Seeking to provide empirical evidence to international non-profit organizations on the best way to issue an apology, and why they should care about giving a good apology.
International development agencies can cause harm to the populations that they are trying to serve, and for that reason, we started our problem as
“How can organizations issue apologies for caused harm in a way that allows the population being served to feel the most respected?”
A lot of the research we reviewed highlighted their own special frameworks but there was a lot of overlap between scholars. In particular, this list of seven things below was what literature consistently outlined as components of a good apology. The overall takeaway from this list is that apologies need to acknowledge and take responsibility for a wrong-doing, explain why it happened, offer reparations, and quite literally say “I’m sorry”.
For our study, we chose to focus on the efficacy of reparations as there is not much research on the actual effects of including reparations in apologies. It is easy to issue a general statement, but we need more evidence to push organizations to recognize the importance of including reparations.
Given the findings from our literature review, our research question is,
“How can the offer of reparations help organizations in the International Development issue more effective apologies?”
In order to answer the research question, we decided to conduct an online experiment.
For the study population, we chose to focus on Colombia. The reason we chose Colombia is that it was the country that received the most foreign aid in the Spanish-speaking world, and we had access to a Spanish translator and wanted to focus on a country outside of the US that had some proximity to foreign aid organizations. In addition, there is also a need for more research on dignity and respect for dignity in south America.
We ran an online study on SurveyMonkey to collect our sample. The general design of the study was to present participants with a vignette outlining the harm that an International Aid organization had caused in their community and an apology issued for that harm. For our measures, we asked participants a series of questions about their perception of the apology and how it respected their dignity, amongst a few other things.
For all treatments we started by asking participants general demographic questions, then we presented them with our vignette outlining harm. After that participants were asked attention check questions and participants were then presented with an apology issued for the harm outlined in a vignette.
In all conditions, the base apology issued contained all of the elements of a ‘good’ apology that were identified in our literature review. The only difference between the conditions was the absence or presence of an offer of reparations in the apology and the vagueness of this offer, which varied based on the condition. After the apologies, we asked participants to answer a series of Likert scale questions relating to our measures.
For the vignette, while doing their work, the organization had caused an outbreak of cholera when waste from the organization’s headquarters contaminated the local water supply. Participants were told that 50 people in the community were killed by the outbreak, including a cousin of theirs.
We based the scenario on a real event that occurred in Haiti in 2010 when UN peacekeepers caused an outbreak of cholera that devastated the community they were attempting to aid. We found that this occurrence encapsulated a common type of offense that occurs in International development that we called, “serious but unintended harm”. When harm is caused in International development, the organizations involved often intend to help a community, but harm occurred by “accident” or by “mistake”.
By creating a vignette focusing on a similar kind of harm, we could encourage organizations to consider the importance of apologies in these cases when blame isn’t entirely clear and harm isn’t caused intentionally, but the consequences are still real and important to the communities being affected.
For our study, we split participants randomly into 3 treatment groups — the control condition, the vague offer of reparations condition, and the concrete offer of reparations condition.
As stated previously, after reading the vignette, participants were shown an apology issued by the CEO which contained a different offer of reparations depending on the condition.
In the control condition, the apology had no offer of reparations at all.
In the vague offer condition (Treatment 1), participants were shown the vague promise that the sentence above in the apology statement they received. In this condition there’s no clear promise what exactly the organization plans to do, only that they are planning to do something.
Lastly, in the concrete offer condition (Treatment 2), the apology contains a monetary offer of reparations to both ends the cholera outbreak and to help the families that have been directly affected. The monetary amounts are based on the offer made by the UN in the real occurrence but adjusted to our scenario and local currency.
We came up with 5 outcomes to test. For each measure, we presented participants with statements and then a 7-point Likert scale asking how much they agreed with the statement.
Perceived goodness of the apology statement
Perceived “respectedness” of participants’ dignity
Perceived believability of the apology statement
Power restoration
Willingness to welcome the NGO back to their community
The main hypotheses for our study are:
The inclusion of an offer of reparation will have a positive effect on all measures in our study in comparison to the control group (no offer of reparation).
The concrete offer of reparations will have an even greater positive effect than the vague offer.
To examine whether the offer of reparations affects the effectiveness of an apology, we conducted these three tests to check if there are any group differences across our three conditions.
After checking the residuals of our dependent variables, we found that they are not normally distributed. So, we decided to use the Kruskal-Wallis test, which is the non-parametric equivalent of the one-way ANOVA test. This test is typically used to compare the differences between three or more independent groups.
We also ran linear regression to examine if the offer of reparations in our treatment conditions would be statistically significant from the control condition.
We found no significant difference between our two treatment groups and control, as well as between the two treatment groups.
The means of all dependent variables lie between 4 and 5 out of a 7-point Likert Scale. This shows that participants in all of our conditions are neutral to somewhat agree to all of our measurements.
Overall, our results found the apologies with offers of reparations did not yield statistical differences than the one without. And a vague offer of reparation is essentially the same as a concrete one in our experiment.
In International Development, organizations can cause harm to the populations that they are trying to serve; how can organizations issue apologies for caused harm in a way that allows the population being served to feel the most respected?
For the next steps, we recommend more research be conducted to really start understanding the relationship between mistakes in international development and apologies.
Conducting field studies with actual apologies
We think field studies with actual apologies would be interesting - how does actually following through on an apology affect feelings of dignity and respect?
Conducting more experiments
We also think that lab or online experiments could be helpful in figuring out what ideal reparations look like for different scenarios, how timing can impact feelings of respect, and how people would respond to reparations for offenses that don’t really fall into our serious but unintended harm
Beyond a standard apology, we believe there may be other interventions that can be implemented to help increase the respect for a population’s dignity after harm has been committed. So, we came up with some potential interventions that are based on research but would, of course, need to be tested for actual effectiveness. They target both interactions with the community and actions within the organization.
We start by suggesting holding public forums with communities after transgressions. We also suggest facilitating interactions outside of giving aid to build an actual relationship with the community, to ensure that the community and the organization understand each other. Having respected community members in positions of leadership working with the organization can also have a greater impact than just issuing an apology. Understanding the needs, concerns, and wants of the harmed community by visiting members of it can give insights into the community’s expectations. And finally committing to internal re-designs after transgressions have occurred.
“Apologies aren't meant to change the past, they are meant to change the future.”
— Kevin Hancock
As this quote says, apologies aren't meant to change the past, they are meant to change the future.
Even though our study returned null results, we feel strongly about the topic of apologies and we hope that our work can contribute to the research of apologies and their importance in International Development and that this research might help to create a better future.
We believe that organizations should make an effort to take responsibility for wrongdoing when it occurs and take steps to make reparations to the communities that they’ve wronged and to respect their dignity.
If apologies are not enough to make people feel truly respected, which our results indicate, then organizations should try and take further steps to engage with the community they’ve harmed, and maybe even consider taking similar steps from the start, which may be able to prevent some harms from occurring in the first place.
Dignity Project — https://dignityproject.net/
Published report — https://dignityproject.net/our-research/respectful-apologies/
Unsplash stock image — Photo by Flavia Carpio on Unsplash