Looking Out for Number One

Andreas Mogensen

Once we are made aware of the extent and severity of global poverty, it is difficult to avoid the conclusion that we ought to do something to help. Unfortunately, the gap between what we ought to do and what we actually put into action is famously wide. We have trouble doing what we ought even when we are the ones who stand to benefit: we know that we should get more exercise or save more for the future, and yet fall short of even these commitments.

In cases like the above, it is the lure of immediate gratification that diverts us from doing what we should. When it comes to donating money or time to the benefit of others, the perceived importance of self-interest can similarly swell beyond what is reasonable. If we are rightly convinced that we ought to help those in need, we can nonetheless be inspired to hold back by the worry that this will require us to make a substantive sacrifice, and even if we recognise this sacrifice as being scarcely comparable to the benefit that we can achieve for others.

Ethical reasoning is arguably ineffectual against this worry and its demotivational force. It can, however, be confronted in a more head-on fashion. In the following, I present empirical evidence to show that donating 10% of your income will most likely require no substantive sacrifice on your part, and may even be better from a purely self-interested perspective than spending your income in the common fashion. This, I hope, will help to close the gap between your actions and your knowledge of what should be done.

It is undeniable that giving away 10% of your income will reduce the amount of money that you have available to spend on yourself. However, money surely has no intrinsic value: it is something we accumulate not for its own sake, but for what we can get with it. Any sensible self-interested worry that fuels our reluctance to donate 10% of our income to the good of others must therefore be a worry that we will not be as happy as we might have been had we kept such money for ourselves. It is the merit of this worry that is at issue in the following.

Of course, different people have different relationships with money, and draw their happiness from different sources. Someone who gains the major part of her satisfaction from an expensive hobby such as skiing or gourmet cooking obviously needs greater income in order to be as happy as someone who is just as pleased playing football in the park and eating homemade pizza. I cannot tailor my approach to suit the peculiarities of every possible situation, and what is provided in the following is therefore limited in this respect (you will find it limited in certain other respects also). In order to determine exactly how your wellbeing depends upon a certain level of income, some personal judgment is called for. You should, however, be aware of your own fallibility in this regard. Psychologists have studied our accuracy in predicting the effect of various outcomes on happiness (these predictions are called affective forecasts in the jargon of positive psychology). They have found that we are not as good at this as we might expect.

We are generally good at predicting whether an outcome will be pleasant or unpleasant: we know full well that going on holiday is a good time and that going to the dentist is not. The problem lies elsewhere: in our estimates of quantity, rather than quality. Our affective forecasts are prey to an impact bias: we overestimate the duration and intensity of the pleasant or unpleasant feelings associated with outcomes (Wilson & Gilbert 2005).1 People routinely misjudge the negative impact of certain medical conditions on quality of life (QOL). One study found that patients without colostomies estimated the QOL of patients with colostomies to be 0.8 (on a scale from 0 to 1), whereas patients with colostomies rated their own life at 0.92 (Boyd et al. 1990);2 another found the general public estimating the QOL of those on chronic dialysis to be 0.39, whereas it was reported as 0.56 by actual dialysis (Sackett & Torrance 1978).3 The impact bias is not restricted to medical contexts: lovers overestimate the negative impact of the dissolution of their relationship; academics overestimate the negative impact of being denied tenure; and voters overestimate the negative impact of having their candidate lose the election (Gilbert et al. 1998).4 The impact bias also colours people’s perceptions of the relationship between income and happiness. Kahneman et al. (2006)5 asked a sample of working women in the United States to estimate the proportion of time typically spent in a bad mood by someone with low income (less than $20,000): they found the predicted prevalence of bad mood to be “grossly exaggerated.” Aknin, Norton, & Dunn (2009)6 found that a representative sample of Americans “vastly underestimated the happiness of people earning lower levels of household income (US$55,000 and below).” Keep in mind, then, that giving away 10% of your income will most likely not reduce your happiness by nearly as much as you think (assuming you think it will do this at all).

Reference:

1.	Wilson, Timothy & Gilbert, Daniel (2005). “Affective Forecasting: Knowing What to Want”, Current Directions in Psychological Science 14, 131-134.

Reference:

2.	Boyd, Norman, Heather Sutherland, Karen Heasman, David Tritchler, & Bernard Cummings (1990). “Whose Utilities for Decision Analysis?”, Medical Decision Making 10, 58-67.

Reference:

3.	Sackett, Daniel & Torrance, Geoerge (1978). “The utility of different health states as perceived by the general public”, Journal of Chronic Diseases 31, 697-704.

Reference:

4.	Gilbert, Daniel, Elizabeth Pinel, Timothy Wilson, Stephen Blumberg, & Thalia Wheatley. (1998). “Immune neglect: A source of durability bias in affective forecasting”, Journal of Personality and Social Psychology 75, 617–638.

Reference:

5.	Kahneman, Daniel, Alan Krueger, David Schkade, Norbert Schwarz, & Arthur Stone (2006). “Would You Be Happier If You Were Richer? A Focusing Illusion”, Science 312, 1908-1910.

Reference:

6.	Aknin, Lara, Michael Norton, & Elizabeth Dunn (2009). “From wealth to well-being? Money matters, but less than people think”, The Journal of Positive Psychology 4, 523-527. p. 54

Two sources of the impact bias have been identified (Wilson & Gilbert 2005):7 a focusing illusion and a failure to take psychological adaptation into account. If someone non-disabled is asked to consider what life would be like with a disability, their mind is immediately drawn to those aspects of life that would be negatively impacted by this condition. The particular salience of these considerations gives them a disproportionate place in the picture they form of a life with disability: they forget to take due account of the many other things that contribute to personal wellbeing but which are unaffected by disability. This is the focusing illusion: we see the outcome that we are thinking about as mattering more to our happiness simply because we are focusing on it at the time. The second source of the impact bias is a failure to take psychological adaptation into account. People typically adjust to both bad and good things. This much is common sense: we know that a person who eats out at a fancy restaurant every day eventually derives diminishing pleasure from their dining experiences, as gourmet food becomes the standard against which future gustatory experiences are measured. Nonetheless, we often fail to take such psychological mechanisms into account in our affective forecasts (Gilbert et al. 1998).8 Although we do not in fact adjust to absolutely everything (see, e.g., Weinstein 1982 on traffic noise)9, there is much evidence that we adapt readily to our level of income. Rainwater (1990)10 found that in the United States over the period 1950 – 1986 perceptions of the minimum income required simply to get by rose in proportion to increases in real income. In a study of Swiss households, Stutzer (2003)11 found that a 10% higher household income was associated with a 4.5% increase in the figure cited in response to the question, “What household income per month would you consider an absolute minimum in order to make ends meet and without running into debt even if you reduce your needs to a minimum?”

Reference:

7.	Wilson, Timothy, Thalia Wheatley, Jonathan Myers, Daniel Gilbert & Danny Axsom (2000). “Focalism: A Source of Durability Bias in Affective Forecasting”, Journal of Personality and Social Psychology 78, 821-836.

Reference:

8.	Gilbert, Daniel, Elizabeth Pinel, Timothy Wilson, Stephen Blumberg, & Thalia Wheatley. (1998). “Immune neglect: A source of durability bias in affective forecasting”, Journal of Personality and Social Psychology 75, 617–638.

Reference:

9.	Weinstein, Neil (1982). “Community noise problems: evidence against adaptation”, Journal of Environmental Psychology 2, 87-97.

Reference:

10.	Rainwater, Lee (1990). “Poverty and Equivalence as Social Constructions”, paper presented at the seminar on Families and Levels of Living: Observations and Analysis, European Association for Population Studies, Barcelona, October 29-31 (Luxembourg Income Study Working Paper 551).

Reference:

11.	Stutzer, Alois (2003). “The role of income aspirations in individual happiness,” Journal of Economic Behavior and Organization 54, 89-109.

You can thus improve your accuracy in estimating the impact of a 10% reduction in personal income by keeping in mind the many sources of happiness that would be unaffected by this reduction, as well as your ability to adjust your desires and aspirations to match your financial circumstances. However, I also hope to have impressed upon you the worth of looking further into what empirical studies can tell us about happiness and its relation to income and spending. Given that we have established that giving 10% of your income will not be as bad as you first predicted, how bad (if at all) is it likely to be? We now proceed to this matter.

If you are sufficiently idealistic, you might believe that income has no significant impact upon wellbeing once certain basic needs are met: money does not buy happiness once one has adequate food, shelter, etc. You might therefore be perfectly happy to give large amounts of money to others who are in need.

There is evidence that might be thought to lend support to this idealistic contention. If one plots the percentage of persons within a country reporting themselves as being ‘quite happy’ or ‘very happy’ against national income per person, one finds a curve that is very steep at low levels of income, but which gradually flattens as one reaches developed, Western nations (Wilkinson & Pickett 2010: 9)12. There are also longitudinal studies of developed nations that suggest no correlation whatsoever between happiness and rising real income. Easterlin (1995)13 found that there has been no upward trend in happiness in the United States since the 1950s, despite a doubling of real GDP per capita over this period; he found a similar non-trend in happiness in Japan between 1958 and 1987, despite a five-fold increase in real income per capita in this period. From this you might be drawn to conclude that once a certain standard of living is reached, greater income does not yield significant gains in personal wellbeing.

Reference:

12.	Wilkinson, Richard & Pickett, Kate (2010). The Spirit Level. London: Penguin.

Reference:

13.	Easterlin, Richard (1995). “Will raising the incomes of all increase the happiness of all?”, Journal of Economic Behavior and Organization 27, 35-47.

There is evidence that casts significant doubt upon this view, however. Cross-sectional studies consistently show that at any given time, within any given nation, income is positively correlated with happiness (see Diener & Biswas-Diener 2002: 122-127 for overview and discussion).14 Thus, the World Value Survey II found the percentage of respondents in the UK reporting above neutral life-satisfaction to be 19% higher for high income groups compared to low income groups; in France the difference was 29%, in the Netherlands it was 6%, and the global average was 17% (World Value Survey Group 1994).15

Reference:

14.	Diener, Ed & Biswas-Diener, Robert (2002). “Will Money Increase Subjective Well-Being? A Literature Review and Guide to Needed Research”, Social Indicators Research 57, 119-169.

Reference:

15.	World Value Survey Group (1994). World Values Survey, 1981–1984 and 1990–

Taken together, these findings are highly surprising. How can we explain the cross-sectional correlation between income and happiness in a manner consistent with the failure of happiness to increase in Japan and the United states despite massive economic growth? The most natural answer is this: when it comes to income, it matters far more how much we have compared to others than how much we have considered in absolute terms. Thus, if you are richer than others, you will be happier, but if everyone gets richer whilst keeping their relative positions constant, no one grows significantly happier. Easterlin’s findings have been contested (e.g., by Haggerty & Veenhoven 2003),16 but the ‘natural answer’ is widely recognised as stating an important truth of happiness economics. The relation between income and happiness depends much less upon how well off we are in non-comparative terms, and far more upon how well off we are as compared to some relevant standard. More specifically, it depends upon the income of those around us and the income that we expect to enjoy (Easterlin 2001; Layard 2006: 41-54).17

Reference:

16.	Haggerty, Michael & Veenhoven, Ruut (2003). “Wealth and Happiness Revisited: Growing wealth of nations does go with greater happiness” Social Indicators Research 64, 1-27.

Reference:

17.	Easterlin, Richard (2001). “Income and Happiness: Towards a Unified Theory”, The Economic Journal 111, 465-484. Layard, Richard (2006). Happiness: Lessons From a New Science. London: Penguin.

The suggestion that your satisfaction with your own economic situation should depend upon how well others are doing is hardly counterintuitive: we are all familiar with the phenomenon of keeping up with the Joneses. You might nonetheless be surprised by just how much more relative income matters compared to absolute income. Ball & Chernova (2008)18 found that for the median individual, “the effect of a marginal change in relative income on happiness is several times larger than the effect of a marginal change in absolute income,” such that the median individual should be indifferent between a tripling of absolute income and a rise from the 50^th to the 70^th percentile (keeping absolute income constant).

Reference:

18.	Ball, Richard & Chernova, Kateryna (2008). “Absolute Income, Relative Income, and Happiness”, Social Indicators Research 88, 497- 529.

The place of personal income aspirations in mediating the relationship between income and happiness must also be taken into account. Consider another paradoxical finding reported by Easterlin (2001)19: despite cross-sectional studies showing those with greater income to be happier than those with less at any given time, studies of life-cycle patterns show that although income generally improves substantially up to retirement, after which it decreases, there is no corresponding upward trend in personal happiness correlated with this rise, nor a downward trend associated with retirement. To explain this finding, it seems that we must invoke the capacity of income aspirations to march in step with the changes in economic situation typically experienced over the course of a person’s life.

Reference:

19.	Easterlin, Richard (2001). “Income and Happiness: Towards a ' Unified Theory”, The Economic Journal 111, 465-484.

We can see, then, that money can buy happiness under certain conditions. What, if anything, can we conclude from this regarding the probable effects of giving away 10% of our income? We might choose to think of this as equivalent to simply exchanging our level of happiness for that attributed by a cross-sectional survey to someone within our country whose income is 10% lower than our own. Since we know that income is positively correlated with happiness in cross-sectional surveys, we should then expect that the effect of donating our income will be to decrease happiness.

However, we must be careful in inferring causation from correlation. The existence of a positive correlation between happiness and income might be partly due, for example, to the influence of happiness on income: people of a cheerful disposition might end up wealthier than others. In fact, an effect of this kind exists (Diener & Biswas-Diener 2002: 134-135).20 Similarly, the correlation between income and happiness might be partly due to any number of third variables that cause both higher income and higher happiness. And, in fact, the correlation between income and happiness is weakened if one controls for variables such as education and unemployment (Easterlin 2001: 468; see Diener & Biswas-Diener 2002: 128 for general discussion).21

Reference:

20.	Diener, Ed & Biswas-Diener, Robert (2002). “Will Money Increase Subjective Well-Being? A Literature Review and Guide to Needed Research”, Social Indicators Research 57, 119-169.

Reference:

21.	Easterlin, Richard (2001). “Income and Happiness: Towards a Unified Theory”, The Economic Journal 111, 465-484.

On top of this, it might be argued that the available data inflates the size of the correlation between income and happiness. In surveys of this correlation, wellbeing is almost always measured by asking subjects to provide global reports of life-satisfaction: for example, they may be asked, “All things considered, how satisfied are you with your life as a whole these days?” However, there exists another technique for measuring happiness, experiential sampling, where subjects are asked to report their instantaneous feelings of happiness or unhappiness at several points over an extended period of time. It has been argued that experiential sampling provides a superior method of measuring happiness, because it overcomes the biases and imperfections associated with the cognitive capacities required to provide an accurate estimate of the happiness associated with one’s life as a whole (Kahneman 1999; Stone, Shiffman, & deVries 1999).22 The choice of method is of importance, because the correlation between income and happiness is weakened by the replacement of such experiential measures for global self-report measures (Kahneman et al. 2006; Diener et al. 2010).23 If experiential sampling provides a more objective measure of wellbeing, then money buys less happiness than most studies indicate.

Reference:

22.

Diener, Ed & Biswas-Diener, Robert (2002). “Will Money Increase Subjective Well-Being? A Literature Review and Guide to Needed Research”, Social Indicators Research 57, 119-169. Stone, Arthur, Saul Shiffman, & Marten deVries (1999). “Ecological Momentary Assessment” in Daniel Kahneman, Ed Diener, & Norbert Schwarz, eds. Well-Being: The Foundations of Hedonic Psychology. New York, NY: Russell Sage, 1999, 26-39.

Reference:

23.

Kahneman, Daniel, Alan Krueger, David Schkade, Norbert Schwarz, & Arthur Stone (2006). “Would You Be Happier If You Were Richer? A Focusing Illusion”, Science 312, 1908-1910. Diener, Ed, Weiting Ng, James Harter, & Raksha Arora (2010). “Wealth and Happiness Across the World: Material Prosperity Predicts Life Evaluation, Whereas Psychosocial Prosperity Predicts Positive Feeling”, Journal of Personality and Social Psychology 99. 52-61.

If we think of the ‘true correlation’ between income and happiness as being that portion of the observed correlation that is actually due to the causation of happiness by income, therefore, we may conclude that the true correlation is smaller than the typically observed correlation. Since giving away 10% of one’s income involves only the loss of money and no sacrifice of one’s educational achievement or sunny disposition, it is the true correlation that should interest us when we are attempting to determine the likely effects of donating to charity. Using the observed correlation as a proxy, and keeping in mind the smaller size of the true correlation, we can attempt to gain a more exact grasp of the size of the expected decline in happiness. The data shows the correlation between income and happiness to be small, and significantly smaller than people estimate (Aknin, Norton, & Dunn 2009: see Figure 1).24 Based on General Social Survey Data for the US in 1994, Easterlin (2001)25 calculated the mean happiness of those earning between $30,000 – 39,999 as only 0.2 points greater than those earning $20,000 - 29,999, when happiness was scored from 0 to 4 (keep in mind that this is before controlling for education, unemployment, etc.). Based on Helliwell (2003),26 Layard (2006)27 calculates that “[a] fall in income by one third (holding national income constant) causes a fall in happiness of 2 points on the scale of happiness (from 10 to 100).” The modest size of the correlation ensures that the effects of income on happiness are dwarfed by other non-pecuniary factors. On one analysis of their data, Ball & Chernova (2008)28 calculate that for the median single individual, the happiness boost produced by marriage is matched only by a 767% increase in absolute income, or by an increase in relative income from the 50^th to the 88^th percentile; the boost associated with moving from a health rating of 3 to 4 (when health is scored from 1 to 5) is matched only by a 6,531% increase in absolute income, or by a move from the 50^th to the 100^th percentile in relative income (see also Table 1).

Reference:

24.	Aknin, Lara, Michael Norton, & Elizabeth Dunn (2009). “From wealth to well-being? Money matters, but less than people think”, The Journal of Positive Psychology 4, 523-527.

Reference:

25.	Easterlin, Richard (2001). “Income and Happiness: Towards a Unified Theory”, The Economic Journal 111, 465-484.

Reference:

26.	Helliwell, John (2003). “How’s life? Combining individual and national variables to explain subjective well-being”, Economic Modelling 20, 331-360.

Reference:

27.	Layard, Richard (2006). Happiness: Lessons From a New Science p. 65 London: Penguin.

Reference:

28.	Ball, Richard & Chernova, Kateryna (2008). “Absolute Income, Relative Income, and Happiness”, Social Indicators Research 88, 497- 529.

Supposing, then, that we adopt the approach suggested above and think of donating 10% of our income as equivalent to simply exchanging our level of happiness for that attributed by a cross-sectional survey to someone within our country whose income is 10% lower than our own (keeping other variables constant), we may conclude that this should produce a very small reduction in personal happiness.

However, the methodology relied upon here is open to challenge. Most importantly, we might ask why we should think of giving away 10% of our income as equivalent to earning 10% less. After all, that assumption isliterally false: we will not have a lower income, just one that is spent differently. You might have been tempted to go along with the assumption in the above because you thought it plausible that giving money to someone you will never meet or otherwise interact with is as good from the perspective of your self-interest as simply throwing that money away. A little reflection shows this to be quite unlikely.

Many people have thought that genuine altruism does not exist: every apparent act of charity or mercy, they claim, is secretly or subconsciously driven by ruthless self-interest. Upon closer examination, this view appears thoroughly mistaken (Blackburn 1998: 122-160).29 However, for it to have been endorsed by so many intelligent persons, it stands to reason that benefitting others must reliably benefit us: if altruistic acts typically involved a dead-weight loss for the altruist, no one would think it particularly plausible to suppose that these acts were secretly driven by the desire for personal gain. Some of the ways in which helping others helps us are known to common sense. If I scratch your back, you might scratch mine, and I might then be better off than if I had kept my hands to myself. Helping someone might also raise our esteem in the eyes of third-parties, making them more likely to help us at some future time. Less tangibly, benefitting others might produce a ‘warm glow’: we might feel good about doing good, or simply about the fact that some good is being done. The existence of such a ‘warm glow’-effect has been demonstrated using functional magnetic resonance imaging (fMRI): reward centres in the brain were activated when money was transferred to a local food bank from the accounts of participants in an experiment designed by Harbaugh, Mayr, & Burghart (2007),30 and even in cases where the transfers were mandatory and tax-like. Finally, helping others who are in need might serve to ease our distress at their situation, whether this be due to pangs of conscience or powerful sympathy.

Reference:

29.	Blackburn, Simon (1998). Ruling Passions. Oxford: Oxford University Press.

Reference:

30.	Harbaugh, William, Ulrich Mayr, & Daniel Burghart (2007) “Neural responses to taxation and voluntary giving reveal motives for charitable donations”, Science, 316, 1622-1625.

Given this, it seems implausible to equate giving money to charity with never having had that money in the first place: whereas money that we have never had might not bring us any benefit, money that we give to aid the worst off might well bring some good back to us. To determine the probable effects of donating our income we should examine empirical studies that address the correlation between different kinds of spending and happiness, rather than the correlation between income and happiness.

Unfortunately, relatively little research has explored the former issue, though it has gained increasing attention of late (see Dunn, Gilbert, & Wilson in press for an overview).31 Certain important results have already been established: for example, spending money on experiences, such as holiday trips, typically yields greater happiness than spending money on material goods (Van Boven & Gilovich 2003).32 There is also evidence to indicate that benefitting others might benefits us more than does benefitting ourselves.

Reference:

31.	Dunn, Elizabeth, Daniel Gilbert, & Timothy Wilson (in press). “If Money Doesn’t Make You Happy Then You Probably Aren’t Spending it Right”, Journal of Consumer Psychology.

Reference:

32.

Imagine the following scenario. You are a participant in a psychological experiment: you are given an envelope containing a small sum of money, which you are asked to spend within 24 hours. The experimenter can assign you to one of two conditions: she can require that you spend the money on yourself (paying a bill or buying yourself a treat) or she can require that you spend the money on others (buying a present for someone or donating the money to charity). Which condition do you suppose would bring the greatest happiness: spending the windfall on yourself or in an altruistic fashion?

If you are like the typical participant in an experiment of this kind, recently conducted by Dunn, Aknin, & Norton (2008),33 you believe that spending on yourself brings greater happiness. If you are of this opinion, your affective forecast is also most likely mistaken: the experimenters found that subjects in the prosocial spending condition reported greater happiness after spending their windfall than did those in the personal spending condition. This was not an isolated result. Dunn et al. also conducted a longitudinal study of 16 employees at a Boston-based company who received a profit-sharing bonus, finding that those who devoted more of their bonus to prosocial spending experienced greater happiness as a result of spending their windfall; a cross-sectional study of a representative sample of Americans also found greater prosocial spending correlated with significantly greater happiness, while personal spending turned out to be unrelated to happiness. Surprisingly, it seems that benefitting others might benefits us more than does benefitting ourselves.

Reference:

33.	Dunn, Elizabeth, Lara Aknin, & Michael Norton (2008). “Spending Money on Others Promotes Happiness”, Science 319, 1687-1688.

Of course, we must be cautious in extrapolating from these results to the conclusion that donating 10% of our income to charity will be better for us than would be keeping this money for ourselves. The sums involved in the experiment described above are very small by comparison. The experiment also involved windfall gains, such that subjects were not using money they had already planned to spend on themselves. Moreover, insofar as the effects of donating 10% of our income might be expected to resemble those achieved for the prosocial spenders in the experiment described above, it is important that this 10% be taken from spending that would otherwise have gone on ourselves: the experiment does not support the view that giving money to charity is better than not giving if the money donated would instead have been spent on presents for friends or family. Nonetheless, it does lend some support to the view that donating 10% of our income could be better for us than keeping this money for ourselves.

It must be conceded that we have not been able to arrive at any definite verdict regarding the expected effects of donating 10% of our income to charity. Nonetheless, we have uncovered strong evidence to support the view that this will involve no great sacrifice on our behalf, and certainly less sacrifice than we initially expected. If the money we donate is as good from the perspective of our self-interest as money that is straightforwardly foregone, we should expect only a minor decrease in our wellbeing. However, we have also seen evidence that choosing to spend our income on charitable giving can bring benefits to us, and that these benefits might be greater than we anticipate. The worry that helping others might hurt us greatly ought therefore to be robbed of its force. I hope this makes it easier for you to do what you know to be right.

Figure 1. Actual versus predicted happiness as a correlate of household income in the USA (Aknin, Norton, & Dunn 2009: 254)35

Reference:

35.	Aknin, Lara, Michael Norton, & Elizabeth Dunn (2009). “From wealth to well-being? Money matters, but less than people think”, The Journal of Positive Psychology 4, 523-527.

Figure 2. The relationship between (un)happiness (scored from 10 to 100) and various key factors (Layard 2006: 64, based on Helliwell 2003).