The Impact of Within-Group Conflict on Trust and Trustworthiness

Yves here. Readers will no doubt notice the importance the authors place on trust as important in making society more efficient. It would be nice if they had unpacked that. What they are presumably referring to is that, for example, in a high-trust society, you can do a lot of business based on a handshake, while in a low-trust society, dealings need to be done on a contractual or governmentally-protected footing. A more mundane example: in a high-trust society, it’s not nutty to leave your door unlocked, while low-trust societies pay more for policing.

We’ve pointed out how the US has become a visibly lower trust society over the last 30 years. For instance, most contracts have become vastly more elaborate. And sneaky fees and “gotcha” provisions have become widespread, if not pervasive. Yet no one has seen fit to call out the rise in sharp practice as bad for commerce generally.

By Alison Booth, Professor of Economics at the Australian National University, and Xin Meng, Professor, Research School of Economics, The Australian National Universit. Originally published at VoxEU

The literature examining the effect of conflict on trust and trustworthiness has reached contradictory conclusions. This column studies the long-term behavioural impact of the Cultural Revolution in China, which was a major in-group conflict. It finds that the children and grandchildren of those who were mentally or physically abused during the Revolution are less trusting, less trustworthy, and less likely to be competitively inclined relative to peers whose parents/grandparents experienced the Cultural Revolution but were not directly mistreated.

Many social scientists are interested in the longer-term effects of socio-political upheavals on economically relevant preferences such as trust and trustworthiness. Trust and trustworthiness are often regarded as important contributors to social capital, which helps to “improve the efficiency of society by facilitating coordinated actions” (Putnam 1993). Societies with less trust and trustworthiness tend to be less coherent and less stable, which may in turn affect long-term economic performance (Nunn 2008). Because of the importance of this topic, economists have investigated what factors affect the level of trust and trustworthiness in a society. In recent years, some studies have found that experience of wars or other conflicts in the society generates distrust and other anti-social behaviour (Fehr 2009, Nunn and Wantchekon 2011), while others have emphasised how war or violent experience can foster pro-social behaviour (for a detailed review, see Bauer et al. 2016).

The contradictory findings may, to a large extent, be related to the nature of the violence or the conflicts experienced. In general, if trauma occurred due to intra-group conflict, trust in others is reduced as ‘betrayal aversion’ comes into effect (Bohnet et al. 2008, Fehr 2009). On the other hand, if the trauma was brought about by inter-group conflict, the within-group bond strengthens and individuals become more cooperative within that group.

In a recent paper on the Cultural Revolution and its intergenerational behavioural impact (Booth et al. 2018), we investigate how this large-scale political upheaval has affected the behaviour of children and grandchildren of those who were mentally or physically abused during the Cultural Revolution, as compared to those who otherwise participated in the Cultural Revolution but were not directly abused.

The Cultural Revolution: A Major Within-Group Conflict

The Cultural Revolution, which began in 1966 and ended ten years later, was the longest lasting political upheaval in the history of the Chinese Communist Party and had the most pervasive impact on contemporary Chinese communities. During the Cultural Revolution, authorities at different levels were challenged and then removed. Moreover, their representatives (government and Communist Party leaders), together with intellectuals, scientists, and artists as well as people who were deemed to be counter-revolutionaries, were openly criticised, denounced, physically tortured and emotionally humiliated. Many were killed and many committed suicide. There are no official statistics recording ‘unnatural’ mortality and the number of people otherwise affected during the Cultural Revolution. Unofficially, the estimates place the death toll at around 1.1 to 7 million and the number of direct victims of some form of political persecution between 11 to 30 million (Yang 2013, Walder 2014). During the first few chaotic years of the Cultural Revolution, perpetrators of mistreatment were in general former colleagues, subordinates, or neighbours. In some cases even friends, relatives, and family members were directly involved. It therefore represents a major within-group conflict.

The Cultural Revolution was a traumatic experience that extended to everybody in all age groups. Previous economic studies of the impact of the Cultural Revolution mainly focus on events that affected society or certain cohorts as a whole, including large-scale schooling interruptions in urban areas and the sent-down youth movement, which assigned hundreds of thousands of urban middle- and high-school graduates to rural China during the Cultural Revolution. Yet, to our knowledge, no economic study apart from ours examines the long-run effects on children or grandchildren of those who directly experienced the denunciation, torture, humiliation and social isolation during the Cultural Revolution.

Intergenerational Behavioural Effects of the Within-Group Conflict

Our study is based largely on a suite of laboratory experiments we conducted in Beijing in 2015, in order to elicit our subjects’ preferences for risk, trust, trustworthiness and competition using games widely used in the experimental literature. The details of the experiment are given in full in Booth et al. (2018). We also complemented our analysis from these laboratory experiments with statistical analysis of two large-scale household survey datasets, the China General Social Survey 2003 and the China Family Panel Survey 2012.

In the experiment, we included a module in the exit-survey questionnaire specifically asking people whether their parents and/or grandparents during the Cultural Revolution experienced one or more of seven different types of ill-treatment, humiliation, torture, or political persecution. This information is then used to identify individuals whose parents/grandparents were directly affected and the degree of the effect. We find that those whose parents/grandparents experienced some types of mistreatment during the Cultural Revolution are less trusting, less trustworthy, and less likely to be competitively inclined relative to their peers whose parents/grandparents were not directly mistreated, but experienced the Cultural Revolution nevertheless.

We further investigated whether the correlations we observed between mistreatments of parents/grandparents during the Cultural Revolution and their children/grandchildren’s lack of trust, trustworthiness, and competitiveness are causal relationships or simply an artefact. By artefact, we mean that those mistreated during the Cultural Revolution might have been a special group of people more likely to have the above mentioned behavioural traits, and those traits might have been transmitted to their offspring with or without their mistreatment due to the Cultural Revolution. To this end, we utilised three different methods.

First, we employed the Altonji et al. (2005) and Oster (forthcoming) style of test by adding a very rich set of parental controls. The idea of the test is simple. If our results do not change much after controlling for the rich additional parental characteristics, this suggests that the link between children/grandchildren’s traits is not related to these characteristics of their parents. Thus, it is indicative that the observed correlation may not be linked to mistreated parents/grandparents during the Cultural Revolution who are characterised by lack of trust, trustworthiness and lower competitiveness to start with, but it was instead the mistreatment during the Cultural Revolution that made them so. Our results confirm this.

Second, using large-scale survey data, we conducted a differences-in-differences analysis, in which we identified individuals who were in their developmental age during the first three years of the Cultural Revolution as being in the Cultural Revolution cohort, and those born afterwards as being in the non-Cultural Revolution cohort. Within each cohort, we also identified those whose parents were party members and were working as government official or intellectuals when they were young. As we indicated above, government officials and intellectuals were the main target of the Cultural Revolution. Using differences-in-differences between the Cultural Revolution and non-Cultural Revolution cohort and those whose parents were the targeted group of the Cultural Revolution and those whose parents were not, we investigate if those whose parents were more likely to be mistreated during the Cultural Revolution were less trusting. Our results confirm this was the case.

Third, we utilised a special feature in the exit survey of our lab experiment to examine whether individuals whose grandparents were mistreated during the Cultural Revolution but who themselves had no direct contact with their grandparents had similar behavioural traits as their counterparts who had contact with grandparents. The answer to this question is no. This suggests that the intergenerational transmission of anti-social traits is not due to genetics (nature), but to role-model effects (nurture).

To conclude, we note that our findings of the intergenerational behavioural effect of the Cultural Revolution should be regarded as lower-bound estimates. The Cultural Revolution affected everybody in society. It affected not only people who were mistreated during the Cultural Revolution (our treatment group) but also those in the control group who witnessed the untrustworthy behaviour of their neighbours and friends even though these behaviours were not directed at them. This should affect their behaviour too. Our finding of statistically significant differences in behaviour traits between the treated and control groups indicates that such society-wide experience had a much stronger impression on those who bore the brunt. But it should not be forgotten that this is not the only group that was affected.

See original post for references

Print Friendly, PDF & Email


  1. Off The Street

    American scholars may have some fertile ground for research on the effects of its own culture wars over the past three decades.

    For supplemental reading, Francis Fukuyama has some observations about trust from a book early in that period.

  2. GeoCrackr

    I’m guessing I’m not the only one who saw the title of this post and thought it was about what the Democratic Party has done to itself over the last couple decades (and the last 3 yrs in particular).

  3. flora

    Thanks for this post. This line seems important:

    During the first few chaotic years of the Cultural Revolution, perpetrators of mistreatment were in general former colleagues, subordinates, or neighbours. In some cases even friends, relatives, and family members were directly involved. It therefore represents a major within-group conflict.

  4. Susan the other`

    This is timely research. The researchers say, preemptively, that intergenerational effects are not genetic but social, learned from a role model. This almost goes without saying because whole families have a personality. Betrayal is the most unforgivable sin; this also goes without saying. Distrust lingers over two generations, they indicate. I think it works it’s way through a third if we are talking about role modeling because traits are seen from grandparents through grandchildren. Makes me wonder about epigenetics at least. Robert Plomin contends that personality is genetic – that everything is genetic in both similarities and differences. He bases this on 50 years of identical twin research. So maybe a better way to look at the effects of trust on a society would be to think in terms of what turns distrust on (genetically) and off again. And politicians might want to take note. When revolution is at hand it is because people feel betrayed and abused. Those feelings last for generations. My grandparents suffered the great depression; my parents were always distrustful of the economy; I have always been a skeptic; and my daughter who never met her great grandparents is a skeptical person as well. In fact, my oldest grandson has all the traits already and he’s only 7. If it is role modeling, it lasts a century by my experience. I would also be willing to speculate that distrust is not as socially destabilizing as this research implies. It may well be a counterbalance to further abuse.

  5. Synoia

    Rudyard Kipling’s Quote seems appropriate:

    The American have an unlimited and meticulous legality, but of law abidingness, not a trace.

    Which certainly speaks to trust, and provides pointed commentary.

    This at a time where in the UK many deals were done by handshake. In the early 50s, I saw my father secure a 100 years lease on commercial property for a tank farm by handshake, in front of his American Boss.

    The boss asked, where’s the contract, and my father’s response was a puzzled “We shook hands on it, we have a deal, the paperwork will follow”.

    1. Cal2

      It’s called cultural affinity. We used to have it in this country.

      The 1924 Johnson-Reed Act established a quota system based on national origins. It directed nearly 70 percent of the immigration slots to northern Europeans, cutting back drastically on immigration from southern and eastern Europe. It maintained formidable barriers against immigration from Asia and Africa, while leaving immigration from the Western Hemisphere unrestricted — a gesture of hemispheric solidarity that also served the cheap-labor interests of American employers.

      The 1965 legislation was named the Hart-Celler Act for its principal sponsors in the Senate and House of Representatives. It abolished the quota system, which critics condemned as a racist contradiction of fundamental American values.

    2. dearieme

      Some years ago I had a near endless series of time-consuming discussions with an American corporate lawyer because she had looked at a nearly five year old royalties contract for some software of mine. She wanted me to sign several new documents. At first it seemed to me that she could not get it into her head that the original deal was that after five years the software would belong to her company and I would have no farther claim on it. Moreover I would have no claim on any improvements her company had wrought.

      It took a while for it to dawn on me that she must presumably have found some tiny flaw in the original wording. But if so she was not prepared to be frank about it. I repeatedly assured her that I gave not a hoot about any potential flaw or ambiguity and that as far as I was concerned the software would be all theirs in a couple of weeks time. I washed my hands of it. That was the intention, the spirit, of the original deal, and I had no intention of ratting on that.

      It was quite a culture clash. I won in the end but only by point blank refusing to sign any new document at all. None. Ever. She probably decided that Britons are an obstinate lot.

      1. Active Listener

        The lawyer likely had an ethical duty to her client not to tell you of the flaw she’d discovered in the existing contract, lest it be exploited against her client. Her job was to protect her client and try to get it fixed. Anything else would be some shade of malpractice. Even if she informed you of the contractual glitch and you chose to honor the intent of the agreement, she’d still know she had committed an ethical breach and would probably have a duty to disclose to her client and her malpractice insurer. Lawyers have a ton of power in this society, but it’s ironic how constrained they are at the same time against doing what “feels good” when they’re representing clients because part of their job is to assume the worst is going to happen and prepare their client for it.

  6. David

    “What they are presumably referring to is that, for example, in a high-trust society, you can do a lot of business based on a handshake, while in a low-trust society, dealings need to be done on a contractual or governmentally-protected footing.”

    Yes, exactly. There are many societies where commercial dealings are a subset and a consequence of good personal relationships, and a certain level of trust that the other person is not going to try to cheat you. I was always struck by the fact that in Japanese the same word is used for a client as for a guest in your home. (I’m told that the same is true in other Asian languages as well.) Trust is remarkably efficient, especially when the alternative is lawyers. I like Goedel’s Incompleteness Theorem, which says essentially that no matter how complex the rules you make, there will always be situations that cannot be predicted from them.
    But it’s also a question of size and proximity. If you know your neighbours and your customers, not only are you likely to trust them, but they are also likely to be more trustworthy because they care about their reputation. In village shops and markets in many countries the trader will have your usual order already and if he hasn’t got the change then it’s “pay me next week.” In our modern, liberal, efficient societies we have essentially replaced these relationships of proximity and trust with contracts. And as Fukuyama’s (actually rather good) book points out, it takes a very long time to create trust, but a very short time to destroy it.

    1. flora

      In our modern, liberal, efficient societies we have essentially replaced these relationships of proximity and trust with contracts.

      I agree. Interesting that relationships of proximity carried their own corrective forces: someone of known bad character would soon find it hard to do business in the area. Word would get around.

      With the rise of desktop computing/network infrastructure, suddenly a scam artist’s activity wasn’t restricted to local areas; he could reach across the country looking for marks. At the same time, the neliberal govt in US looked the other way when scam artists found new ways to defraud people… as long as the scam artists were raking in enough money to give politicians big campaign “contributions”. Gresham’s dynamic came into play. * imo.


  7. thesaucymugwump

    It’s interesting that the single-digit millions fatalities caused by the Cultural Revolution were mentioned, yet the fatalities caused by the Great Leap Forward, generally accepted to be 45 million, were not mentioned.

    Fun fact: it’s common for Chinese couples to have wedding photos taken with them dressed as Red Guards. The RFA article “Everyone is Responsible For This Evil” has an amusing photo.

  8. Oregoncharles

    Responding to Yves’ preface: my father, an investment manager, made the same point 50 years ago – before the decline in trust. He argued that trust – honesty – was worth a lot of money. Yves mentions some examples; his included tenant farm contracts. The owner has to trust the tenant to come up with the numbers. If they don’t, they have to hire accountants and lawyers to make sure. And how do they trust them? But tenancy is usually in the tenant’s interest: they pay a share instead of a fixed rent, hence have lower risk, and they’re entrepreneurs like farm owners.

    His company also made direct investments, not just the stock market; if they smelled anything at all shady, they stayed away, again because monitoring was a large extra expense.

    Just a confirmation from a much earlier era.

    1. flora

      50 years ago the financial malefactors were punished; the law was used to stop fraud and other bad behavior. Fast forward to today… when the frauds in the subprime mortgage market and other Wall St. and TBTF bank scams were brought to light in the GFC the public voted in full Dem control of Congress and the Pres. I knew life-long Repubs who voted for Dems and for Obama thinking the Dems would clean out the financial corruption. Instead, the Dems ‘foamed the runway’ for the banks and protected the crooks. (That was a watershed moment of betrayal, imo.) How does one trust anyone when the govt protects the crooks, when the govt isn’t on the side of the good guys? (Hello, Wells Fargo…) The Clintonist Dems might be courting the moderate suburban Repub voters, but they burned that bridge 10 years ago. My 2 cents.

    2. Yikes

      Capital was costly, not to be wasted, and trust therefore was critical. Now, capital is so cheap and easy to access with a simple bribe such that trust is an impediment to profit, just ask Mitt Romney.

  9. vlade

    I have said now for a long time that all most successful teams/companies/societies share one thing – high level of mutual trust, and a primary role of a leader should be to establish and deepen that.

    1. Epistrophy

      You have to trust others to achieve goals within an organization; even when others are not trustworthy.

  10. tiebie66

    Why would people become untrustworthy? I can understand developing a lack of trust in response to such experiences and perhaps transmitting that, wittingly or unwittingly, to progeny. Becoming untrustworthy oneself seems to reflect a very profound distrust of the system. Is it because being trustworthy in system that cannot be trusted is simply a waste of personal resources?
    It would be interesting to know if there were some that responded the opposite way – becoming more trustworthy themselves. Many times, persons who suffer a great calamity respond in ways to “ensure that it never happens again” or “never happens to someone else”.

  11. cat sick

    In very low trust environments there will often exist sub groups with high trust intra group to whom this becomes a massive advantage to be able to do business with much lower risks within your network, In India this might be extended families, in Asia the Chinese dispora and in many parts of the world Jewish or Masonic networks, being part of one of these groups where you are honour bound to be fair to your fellows but under no obligation to the wider community creates huge relative advantage.

  12. RJMac

    A field study in trust and trustworthiness was conducted fifty plus years ago in Vietnam that I personally observed in III Corps at the 12the Evac Hospital in CuChi in 1970 before, during and after the Cambodian Invasion in May-June. Arriving in February and through March and April, the morale of GI’s ( 11Bravos -the infantry) was surprising good with soldiers of the 25th Division and 1st Air Cav sick with malaria, scrub typhus, etc. wanting to get back to their units. Morale tanked with the Cambodian Invasion with GI’s coming back angry with stories of friends killed pointlessly in undermanned units (up to 50% undermanned), inadequate resupply of ammunition and water as well as inadequate or absent artillery and air support. It was clear that the trust in their leadership had evaporated. This is the story I heard from privates to colonels. At about the same time highly purified heroin (90%) became available from the Golden Triangle and in the ensuing months heroin use sky-rocketed, such that at one point in the summer I had 50 GI’s under my care for either heroin overdose or withdrawal. The lost of trust in an untrustworthy authority in a situation in which one has little control made a deep impression. It is striking how the current narcotic epidemic has a similar origin of loss of trust in an untrustworthy authority ie either corporate America’s outsourcing of millions of jobs or government’s unwillingness or inability to deal with same. Trust seems the central issue to me in our current cultural situation.


Leave a Reply

Your email address will not be published. Required fields are marked *