens. Person B then decides how many tokens to return to Person A. Person A's earnings are the tokens they kept plus any returned to them. Person B's earnings are what they received minus what they returned.' In practice, players in the first sub-experiment are assigned the role of person B (the trustee) and players in the second sub-experiment are assigned the role of person A (the trustor).

The trustees are asked to state how many tokens they would like to return to person A for each possible amount that they could receive (0, 3, 6 or 9 tokens). This is a strategy method of elicitation. The trustees are told that their partner will make their decision also and this, together with their own decisions, will determine their payoff.

In the dictator game, players are given the following instructions: 'Person A first receives 5 tokens while person B receives nothing. Person A may send some, none, or all of their tokens to person B. Person A's earnings from this round are the tokens they do not send. Person B's earnings equal the tokens they receive.' In practice, players in the first sub-experiment are assigned the role of person A (the dictator) and players in the second experiment are assigned to the role of person B (the receiver). The dictators are asked to decide how many tokens they wish to send. The receivers do not make a decision in this stage.

Before each game, participants are required to pass a comprehension test before they proceed.

3/ The spectator

Spectators are recruited to guess the distribution of outcomes among 10 pairs who played the game. Spectators must also pass the pre-screening test. This way we can measure the trust of Northerners and Southerners separately. Spectators are paid £1.00 for their participation in the survey.

For each game, the spectators are given the same instructions as the players of the game and are then told that they must guess the frequency at which each possible decision was made among 10 trustees. Thus they are asked to elicit their beliefs of 3 distributions (one for each conditional transfer amount from the trustor) and are awarded £2.00 if they guess the exact frequency distribution for each question. After being given a chance to adjust their estimates, the spectators are then asked to guess the distribution of initial transfers in the trust game among 10 trustors.

The same procedure occurs for the dictator game where the spectators must state their beliefs about the distribution of transfers made by 10 dictators in the game. Spectators are awarded £2.00 if they guess the exact distribution.

As a final check, I ask spectators how much they agree with the following statement using a 7 point Likert scale. I then ask them to state their belief of how many of 100 individuals of a specific identity agree or strongly agree with this statement. The statement is 'Most people can be trusted'. Spectators are awarded £0.50 for each guess that is within 3 of the correct answer.

4/ Treatments

Before making any decisions, spectators are assigned to a treatment cell. I inform spectators that they will guess the distribution of outcomes among 10 decision makers. In the general treatment, I inform spectators that the decision makers are English and no further information is given. In the in-group treatment, I inform spectators that the decision makers are of the same identity as the spectators. In the out-group treatment, I inform spectators that the decision makers are of the other experimental identity to the spectators.

When I ask individuals how many of 100 individuals of some identity agree with the statement, I use other spectators to form this group of 100 and thus assign the identity that they observed in the other games to be the reference group.


Randomization Method
Randomisation performed by oTree programme during experiment.
Randomization Unit
Individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
N/A
Sample size: planned number of observations
900 spectators.
Sample size (or number of clusters) by treatment arms
150 spectators per treatment arm x identity cell.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Using a chi-squared and regression test. To detect an 8% difference in conditional expectation of trustworthiness between Northerners and Southerners, a sample size of 80 was needed for constant standard deviations. A larger sample size is opted for to identity difference in difference analyses.
IRB

Institutional Review Boards (IRBs)

IRB Name
NHH IRB
IRB Approval Date
2024-09-12
IRB Approval Number
NHH-IRB 42/22
Analysis Plan

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

Request Information

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