Parenting by lying

Abstract

The present set of studies identifies the phenomenon of `parenting by lying', in which parents lie to their children as a means of influencing their emotional states and behaviour. In Study 1, undergraduates (n = 127) reported that their parents had lied to them while maintaining a concurrent emphasis on the importance of honesty. In Study 2 (n = 127), parents reported lying to their children and considered doing so to be acceptable under some circumstances, even though they also reported teaching their children that lying is unacceptable. As compared to European American parents, Asian American parents tended to hold a more favourable view of lying to children for the purpose of promoting behavioural compliance.

Introduction

Prohibitions against lying have a long history in Western society. Aristotle asserted, `the least initial deviation from the truth is multiplied later a thousand fold' and St. Augustine declared, `when regard for truth has been broken down or even slightly weakened all things will remain doubtful'. Kant went as far as to suggest that even if a would-be murderer were to inquire as to the whereabouts of his intended victim, one is morally required to tell him the truth. In contemporary American society such prohibitions are alive and well; to call someone a liar is to deliver a significant insult (Lewis, 1993). A recent poll found that Americans consider a candidate's honesty to be the single most important factor in choosing a president, ranking it as more important than positions on the issues and questions of leadership, experience and intelligence (Fournier & Tompson, 2007).

Given the strong prohibition against lying it is not surprising that an emphasis on honesty is central to parents' efforts to socialise their children (Stouthamer-Loeber, 1986; Barnes, 1994) and that children's lies are frequently met with punishment (Lewis, 1993; Robinson, 1996). Parents often emphasise the importance of truth-telling with stories such as The boy who cried wolf and the apocryphal story about young George Washington declaring `Father, I cannot tell a lie' when asked if he had chopped down a cherry tree.

Despite the broad disapproval of lying, parents do not always indicate to their children that lying is wrong. For example, a parent might lie to a child to protect the feelings of others or to keep family matters private (Robinson, 1996). In addition, parents sometimes lie to other adults in their children's presence (Lewis, 1993; Robinson, 1996).

This research considers another way in which parents might condone lying: by making false claims to their children. Although it has been noted that parents often make false claims to their children within the contexts of fantasy play or teasing, or regarding entities such as the Tooth Fairy or Santa Claus (Clark, 1995), almost nothing is known about whether parents do so in other contexts. Suggestive that they might do so is evidence that adults lie to each other quite frequently, for a wide range of reasons (Camden et al., 1984; Ekman, 1985; Miller & Tesser, 1988; Bell & DePaulo, 1996; DePaulo & Bell, 1996; DePaulo et al., 1996; DePaulo & Kashy, 1998). For example, in a diary study involving a community sample, DePaulo and Kashy (1998) found rates of reported lie telling as high as one in every five social interactions. However, there are also reasons to believe that parents would be strongly motivated to avoid lying to their children. By telling their children only what they actually believe, parents have the opportunity to demonstrate the value of honesty to their children. Parents might also be eager to insure that their children see them as reliable sources of information.

In one of the only published papers on the topic of parental lying, Brown (2002) reports findings from ethnographic research conducted with Tzeltal-speaking Mayan corn farmers who live in the rural community of Tenejapa, in Southern Mexico. Parents in this community frequently lie to young children while attempting to control their behaviour. They often do this by making idle threats. For example, children are commonly threatened with wasp stings or dog bites or with the prospect of being kidnapped. Parents justify these lies in terms of the ends they are hoping to accomplish, which typically involves controlling the child's behaviour. Parents do not appear to view these lies as morally problematic and explicitly teach child caregivers to engage in this practice. This research provides strong and consistent evidence that there is no culturally universal taboo against parents lying to their children. However, it is important to note that these practices take place within a cultural context in which it is assumed that everyone lies in the service of self-interest and lying is seen as morally neutral (Brown, 2002).

It is unclear whether a similar acceptance of parental lying would be seen in a society such as that of the USA, where lying is subject to widespread disapproval. Finding an answer to this question should provide important insights into how children are socialised to reason about different types of lies. This socialisation process is important because honesty is a key issue in moral development and by adolescence a complex set of beliefs about the contexts in which lying is acceptable has usually been developed (Perkins & Turiel, 2007). Even by the time they reach early elementary school, children have learned to make distinctions among different types of lies (Bussey, 1999; Heyman et al., 2009; Peterson et al., 1983). For example, they find lies more acceptable when the purpose is to protect the feelings of others than when the purpose is to conceal one's transgressions. However, little is known about what types of direct and indirect messages children receive from socialising agents that might serve to guide their developing beliefs about lying.

The research

We investigated whether parents in the USA lie to their children to influence their behaviour or emotional states, a practice we refer to as parenting by lying. Participants were asked to respond to scenarios that describe parents lying to their children to encourage appropriate behaviour or to make the children happy. In Study 1, undergraduate students were asked to report on their parents' behaviour. In Study 2, parents were asked to report on their behaviour toward their children. The goal of each study was to determine whether there would be any evidence of parenting by lying and, if so, whether the parents who most strongly promoted the importance of being honest with their children would be less likely than other parents to engage in parenting by lying.

Because parenting practices play a key role in the transmission of cultural values, we sought to determine whether the phenomenon of parenting by lying is related to broader cultural factors, in light of evidence of cross-cultural differences in moral evaluation related to beliefs about lying. For example, modesty-related lies that involve a denial of responsibility for performing good deeds are viewed more favourably among East Asians than North Americans (Lee et al., 1997, 2001). In one such study, Lee et al. (1997) found that Chinese children gave more positive ratings to protagonists who falsely denied helping clean up at school than did Canadian children. Cross-cultural differences have also been documented in situations in which individuals are asked about lying to promote the interests of one's group. Fu et al. (2007) found that Chinese children evaluated these types of lies more favourably than lies told to benefit an individual friend, whereas Canadian children tended to show the reverse pattern. Lee and colleagues (e.g. 1997) have interpreted these patterns of results as reflecting differences in sociocultural practices, with a greater emphasis on collectivist values in East Asia and on individualism in North America (Triandis, 1995; Oyserman et al., 2002). Specifically, the strong emphasis on maintaining harmonious interpersonal relationships in East Asia may lead to a greater acceptance of lies that help individuals fit in with the group and that promote the needs of the group (Lee et al., 1997).

As a starting point in examining parental lying in relation to broader cultural factors, we examined parenting by lying among Asian American versus European American parents. Previous research suggests that these groups may be subject to different cultural norms concerning the moral evaluation of lying including a greater emphasis among Asian Americans on social values other than honesty (Gao et al., 1996; Seiter et al., 2002). Additionally, there is evidence that European American parents focus more on children's self-esteem (Chao, 1995; see also Miller et al., 2002) and that Asian American parents focus more on promoting obedience and respect (Lin & Fu, 1990; Chao, 1995). We predicted that European American parents would be more likely to lie to promote positive feelings and that Asian American parents would be more likely to lie to promote appropriate behaviour.

Study 1

Methods

Participants

Participants were 127 (39 male, 86 female, two unreported) undergraduates who were enrolled in a psychology class. According to participants' reports, the sample was 35% European American, 40% Asian American, 13% Hispanic American, 1% African American and 11% multiple categories or `other'.

Materials and procedure

Participants were asked to read and rate nine scenarios, each of which concerns a mother who lies to her six-year-old child. The scenarios were developed with the aid of pre-testing that involved two groups of undergraduates who did not participate in the primary studies. The first group of undergraduates (n = 40) were asked to report an instance in which one or both of their parents told them something that they did not believe and to specify their approximate age at the time of the event. Nearly every undergraduate reported at least one instance. From this set of responses, we selected a group of 15 false claims that could be easily adapted into scenarios that would be appropriate for children aged six and younger. In making these selections, we excluded statements that involved adult themes, such as lying about drug use. We also excluded stories that were identified as fantasy play or teasing, or mythical characters like Santa Claus or the Easter Bunny.

A second group of undergraduates (n = 20) was asked to classify the apparent motive of the mother in each of the 15 scenarios as either shaping the child's behaviour or promoting positive feelings. These ratings were done individually and participants were asked to select which of the two possibilities was the best fit. Because we were particularly interested in examining the distinction between behaviour and emotion scenarios, we only selected scenarios on which participants showed a high level of agreement in these classifications. Specifically, we selected the nine scenarios for which there was at least 90% agreement on the mother's goal: six scenarios that involved a lie intended to shape the child's behaviour (behaviour scenarios), and three scenarios that involved a lie intended to promote positive feelings (emotion scenarios). Further information about these scenarios appears in Appendix 1. The first example below is a behaviour scenario; the second is an emotion scenario:

• Reyna's child was crying when they went shopping together. Embarrassed, she told the child, `The police will come to make sure that you behave if you don't stop crying now', even though Reyna knows that it is not true.

• Catherine has a child who likes rainbows. When she and her child saw a rainbow, she told the child `the rainbow came out just for you' to make the child happy, even though Catherine knows that it is not true.

Following each scenario, participants were presented with a lie-telling measure in which they were asked whether their parents had said something similar to them, with a response scale from 1 (absolutely no) to 7 (absolutely yes). In this and other scales in the present study, only scale endpoints were labelled.

After the nine scenarios, participants were presented with a set of honesty emphasis measures in which they were asked to characterise their parents' approach to lying and truth-telling. One was an honesty promotion measure concerned with how strongly their parents encouraged them to be honest, from 1 (not very strongly) to 7 (very strongly). In addition, a consequences measure concerned how severely their parents would discipline them for lying as compared to stealing or fighting, on a scale from 1 (much less severely) to 7 (much more severely).

Finally, in two open-ended measures, participants were asked what their parents had taught them about lying and given an opportunity to provide an example of a lie a parent had told them.

Results and discussion

Honesty emphasis

Participants reported that their parents had strongly encouraged honesty. The mean response on the honesty promotion measure was 6.19 (SD = .97) and the mean response on the consequences measure was 4.45 (SD = 1.54). There were no significant differences between European Americans and Asian Americans on either of these ratings.

Further evidence of parents' emphasis on honesty came from participants' open-ended responses: 79% reported being taught that lying is unacceptable (e.g. `there is no such thing as a white lie; all lies are bad'); 17% reported being taught that lying is acceptable under some circumstances (e.g. `little white lies are sometimes okay', `lie when it is for people you don't know [but] you can't lie to your parents and family', `tell the truth unless you have to lie to keep a secret'); and 4% fell into neither category (e.g. `they told me to trust my judgment and use my own discretion').

Lie-telling

Table 1 presents results from the lie-telling measure for each scenario. These results are not consistent with the possibility that parents strongly and consistently avoid lying to their children.

Table 1.

Mean lie-telling scores from Study 1

Scenario	Lie-telling score (SD)
Behaviour
Food	3.20 (2.10)
Police	3.37 (2.04)
Money	3.67 (2.00)
Bogeyman	3.33 (1.97)
Monster: sleep	3.36 (2.03)
Monster: vegetables	2.25 (1.49)
Emotion
Uncle	3.71 (2.13)
Rainbow	3.20 (1.98)
Cleaning	3.61 (2.26)

Note: Response scale ranged from 1 = definitely no to 7 = definitely yes.

Also notable is that 88% of participants gave a positive response (5 or greater on the 7-point scale) to at least one of the lie-telling questions and many offered examples of parental lies. Some of the lies appeared to be directed at promoting positive feelings, such as complimenting a child's cooking even though the outcome was `terrible'. Others involved threats of abduction by witches (`if you go outside alone, a witch will fly off with you'), beggars (`if you go outside alone, beggars will kidnap you') or the bogeyman (`if you don't pay attention, the bogeyman will steal you'). Additional examples are shown in Appendix 2.

Scenario and ethnicity effects

To examine scenario and ethnicity effects, an ANOVA was conducted on mean lie-telling scores from the European American and Asian American participants only, with Ethnicity (European American, Asian American) as a between-subjects factor and Scenario (behaviour, emotion) as a within-subjects factor (see Table 2). There was a main effect of Scenario, with emotion scenarios receiving higher lie-telling scores, F_{(1, 94)} = 4.7, p < .05, p_rep = .90, η_p² = .05. There was also a significant interaction, F_{(1, 94)} = 8.52, p < .005, p_rep = .98, η_p² = .08: despite a lack of significant ethnic differences in the emotion scenario ratings, Asian American participants gave higher ratings to the behaviour scenarios than did European Americans, t₍₉₄₎ = 2.63, p < .01, p_rep = .96, suggesting that their parents were more likely to lie to them to attempt to control their behaviour.

Table 2.

Mean lie-telling scores from Asian American and European American participants in Study 1

Scenario	Lie-telling score (SD)
Behaviour scenarios
European American participants	2.79 (1.24)
Asian American participants	3.47 (1.30)
Emotion scenarios
European American participants	3.74 (1.40)
Asian American participants	3.33 (1.53)

Note: Sub-sample included 45 European American and 51 Asian American participants.

Are parents who strongly promote honesty less likely to lie?

There was no evidence that the parents who strongly promoted the importance of honesty were less likely to lie to their children than were other parents. The only significant correlation between the honesty emphasis and lie-telling measures was in the reverse direction: for the behaviour scenarios, the lie-telling and consequences measures were positively correlated, r₍₁₂₅₎ = .22, p < .05, p_rep = .88. This suggests that the parents who were the most punitive about lying, as compared to other transgressions, were also more likely to lie to their children.

In sum, participants in Study 1 reported that their parents expressed the opinion that lying is unacceptable, but engaged in parenting by lying, nevertheless. Additionally, Asian American participants were more likely to report that their parents had lied to them as a means to influence their behaviour.

Study 2

The goal of Study 2 was to determine whether the results of Study 1 would be replicated in a sample of parents.

Participants

Participants were 127 parents (22 male, 100 female, 5 unreported) who reported living in the USA and having one or more children of at least two years of age. The study was conducted through an online questionnaire. About a quarter of the sample was recruited through online parent groups and the rest were recruited using other strategies, which included recruitment through schools and workplaces.

According to participants' reports, the sample was 52% European American, 32% Asian American, 7% Hispanic American, 2% African American and 7% multiple categories or `other'. Of the Asian American parents, 18% reported that they were born in the USA.

Participants were asked about their annual family income: 2% reported that it was less than $25,000, 13% between $25,000 and $50,000, 13% between $50,000 and $75,000, 17% between $75,000 and $100,000, 31% between $100,000 and $150,000, 7% between $150,000 and $200,000, 10% greater than $200,000 and 7% declined to give a response. Participants were also asked about their highest education level: 2% reported some high school, 5% a high school diploma, 12% some college, 11% associates degree, 36% bachelors degree, 32% graduate degree and 1% declined to give a response.

Materials and procedure

The scenarios and measures were the same as in Study 1, except that participants were asked about their own parenting practices. An evaluation question was included in which participants were asked to make a moral judgement about what the parent in each scenario had said, on a scale from 1 (very bad) to 7 (very good).

Results and discussion

Preliminary analyses

Preliminary analyses suggested that patterns of response were highly similar for those recruited online versus those recruited through other strategies. There were no significant differences between reported income or education levels between European American and Asian American participants.

Honesty emphasis

Results indicate that participants perceive themselves as strongly committed to teaching their children that lying is wrong. The mean response on the honesty promotion measure was 6.49 (SD = .92) and the mean response on the consequences measure was 4.67 (SD = 1.33). There were no significant differences in ratings between European American and Asian American participants on the honesty promotion measure. However, Asian Americans were significantly more likely to provide higher ratings on the consequences measure, t₍₁₀₀₎ = 3.19, p < .005, indicating that they were more likely to think that children should be disciplined more severely for lying as compared to stealing or fighting.

Further evidence of this emphasis on honesty can be seen from the open-ended reports of what participants teach their children: 74% reported teaching that lying is unacceptable (e.g. `Baby Jesus knows when they lie so they should always tell the truth', `we do not lie in this family—it is a sin; the truth is always told', `lying is for bad people and witches; good people and fairies never tell lies'); 16% of participants reported teaching that lying may be acceptable under some circumstances (`white lies are okay, especially to avoid hurting people's feelings or to save somebody from physical harm') and 10% fell into neither category (`we haven't really discussed it yet since he's only three').

Lie-telling and evaluation

Results of the lie-telling and evaluation measures are shown in Table 3. As in Study 1, the results are inconsistent with the notion that parents strongly and consistently avoid lying to their children. Indeed, parents sometimes viewed parental lying favourably.

Table 3.

Mean lie-telling and evaluation scores from parents in Study 2

Scenario	Lie-telling score (SD)	Evaluation score (SD)
Behaviour
Food	1.84 (1.48)	1.96 (1.34)
Police	2.28 (1.82)	2.22 (1.44)
Money	2.80 (2.01)	2.86 (1.61)
Boogeyman	2.29 (1.89)	2.26 (1.57)
Monster: sleep	2.02 (1.70)	1.97 (1.14)
Monster: vegetables	1.39 (1.02)	1.57 (1.02)
Emotion
Uncle	3.53 (2.20)	4.80 (1.66)
Rainbow	3.83 (2.13)	4.57 (1.49)
Cleaning	3.82 (2.13)	3.92 (1.82)

Note: Response scale ranged from 1 = definitely no to 7 = definitely yes on the lie-telling measure; 1 = very bad to 7 = very good on the evaluation measure.

Also notable is that 78% of parents gave positive responses (5 or greater on the 7-point scale) to at least one lie-telling question. Many provided examples of lies they had told to their children. Two examples follow; others appear in Appendix 2.

• I was attempting to get my four year-old daughter out of the bathtub…I told her that sitting back down was going to make the germs she'd just washed off think she really liked them and would cause them to jump back on.

• My son never lets me leave him. One night my husband and I had dinner plans. When my friend Linda called, my son asked me `Who is it, Mommy?' An idea came to my mind. I said, `Oh it was my friend Linda. The witch got her. You will have to stay with your brother so daddy and mommy can go help her.' I did it because I didn't want him to cry when I left and ruin my dinner.

Scenario and ethnicity effects

As in Study 1, an ANOVA was conducted on mean lie-telling scores to examine scenario and ethnicity effects among the European American and Asian American participants (see Table 4). There was an effect of Scenario, F_{(1, 104)} = 113.39, p < .001, p_rep > .99, η_p² = .52, with emotion scenarios receiving higher lie-telling scores and Ethnicity, F_{(1, 104)} = 16.30, p < .001, p_rep > .99, η_p² = .14, with higher lie-telling scores among Asian Americans. There was also an interaction between Ethnicity and Scenario, F_{(1, 104)} = 4.62, p < .05, p_rep = .90, η_p² = .04: as in Study 1, there were no significant ethnicity differences for the emotion scenarios, but Asian American participants provided higher ratings for the behaviour scenarios, t₍₁₀₄₎ = 6.00, p < .001, p_rep > .99, which suggests that they were more likely than the European American parents to lie to their children to try to influence their behaviour. Corresponding analyses of the evaluation measure showed the same pattern of scenario and ethnicity effects.

Table 4.

Mean lie-telling and evaluation scores from Asian American and European American parents in Study 2

Scenario	Lie-telling score (SD)	Evaluation score (SD)
Behaviour scenarios
European American participants	1.63 (0.71)	1.71 (0.60)
Asian American participants	2.73 (1.20)	2.81 (1.12)
Emotion scenarios
European American participants	3.47 (1.36)	4.24 (1.05)
Asian American participants	3.96 (1.61)	4.65 (1.27)

Note: Sub-sample included 66 European American and 40 Asian American participants.

Are parents who strongly promote honesty less likely to lie?

There was no evidence that the parents who most strongly promoted the importance of honesty were less willing to lie to their children, as indicated by the lack of a significant correlation between the honesty emphasis measure and the evaluation or lie-telling measures. Similarly, there was no evidence that a concern with promoting honesty was associated with negative evaluations of parenting by lying.

Summary

The results of Study 2 were quite similar to those of Study 1. Parents reported teaching their children that lying is unacceptable, but lied to them nevertheless. The ethnic differences seen in Study 1 were replicated, with Asian American participants more likely to endorse lying for the purpose of controlling their children's behaviour.

General discussion

These findings indicate that American parents lie in order to influence their children's behaviour and emotions. Thus we refer to this practice as parenting by lying. This was reported even though parents promoted the value of honesty. There was no evidence that the parents who placed the greatest emphasis on honesty were less likely than other parents to lie to their children.

One might question, based on verbal report data, whether it is appropriate to conclude that parents lie to their children. Although honesty is a domain for which social desirability effects are likely to be strong (see Heyman & Legare, 2005), it should be noted that the key findings were replicated across two studies that used different research strategies: Study 1, in which adult children were asked about their parents, and Study 2, in which parents were asked about their own parenting practices. In addition, any social desirability effects would presumably be in the direction of denying or minimising parental lying, rather than exaggerating it.

Parenting by lying was not limited to the telling of white lies to be polite (e.g. Talwar et al., 2002). Except for the case in which a parent offers false praise, the scenarios did not involve lies that are likely to be motivated by politeness and some are consistent with the possibility that the parent's lie is at least partly motivated by self-interest (e.g. the parent who is embarrassed by a child's crying and warns that the police will come if the crying does not stop). Similarly, the examples of parental lying that were reported by participants went beyond contexts in which politeness or the child's best interests were at stake, such as those that were directed at preventing tantrums or excessive talking.

Why would parents lie to their children when they are so concerned with promoting honesty? One possibility is that parenting by lying is a response to the challenge of coordinating conflicting sets of goals. It may be that parents sometimes consider certain goals to be more important than the need to avoid telling lies. This would be consistent with evidence that adults sometimes lie in the context of non-parenting relationships as a means to accomplish desired ends (Lindskold & Walters, 1983; Camden et al., 1984; DePaulo et al., 1996; DePaulo & Kashy, 1998) and with evidence that individuals often balance different factors when making judgements about honesty (Perkins & Turiel, 2007). For example, a parent may believe that socialising children to behave appropriately outweighs the potentially negative consequences of making false statements. Parents may also lie to protect their children from information that they consider to be harmful, such as when a child hears a news story about a brutal crime and a parent falsely reinterprets the story to make it less threatening.

A related possibility is that motivational concerns sometimes override moral concerns (see Batson & Thompson, 2001). For example, it may be that parents intend to communicate with their children honestly and engage in parenting by lying only when they have difficulty generating other ways to achieve desired goals. As one mother reported, `I have said the police will get my son in trouble in some instances when the threat of mom is not intimidating'. Yet another possibility is that parents do not realise that when they lie to their children, they may be implicitly contradicting their more overt declarations that lying is unacceptable (see Harris and Gimenez [2005], Heyman and Compton [2006] and Rosengren and Hickling [1994] concerning contradictory beliefs in other domains). Finally, parents may assume that lying to a child is problematic only if the child finds out and limit their lying to cases in which they believe it will not be detected.

The present findings indicate that conceptions of lying may differ between Asian Americans and European Americans: although there were no significant group differences concerning lying to promote positive feelings, results from both studies indicate that Asian American parents were more likely to lie to influence behaviour. One possible explanation is that Asian American parents place a greater emphasis on respect and obedience (Lin & Fu, 1990; Chao, 1995) and are therefore more willing to lie to achieve these ends.

These results are also generally consistent with a growing body of evidence of cross-cultural differences in views about truth-telling and lying (Gao et al., 1996; Lee et al., 1997, 2001; Seiter et al., 2002; Fu et al., 2007; Heyman et al., 2007). This evidence suggests that concerns about social cohesiveness are more likely to trump concerns about veracity in East Asia than in North America (see Bond, 1986).

Limitations

There are several limitations in the way this research was conducted that should be considered when evaluating the results. First, the results are based exclusively on self-report data and many of the participants were reporting events that occurred many years earlier. Consequently, it is possible that participants misremembered or misinterpreted the events that they reported. It is also possible that responses were influenced by concerns about appearing socially desirable. However, the fact that the data were collected anonymously rather than in person is likely to have reduced these concerns. In addition, since it would probably be more socially desirable to deny having lied, any such distortions would probably lead to an underestimation of the acceptance of parental lying.

There are also limitations related to the selection of participants. Although the combination of an undergraduate sample in Study 1 and a parent sample in Study 2 helps to rule out the possibility that the major findings can be explained in terms of the perspective or bias of one particular group, it would have been ideal to have obtained data from both parents and children in the same parent–child dyad. This would have allowed for an examination of similarities and differences in perceptions within the same family and it also would have allowed us to control for a range of factors such SES. In addition, since the sample was predominantly female, we could not effectively assess potential gender differences.

Future research

Further research will be needed to examine the consequences of parenting by lying. One possibility is that parenting by lying encourages children to lie (see Stouthamer-Loeber, 1986). A second possibility is that parenting by lying promotes the development of scepticism about what others say (Heyman et al., 2007) because it increases children's awareness that `there is often a difference between the state of the world and words used to describe or predict it' (Brown, 2002, p. 252). It will also be important to examine whether parenting by lying has the intended emotional and behavioural consequences. For example, when parents tell children that bad things will happen if they do not eat their vegetables, does it influence their willingness to eat vegetables on that occasion and also at other times?

Another psychological consequence worthy of consideration is the development of trust within the parent–child relationship. A key theme from participants' open-ended responses was that lying undermines trust, as evidenced by a father who remarked, `my children know that if they ever lie to their mother or myself that it is almost the worst thing they could ever do because I will then question everything they say and I will no longer trust them'. Another parent reported that she had lied about the death of a pet to her young son to protect his feelings and that he had discovered the truth. She described being afraid that she would lose his trust and had decided that from that point forward all of her lies would be `half-truths'. An obvious question is whether parental lying ultimately undermines children's trust in their parents, just as deception and lying undermine trust in elementary school children's friendships (Kahn & Turiel, 1988). Parental lying may even have negative implications for trust before children enter elementary school, as is suggested by evidence that even preschool children tend to discount reports from individuals who have a history of providing inaccurate information (Koenig et al., 2004; Jaswal & Neely, 2006) and understand that lying can carry negative interpersonal implications (Siegal & Peterson, 1998). Consequently, it will be important for future research on parental lying to include measures of trust. In addressing this issue it will be important to examine different types of lies, since different types of lies may have different implications for trust.

It will also be important to investigate the beliefs and practices of different groups of people. For example, the sample of Asian American parents was composed primarily of parents who were not born in the USA. It will be important to examine how timing of arrival in the USA might affect the results and to look at differences among Asian American parents from different cultural backgrounds.

Further work is also needed on the conceptualisation of parental lying. Clearly, there can be a fine line between lying and other forms of communication, such as engaging in fantasy play and joking. Where such lines are drawn may depend not only upon the words that are used, but also upon the emotions that are expressed and the nature of the relationship. Such lines may also be perceived differently by parents and children. For example, a parent may tease a child by suggesting that something bad will happen, but the child may interpret the statement as straightforward prediction. This may be especially likely for young children who can have difficulty with interpreting subtle emotional cues.

Implications for moral education

Our findings show that parents lie to their children even though they maintain that lying is unacceptable. What are the implications of these findings for moral education? One possibility is that parents should only tell their children what they believe to be true. However, this solution is problematic because it implies that the goal of honest communication trumps all other social and moral values. Philosophers have long pointed to cases in which such a position is indefensible, such as when innocent lives are at stake and when only by telling lies can the danger be averted (Bok, 1978). Adults may also tell children things they do not believe in everyday interactions in which the positive consequences outweigh potential negative consequences. For example, adults may tell children things they do not believe in a fantasy play context in which there is a shared understanding that the parties involved have agreed to suspend disbelief (Bok, 1978).

A less problematic approach would be for parents to think carefully about the potential costs and benefits of truth-telling and lie-telling in different contexts. It is likely that when parents make knowingly false statements to their children, their focus is often on the immediate implications for the child's behaviour or emotions rather than the broader implications for the children's developing beliefs about interpersonal relationships. These broader considerations may not seem relevant to many parents, given that it may seem natural to lie to children for the purpose of providing appropriate care and protection. As Bok (1978) wrote, `To shield [children] not only from brutal speech and frightening news, but from apprehension and pain—to soften and embellish and disguise—is as natural as to shelter them from harsh weather' (p. 217). Additionally, it may at times seem appropriate to lie to children to compensate for their lack of knowledge and experience. Nevertheless, in meeting these goals, parents risk unintended consequences. For example, there is a real danger that parents will damage their credibility and send confusing messages about the importance of telling the truth. Consequently, parents should consider whether there are alternative ways of handling challenging parenting situations that do not pose such a dilemma.

Another implication of the present work is that parents might reconsider how they talk to their children about lying and deception. Although simple statements such as `lying is always wrong' seem to send a clear moral message that even young children can understand, they are unlikely to be effective in a context in which children observe lying by those close to them. We argue that a better approach would be to encourage children to think more deeply about the implications of different kinds of lies. For example, when a parent reads a story to a child in which a character faces a decision about lying, the child could be asked what the character should do and why. For example, the child might generate different responses that the character could make, and discuss the implications of different strategies for all of the characters who are involved. Holding such discussions separately from contexts in which the child is personally involved in lying is likely to encourage open discussion rather than defensiveness.

Because being able to reason about deceptive intent is a key part of learning to avoid being manipulated by others, discussions about lying could also be conducted with reference to the statements of others. For example, parents could discuss how and why advertisers present their products in the best possible light by emphasising positive qualities and de-emphasising negative qualities (Moses & Baldwin, 2005).

Our suggestions are in line with a wide body of evidence suggesting that parent-child discussions about deception are likely to have more positive effects than simple declarations that lying is wrong. For example, discourse about feelings and values is associated with the development of the child's conscience (Dunn, 2006). In addition, explanations, including those offered to children, and those that children generate themselves, tend to enhance cognitive development (Keil, 2006; Legare & Gelman, 2008). The present findings combined with a wide body of previous research suggest that children will benefit from extended discussions about lying and truth-telling.

Appendix 1. Scenario themes

Behaviour scenarios

Food. A child is told, `If you do not eat all the food, children in Africa will die.'

Police. A parent is embarrassed by a child's crying and says, `The police will come to make sure that you behave if you don't stop crying now.'

Money. A parent who wants to teach her child the value of money falsely claims that, `we are too poor to get what you want.'

Boogeyman. A child is told, `If you go outside alone, a bogeyman will get you.'

Monster: sleep. Susan knows that staying up late is bad for children, so she told the child, `if you do not go to sleep early, a monster will pull your feet.'

Monster: vegetables. A child is told, `If you don't eat vegetables, a monster will come and kidnap you.'

Emotion scenarios

Uncle. A favourite uncle has just died and the child is told that he has become a star to watch over the child.

Rainbow. A child who likes rainbows is told, `the rainbow came out just for you'.

Cleaning. A child is told, `you did a good job at cleaning up your room' after making things messier.

Appendix 2. Examples of parental lies

Lies participants in Study 1 reported being told by their parents

`You have a limited amount of air and if you keep talking it will run out.'

`When I was bad she drove me to the police station and said she would turn me in.'

`Eat your rice or it will become worms.'

`Mommy has no money but when you get A's mommy gets money.'

`If you play with fire, you'll wet the bed.'

`If you do not go to sleep early, a scary man will take you away from mommy and daddy.'

Lies participants in Study 2 reported telling their children

`When my son was not following directions in kindergarten I told him they would kick him out of school and he would have to go to a school where kids would beat him up.'

`We told our daughter that if she wrapped up all her pacifiers like gifts the `pacifairy' would come and give them to children who needed them…I thought it was healthier to get rid of the pacifiers and it was a way for her to feel proud and special.' `I told my kids that they need to finish all their food or they would have pimples all over their face.'