The Benjamin Franklin Effect: Why Doing Someone a Favor Can Make You Like Them

The Benjamin Franklin effect is the claim that people tend to like someone more after they have done that person a favor. It sounds backward. The usual assumption is that we help people because we already like them, and that the way to win someone over is to do something nice for them. The Franklin effect says the causation can also run the other way: helping someone can produce liking, and so asking a person for a small favor can be a way to make them warmer toward you.

The idea takes its name from a story Benjamin Franklin told in his autobiography about winning over a political opponent by borrowing a book from him. It entered social psychology through a small experiment published in 1969, and it is usually explained using cognitive dissonance theory, one of the best-known ideas in the field. In recent years it has also become a staple of self-help and sales writing, often stated with much more confidence than the evidence supports.

This article covers Franklin's original story, the experiments behind the effect, the competing explanations for why it might happen, the reverse version of the effect, and what the evidence can and cannot support. It also covers the common ways the idea gets distorted in popular retellings.

Franklin's story

The setting

In 1736 Franklin was chosen, without opposition, as clerk of the General Assembly of Pennsylvania, the colony's legislature. The position was renewed every year. It mattered to Franklin for practical reasons as well as political ones. Besides the clerk's pay, it helped him keep the Assembly's printing business: the votes, laws, paper money, and other public jobs that were valuable to a printer.

The following year, 1737, when he was up again, a newly elected member of the Assembly made a long speech against him in favor of another candidate. Franklin describes this man as a gentleman of fortune and education, with talents that seemed likely to give him growing influence in the House. Franklin was chosen anyway. But he did not want a talented and influential member as a lasting enemy.

The book

Franklin says he did not try to win the man over by flattering him or showing him "servile respect." Instead, having heard that the man owned a very scarce and curious book, he wrote asking whether he could borrow it for a few days. The man sent it immediately. Franklin returned it about a week later with a note expressing his thanks.

The next time they met in the House, the man spoke to him, something he had never done before, and did so with great civility. He was afterward ready to serve Franklin on all occasions. The two became friends, and the friendship lasted until the man's death.

The maxim

Franklin presents the episode as proof of what he calls an old maxim he had learned: "He that has once done you a kindness will be more ready to do you another, than he whom you yourself have obliged." He draws the moral that it is more profitable to remove an opponent's hostility this way than to resent it, return it, and keep up the conflict.

What the story does and does not show

Franklin's story is an anecdote, written down decades after the fact by a man who had every reason to present himself as shrewd. It is not an experiment. We do not know what else was going on between the two men, whether the political situation shifted, or whether the man would have warmed to Franklin anyway. It is also worth noticing what Franklin's maxim actually says. It is about the helper's readiness to help again, a claim about future behavior, and only indirectly about liking.

There is a second subtlety. Asking to borrow a rare book is not just a request for help. It is also a compliment. It tells the owner that his library and his taste are valued, and that Franklin trusts him with a small dependency. Some of the man's change of heart may have come from that flattery and trust rather than from the act of lending itself. This point returns in the modern research, where it turns out to matter a great deal.

The psychological background: cognitive dissonance

The Franklin effect became a topic in psychology because it seemed to fit a theory that was hugely influential in the 1950s and 1960s: Leon Festinger's theory of cognitive dissonance, set out in his 1957 book A Theory of Cognitive Dissonance.

The basic idea

Festinger proposed that people experience discomfort when they hold beliefs, attitudes, or knowledge about their own behavior that conflict with one another. He called this discomfort dissonance. Because dissonance is unpleasant, people are motivated to reduce it. They can do so by changing their behavior, changing one of the conflicting beliefs, or adding new beliefs that make the conflict seem smaller.

Often the easiest thing to change is an attitude, because behavior that has already happened cannot be undone. If I have already done something, I can reduce dissonance by deciding I must have had good reasons to do it.

The classic demonstrations

Two famous experiments show the kind of reasoning that later got applied to favors.

In Festinger and Carlsmith's 1959 study, students performed deliberately boring tasks, such as turning pegs on a board, and were then asked to tell the next participant that the tasks had been enjoyable. Some were paid $1 for telling this lie, and others $20. Afterward, the students paid only $1 rated the tasks as more enjoyable than those paid $20. The interpretation was that the $20 students had a sufficient external reason for lying, while the $1 students did not, and so resolved the dissonance by coming to believe the tasks had not been so bad.

In Aronson and Mills' 1959 study, women who went through a more embarrassing initiation to join a discussion group later rated a deliberately dull group discussion as more interesting than women who went through a mild initiation. This is the basis of the idea known as effort justification: we come to value what we have suffered for.

Applying it to favors

The application to the Franklin effect is straightforward. Suppose I do a favor for someone I do not particularly like. Now I hold two thoughts that sit awkwardly together: "I don't like this person" and "I went out of my way to help this person." One way to resolve the tension is to revise the first thought. Perhaps I like this person more than I realized, since otherwise why would I have helped? On this account, doing a favor produces liking as a way of justifying the behavior after the fact.

The key experiments

Jecker and Landy (1969)

The study usually cited as the experimental foundation for the Franklin effect is Jon Jecker and David Landy's "Liking a Person as a Function of Doing Him a Favour," published in the journal Human Relations in 1969.

The design, as described by Wikipedia and in David McRaney's popular account, was as follows. Students took part in what they thought was an intellectual task in which they could win money. The researcher running the session was, by McRaney's description, deliberately rude and demanding. The tasks were rigged so the participants won. After the session, the participants were split into three groups:

Direct favor. The researcher himself approached the participant and asked whether they would return the money. He explained that he was funding the study with his own money and was running short.
Indirect request. A secretary approached the participant and asked them to return the money to the psychology department's research fund. In other words, the request came from an institution, not from the researcher as a person.
No request. The participant kept the money.

Afterward, all participants rated how much they liked the researcher. McRaney reports that nearly everyone agreed to the request. He gives the average liking ratings, on a 12-point scale, as 7.2 for the group that did the researcher a personal favor, 5.8 for the group that kept their money, and 4.4 for the group asked by the secretary. The ordering, with the direct favor group liking him most and the secretary group liking him least, matches Wikipedia's summary. The original paper is not freely available, and those exact figures come from a popular-science source rather than from the paper itself.

The usual interpretation is that the people who gave money back to the researcher as a personal favor needed a reason to justify doing so, especially for someone who had been unpleasant to them, and found it in deciding he was likable after all.

Two features of the result are often skipped in retellings. First, the effect depended on who asked. Returning the money to the department, which was the same action with the same cost, did not raise liking and appears to have lowered it. Second, this is a single study from 1969 with a small sample by modern standards. It has not been the subject of large, preregistered replications of the kind that have become standard since psychology's replication crisis.

Schopler and Compere (1971)

A study by John Schopler and John Compere, published in the Journal of Personality and Social Psychology in 1971 and carried out at the University of North Carolina, looked at the reverse side of the same idea. Participants played the role of "teachers" whose "learners" (actually confederates) had to reproduce patterns of taps on wooden cubes. Some teachers were instructed to encourage their learners when they got things right. Others were instructed to insult them when they made mistakes.

Afterward, the teachers who had been harsh rated their learners as less attractive and less likable than the teachers who had been kind. Wikipedia summarizes the lesson as: "You tend to like the people to whom you are kind and dislike the people to whom you are rude." Here the treatment was assigned by the experimenters, so the teachers' attitudes seem to have followed from their behavior rather than causing it.

Niiya (2016)

The most recent experiment commonly cited on the effect is by Yu Niiya of Hosei University in Tokyo: "Does a Favor Request Increase Liking Toward the Requester?", published in The Journal of Social Psychology in 2016 (online in September 2015). Wikipedia describes it as smaller than the earlier work.

Niiya ran the study with both Japanese and American participants. There were three kinds of situations:

a confederate directly asked the participant for help;
the participant helped the confederate without having been asked directly;
a baseline comparison in which no help occurred.

Participants who had been directly asked for help rated the confederate as more likable than baseline. They also rated the confederate as more sociable and reported feeling closer to them. Helping without a direct request did not produce this increase.

This last result is important. If the Franklin effect were purely about cognitive dissonance, helping should produce liking whether or not you were asked, because in both cases you have done something for the person that needs justifying. Niiya's results suggest instead that the request itself carries social meaning. Being asked for a favor signals that the requester wants a closer relationship, and the helper responds by reciprocating that warmth.

Niiya connected the finding to the Japanese concept of amae, which the psychiatrist Takeo Doi made widely known in his book The Anatomy of Dependence (published in Japanese in 1971 and in English in 1973). Amae refers roughly to the comfortable presumption on another person's goodwill that characterizes close relationships, such as a child's dependence on a parent. In that framework, asking for something is a way of expressing and building intimacy, not only a burden placed on the other person.

Why might it happen? Competing explanations

There are at least four explanations for why doing a favor might increase liking. They are not mutually exclusive, and the evidence does not clearly favor one over the others.

1. Cognitive dissonance

As described above, the helper has to reconcile "I helped this person" with "I don't like this person," and adjusts the attitude. Wikipedia also describes this as maintaining consistency in one's self-concept: people like to think of themselves as acting for good reasons.

This explanation fits Jecker and Landy's direct-favor condition. It fits Schopler and Compere's results in reverse, because harming someone needs justifying too. It struggles with the indirect-request condition in Jecker and Landy and with Niiya's finding that unrequested help does not raise liking. In both cases the helper has still done something costly for the person, so dissonance theory would seem to predict increased liking.

2. Self-perception

In 1967, the psychologist Daryl Bem proposed self-perception theory as an alternative to dissonance theory. Bem argued that people often infer their own attitudes the same way an outside observer would, by watching their own behavior and the circumstances in which it happened. No uncomfortable tension is needed. If I see myself lending a book to someone with no obvious pressure or reward, I may simply conclude that I must like them, the same way I would conclude that about someone else who did the same thing.

Self-perception predicts many of the same results as dissonance theory, and psychologists spent years trying to tell the two apart. For the Franklin effect, it offers a simpler account of the basic result. It shares the same difficulty with the conditions where helping did not produce liking.

3. Reciprocity of liking

Niiya's own explanation is reciprocity. A direct, personal request is read as a friendly overture: it signals that the requester trusts you, values you, and wants a closer relationship. People tend to like those who seem to like them, so the helper returns the warmth. This explains why it matters who makes the request. When a secretary asks on behalf of a department, there is no personal overture to reciprocate.

This explanation also fits the details of Franklin's own story. His request was personal, it flattered the man's taste and learning, and it was followed by a gracious note of thanks.

4. Investment and commitment

A fourth possibility is that helping someone creates a small investment in the relationship. Once a person has given something, they have a stake in the relationship going well, and so are inclined to see the other person favorably. This is related to effort justification. It is also related to the "foot-in-the-door" phenomenon described by Jonathan Freedman and Scott Fraser in 1966, in which people who agree to a small request become more likely to agree to a larger one later. Franklin's maxim, that someone who has done you one kindness is readier to do another, is close to a description of foot-in-the-door.

Summary of the explanations

Explanation	Core mechanism	Fits direct-favor result?	Explains why unrequested or institutional help may not increase liking?
Cognitive dissonance	Need to justify costly behavior	Yes	Not easily
Self-perception	Infer attitude from own behavior	Yes	Not easily
Reciprocity of liking	A personal request signals warmth, which is returned	Yes	Yes
Investment and commitment	Giving creates a stake in the relationship	Yes	Partly

The table is a simplification. Real interactions probably involve several of these processes at once.

The reverse effect: disliking those we harm

If doing good for someone can increase liking, the reverse is plausible as well: harming someone may lead people to dislike them, as a way of justifying the harm. Schopler and Compere's results point in this direction, since teachers who insulted their learners came to rate them less favorably.

Wikipedia's article suggests that this converse process could help explain how people treat those they have mistreated. It mentions soldiers and jailers who come to see their victims as deserving it, and long-running feuds such as the one between the Hatfield and McCoy families, where each act of harm becomes a reason for the next. These are suggestions rather than tested findings. Real conflicts have many causes, and it would be a mistake to treat a laboratory result about insulting confederates as an explanation of historical violence. The general idea, that people justify their own harmful acts by thinking less of their targets, is a recurring theme in research on dissonance and on derogation of victims.

How the effect is studied, and why that matters

Experiments on the Franklin effect share a basic structure, and understanding it helps in judging what the results mean.

The usual design

A researcher needs three things: a situation in which a participant can be asked for a favor, a comparison situation in which no favor is asked, and a way of measuring liking afterward. In practice the favor is usually small and arranged by the experimenters. It might involve giving back money, helping with a task, or doing something for a confederate, an actor who is secretly part of the study. Liking is almost always measured with a questionnaire: participants rate the other person on numbered scales for things like likability, sociability, or closeness.

Problems with self-report

Questionnaire ratings are convenient, but they measure what people say about their feelings in a laboratory, shortly after an unusual experience, not how they behave toward the person over weeks or months. A participant might give a higher rating partly out of politeness, or because they guess what the researcher wants. Psychologists call such cues "demand characteristics." None of the classic studies followed participants to see whether the favor changed any real relationship, which is what Franklin's story is actually about.

The role of the person asking

Another difficulty is that the person asking for the favor is not a neutral party. In Jecker and Landy's study, the researcher who asked for the money was also the person whose likability was being rated, and the request involved an explanation about his personal finances. That explanation could itself make him seem more sympathetic, separately from the act of helping. Disentangling "I helped him, so I like him" from "he confided in me, so I like him" is hard. Niiya's design, which compared requested with unrequested help, was one attempt to do this. Its result suggested that the request itself carried much of the effect.

Cultural context

Niiya's choice to compare Japanese and American participants reflects a long-standing question in social psychology: do findings from mostly American university students generalize to other cultures? Doi's theory of amae was originally presented as a distinctively Japanese pattern of relating. That made it plausible that requests would be read differently in Japan than in the United States. Niiya found the effect of a direct request in both samples, which suggests that the relationship-building meaning of a request is not unique to one culture. It is still a single study, and two countries are a small slice of human societies.

Ethics and deception

Like much social psychology of the 1960s and 1970s, the early studies relied on deception. Participants were misled about the purpose of the tasks, about the researcher's finances, and in Schopler and Compere's case about the identity of the "learners." Modern research ethics still permit some deception, but require review boards and debriefing. This is one reason exact replications of the classic designs are not always straightforward to run today.

Franklin's autobiography as a source

It is also worth knowing what kind of document the original story comes from. Franklin began writing his autobiography in 1771, addressing the first part to his son William. He worked on it at intervals for the rest of his life, and it was left unfinished when he died in 1790. The Assembly episode he describes took place in 1737, more than three decades before he began writing.

The autobiography is a deliberately instructive book. Franklin uses episodes from his life to illustrate maxims and habits he wanted readers to learn, the best known being his list of thirteen virtues. The book-borrowing story is presented in exactly this way, as an example proving the truth of an old saying. That does not make it untrue, but it does mean the story has been shaped to make a point. Franklin does not name the rival in the passage, so it is not possible to check the episode against the other man's side of events.

How strong is the evidence?

The Franklin effect is a good case study in how a modest experimental finding can become a confident popular "law."

What supports it

A memorable historical anecdote from Franklin.
Jecker and Landy's 1969 experiment, in which a personal favor appears to have increased liking.
Schopler and Compere's 1971 experiment, showing that kind or harsh behavior shaped later attitudes toward the person treated that way.
Niiya's 2016 experiment, which found increased liking after a direct request across Japanese and American samples.
A broad, well-established research tradition showing that behavior can shape attitudes, including dissonance and self-perception research.

What weakens it

Small, old studies. The core experiments are few, small by modern standards, and mostly decades old. Several popular sources note that the effect has not been confirmed through large-scale replication. No large preregistered replication of the effect turned up in research for this article.
Dependence on conditions. In both Jecker and Landy and Niiya, the effect appeared only under specific conditions: when the request was personal and direct. That is a significant qualification that most retellings omit.
Uncertainty about mechanism. The standard explanation, dissonance, does not fit all the results, and the alternative explanation, reciprocity, implies that the effect is about requests rather than favors.
Broader replication problems in social psychology. Since the early 2010s, many classic findings in social psychology have proved weaker than originally reported when retested with larger samples. That does not mean the Franklin effect is false. It means that a small number of small studies should not be treated as settled.

A fair verdict

The most defensible statement is something like this: there is some experimental evidence that people who are personally and directly asked for a small favor, and who agree, later report liking the requester somewhat more. The evidence is limited, it depends on how the request is made, and the reasons are debated. It is a plausible and interesting effect, not an established law of human nature.

Practical implications and their limits

Popular articles routinely turn the Franklin effect into advice. Ask your boss, a rival, or a potential client for a small favor, and they will like you more. Some of that advice is reasonable, and some goes far beyond the evidence.

What seems reasonable

People are more willing to help than we expect. This is a separate but related finding. In a 2008 paper in the Journal of Personality and Social Psychology, Francis Flynn and Vanessa Lake reported six studies showing that people asking for help underestimated how likely others were to say yes to a direct request, by as much as 50 percent, across both laboratory and field settings. The help-seekers focused on what it would cost the other person to agree, and overlooked how socially awkward it is to refuse. If you avoid asking for help because you assume people will be annoyed or will refuse, you may be wrong on both counts.

Personal requests can open relationships. Niiya's research and the amae framework both suggest that a direct, sincere request can function as an invitation to closeness. Many people find this true of their own experience. A new colleague who asks for advice is often easier to warm to than one who never asks for anything.

Gratitude matters. Franklin returned the book with a note of thanks. Expressing appreciation for help is widely recommended, and it fits the reciprocity explanation: it confirms that the request was a friendly overture rather than an imposition.

What goes beyond the evidence

"Any favor will make anyone like you." The research suggests the opposite: the favor has to be requested personally and directly, and a burdensome or awkward request can harm a relationship. Popular summaries also tend to ignore that the requests in the experiments were small.

"This is a manipulation technique." Some sales and dating advice presents the effect as a trick. Apart from the ethical problems, the reciprocity explanation suggests the effect works because the request is read as a genuine sign of interest. Requests that come across as calculated may well backfire.

"It works on enemies." Franklin's opponent was a political rival, not a personal enemy, and Franklin had something to offer the relationship. There is no good evidence that asking a favor of someone with real hostility toward you will reliably change their mind.

Hypothetical illustrations

The following examples are hypothetical, constructed to show how the different explanations would predict different outcomes. They are not reports of real events or studies.

Hypothetical 1. A new employee wants to build a relationship with a senior colleague who seems distant. She asks him for his opinion on a draft report, explaining that she admires how clearly he writes. He reads it, gives a few comments, and she thanks him afterward, mentioning a change she made because of his advice. The dissonance, self-perception, and reciprocity explanations all predict he will view her somewhat more warmly. The request was personal, small, and flattering, and it was followed by gratitude.

Hypothetical 2. The same employee instead sends a department-wide email asking anyone available to proofread her report. The colleague helps. Here the reciprocity explanation predicts little or no change in liking, because the request was not a personal overture. Jecker and Landy's institutional-request group and Niiya's no-direct-request condition are the closest experimental parallels.

Hypothetical 3. The employee asks the colleague to spend his weekend rewriting her report. He agrees reluctantly. Dissonance theory might predict that he justifies the effort by deciding he likes her. Common sense, and the warnings in the research, suggest he is just as likely to resent being imposed on. None of the cited experiments tested large, costly favors, so the evidence does not tell us which outcome is more likely.

Related concepts

Several better-studied ideas overlap with the Franklin effect:

Cognitive dissonance. The broader theory most often used to explain it.
Effort justification. Valuing things more because we worked or suffered for them.
Self-perception theory. Inferring our attitudes from our own behavior.
Foot-in-the-door technique. Agreeing to a small request makes agreeing to a larger one more likely. Franklin's maxim describes something very close to this.
Reciprocity. The strong social norm of returning favors, and of returning liking to those who seem to like us. Robert Cialdini's book Influence (1984) made reciprocity famous as a principle of persuasion. Note that standard reciprocity predicts that the person who receives a favor will feel obliged, while the Franklin effect concerns the person who gives one.
Amae. The Japanese concept of comfortable dependence on another person's goodwill, used by Niiya to frame requests as relationship-building.

The Wikipedia article on the effect also mentions Dale Carnegie's How to Win Friends and Influence People (1936), whose advice about letting other people feel important is in a similar spirit. It notes that the same general principle has been applied in fields as different as mentoring and dog training.

Common misconceptions

"Franklin discovered the effect scientifically." He told an anecdote and quoted what he called an old maxim. The experimental work came more than two centuries later.

"The effect is well established and widely replicated." The core evidence consists of a handful of small studies. It has not been tested in large modern replications.

"Doing any favor makes you like the person." In the two main favor experiments, the increase in liking appeared when the request came personally and directly from the person being helped. Institutional requests and unrequested help did not show the same effect.

"It is definitely caused by cognitive dissonance." Dissonance is the traditional explanation, but it does not fit all the results. Reciprocity of liking and self-perception are serious alternatives.

"Franklin's maxim is about liking." Franklin's maxim says the helper will be "more ready to do you another" kindness. That is a claim about future helping. His story does describe a resulting friendship, but the maxim itself is closer to the foot-in-the-door effect than to a pure claim about attitudes.

"The effect is the same as reciprocity." Ordinary reciprocity is about the receiver of a favor feeling obliged to return it. The Franklin effect is about the giver of a favor coming to like the receiver. Niiya's explanation brings them together by treating the request itself as something the helper reciprocates.

Frequently asked questions

What is the Benjamin Franklin effect in one sentence?

It is the observation that people sometimes come to like someone more after doing that person a favor, especially when the person asked them for it directly.

Where does the name come from?

From Benjamin Franklin's autobiography. He described winning over a hostile fellow member of the Pennsylvania Assembly in 1737 by asking to borrow a rare book from him and returning it with a note of thanks.

What was the original experiment?

Jecker and Landy (1969). Participants won money in a rigged task and were then asked to return it, either personally by the researcher or by a secretary on behalf of the department, or were not asked at all. Those who returned money to the researcher personally rated him as most likable.

Is the effect real?

There is some experimental support for it, but the evidence base is small and old, and the effect seems to depend on how the request is made. It is best treated as plausible but not firmly established.

Why would helping someone make you like them?

The traditional explanation is cognitive dissonance: we adjust our attitudes to justify our behavior. An alternative is self-perception, in which we infer our feelings from our own actions. Yu Niiya's research suggests a third: a direct personal request signals that the other person wants to be closer, and we respond to that warmth.

Does the reverse happen?

There is some evidence for it. In Schopler and Compere's 1971 experiment, participants instructed to treat someone harshly later rated that person as less likable. The broader idea that people derogate those they have harmed, to justify the harm, appears often in social psychology. It should not be treated as a full explanation of real-world conflicts.

Can I use it to get people to like me?

Asking a colleague or acquaintance for a small, reasonable favor, and thanking them sincerely, is unlikely to hurt and may help. Research also suggests people are more willing to help than we expect. But large, burdensome, or obviously calculated requests may have the opposite effect, and the research does not support the idea that this is a reliable technique.

Is it the same as "foot in the door"?

They are related. Foot-in-the-door, described by Freedman and Fraser in 1966, is the finding that agreeing to a small request makes people more likely to agree to a larger one later. Franklin's maxim, that someone who has done you a kindness is "more ready to do you another," describes something very similar. The Franklin effect, as studied in psychology, focuses on liking rather than compliance.

Why do people hesitate to ask for favors if it might help?

Part of the answer comes from Flynn and Lake's research. People asking for help tend to focus on the cost to the other person of saying yes, such as their time and effort. They pay less attention to how uncomfortable it is for the other person to say no. As a result they underestimate how often others will agree. Many people also worry that asking makes them look weak or incompetent, and in some situations that worry may be justified. The Franklin effect research suggests that, at least for small and personal requests, the social cost of asking may be lower than people fear.

Does the size of the favor matter?

The classic experiments used small favors: returning modest winnings or helping briefly with a task. There is no good experimental evidence about what happens with large or costly favors. Dissonance theory would predict that bigger sacrifices need more justification and could produce more liking. Everyday experience suggests that big impositions can also breed resentment. Until the question is studied directly, it is safest to assume the effect, if it exists, applies to small and reasonable requests.

Summary

The Benjamin Franklin effect starts from a story and a maxim. Franklin turned a political opponent into a friend by asking to borrow a book, and concluded that people who have helped you are readier to help you again. Psychologists later tested a version of the idea. A 1969 experiment found that people who returned money to a researcher as a personal favor rated him more favorably. A 1971 experiment found that people asked to treat someone harshly came to think less of them. A 2016 experiment in Japan and the United States found that a direct request for help increased liking, while unrequested help did not.

The traditional explanation is cognitive dissonance, but the details of the results, especially the importance of who makes the request and how, point toward a social explanation as well: a personal request is a gesture of trust and friendliness, and people return it. The evidence is thin by modern standards, and the effect is far less universal than popular articles suggest. Still, the core observation, that the act of helping can shape how we feel about the person we helped, is consistent with a large body of research showing that our actions shape our attitudes as much as our attitudes shape our actions.