The Economic Journal,  (Noember), –. # Royal Economic Society . Published by BlackwellPublishers,  Cowley Road, Oxford OXJF, UK and  Main Street, Malden, MA , USA.THE FRAME OF REFERENCE AS A PUBLIC GOOD*Robert H. FrankDoes consuming more goods make people happier ? For a broad spectrum ofgoods, available evidence suggests that beyond some point the answer isessentially no. Much of this evidence is from the large and growing scientificliterature on the determinants of life-satisfaction and psychological well-being."Evidence from this literature also suggests, however, that there are ways ofspending time and money that do have the potential to increase people’ssatisfaction with their lives, and herein lies a message of considerableimportance for policy-makers.The psychologist’s conception of human well-being is somewhat differentfrom the economist’s. Economists speak of an individual’s utility, which intraditional economic models is assumed to be an increasing function of presentand future consumption of goods, leisure, and other amenities that peopletypically view as desirable. Faced with a limited income, the individual isassumed to choose among alternatives so as to maximise her utility. Theanalogous construct in the psychological literature is ‘subjective well-being ’, acomposite measure of life satisfaction, positive affect, and negative affect.Operational measures of subjective well-being take one of several forms. Byfar the most popular approach in the psychological literature has been simplyto ask people how happy or satisfied they are.#For example, people may beasked to respond, on a numerical scale, to a question like, ‘ All thingsconsidered, how satisfied are you with your life as a whole these days ?’ Or,‘Thinking of your life as a whole, would you consider yourself (a) very happy;(b) fairly happy; or (c) not happy.’ Another approach measures the frequencyand intensity of positive affect by asking people the extent to which they agreewith such statements as: ‘ When good things happen to me, it strongly affectsme.’More recently, neuroscientists have also used brainwave data to assesspositive and negative affect. Subjects with relatively greater electrical activityin the left prefrontal region of the brain are likely to indicate strong agreementwith statements like the ones above, while those with relatively greaterelectrical activity in the right prefrontal region are much more likely todisagree with these statements.$The left prefrontal region of the brain is richin receptors for the neurotransmitter dopamine, higher concentrations of whichbeen shown independently to be correlated with positive affect.%* I thank Jeremy Chua, Rajib Das, Nadja Marinova, Rupal Patel, Lisa Shenouda, and AndreaWasserman for their able research assistance."For an excellent and accessible survey of this literature, see Myers ().#See Easterlin ().$Davidson ().%Reported by Goleman ().[  ][. ]       020406080 100Mean satisfactionAnnual income (Thousands)Fig. . Income s. satisfaction in the United States, –. Source: Diener et al.().Satisfaction as identified by any of these measures is predictive of a varietyof observable behaviours that most of us take to be indicative of well-being. Forexample, people who call themselves happy, or who have relatively high levelsof electrical activity in the left prefrontal region, are more likely to be rated ashappy by friends; more likely to initiate social contacts with friends; morelikely to respond to requests for help; less likely to suffer from psychosomaticillnesses; less likely to be absent from work; less likely to be involved in disputesat work; less likely to die prematurely; less likely to attempt suicide ; less likelyto seek psychological counselling.&In short, it seems that what the psychologistscall subjective well-being is a real phenomenon. Empirical measures of it havehigh consistency, reliability, and validity.'In what follows, it is not my claimthat the only goal of a person or a society should be to achieve the highestpossible levels of subjective wellbeing. (Would you prefer to be Socratesdissatisfied or a pig satisfied?) For the purposes of this discussion, I need assumeonly that an increase in subjective well-being counts as a good thing if it isachieved without having to compromise other important values.My claim is that available evidence on the determinants of subjectivewell-being suggests a variety of ways this could be achieved. The basic idea issimple – namely, that, whereas across-the-board increases in many forms ofmaterial consumption goods have little discernible effect on subjective well-being in the long run, the same resources can be used in alternative ways thatdo give rise to lasting increases in subjective well-being.    -Richard Easterlin was the first to call economists’ attention to survey data thatilluminate the relationship between material living standards and subjectivewell-being.(Easterlin saw three significant patterns in the self-reportedsatisfaction data. First, he noted that satisfaction levels across individualswithin a given country vary directly with income – richer people, on theaverage, are more satisfied than their poorer countrymen. This relationship is&For surveys of this evidence see Frank (b, chapter ) and Clark and Oswald ().'Diener and Lucas ().(Easterlin ().# Royal Economic Society 01972Percent very happy20401975 1978 1981 1984 1987 1991Fig. . Percent very happy, United States, –. Source : National Opinion Research Center().illustrated in Fig. , which plots average satisfaction against annual income fora US sample of , persons surveyed between  and .Second, Easterlin noted that the average satisfaction levels within a givencountry tend to be highly stable over time, even in the face of significanteconomic growth. Fig. , for example, plots the percentage of Americanssurveyed who respond ‘very happy’ when asked,