You are on page 1of 7

Spending Money on Others Promotes Happiness

Elizabeth W. Dunn, et al.

Science 319, 1687 (2008);
DOI: 10.1126/science.1150952

This copy is for your personal, non-commercial use only.

If you wish to distribute this article to others, you can order high-quality copies for your
colleagues, clients, or customers by clicking here.
Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.

The following resources related to this article are available online at (this infomation is current as of August 30, 2011 ):

Downloaded from on August 30, 2011

A correction has been published for this article at:
Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
Supporting Online Material can be found at:
A list of selected additional articles on the Science Web sites related to this article can be
found at:
This article has been cited by 21 article(s) on the ISI Web of Science
This article has been cited by 19 articles hosted by HighWire Press; see:
This article appears in the following subject collections:

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2008 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.
finely regulated by correlated activities in a 5. C. D. Kopec, B. Li, W. Wei, J. Boehm, R. Malinow, 23. G. Aicardi et al., Proc. Natl. Acad. Sci. U.S.A. 101, 15788
neuronal network and may consolidate nearby J. Neurosci. 26, 2000 (2006). (2004).
6. Y. Dan, M. M. Poo, Physiol. Rev. 86, 1033 (2006). 24. R. Yuste, W. Denk, Nature 375, 682 (1995).
stimulated synapses by autocrine or paracrine 7. M. Matsuzaki et al., Nat. Neurosci. 4, 1086 (2001). 25. T. Nevian, B. Sakmann, J. Neurosci. 26, 11001
mechanisms. BDNF action was selective on the 8. Materials and methods are available as supporting (2006).
spines that showed immediate enlargement, material on Science Online. 26. F. Santamaria, S. Wils, S. E. De, G. J. Augustine, Neuron
which may act as the structural tag for selective 9. J. Spacek, K. M. Harris, J. Neurosci. 17, 190 (1997). 52, 635 (2006).
10. O. Steward, W. B. Levy, J. Neurosci. 2, 284 (1982). 27. L. E. Ostroff, J. C. Fiala, B. Allwardt, K. M. Harris, Neuron
trapping (26) of the protein-synthetic machinery 11. C. Job, J. Eberwine, Nat. Rev. Neurosci. 2, 889 (2001). 35, 535 (2002).
(11, 27) and the capture of plasticity proteins 12. S. L. Patterson et al., Neuron 16, 1137 (1996). 28. U. Frey, R. G. Morris, Nature 385, 533 (1997).
(2, 28) for long-term spine-head enlargement. 13. L. Minichiello et al., Neuron 24, 401 (1999). 29. This work was supported by grants-in-aid from MEXT of
Thus, BDNF acts as an associative messenger for 14. M. M. Poo, Nat. Rev. Neurosci. 2, 24 (2001). Japan (H.K. and M.M.), the Global COE Program
15. A. Figurov, L. D. Pozzo-Miller, P. Olafsson, T. Wang, B. Lu, (Integrative Life Science) of MEXT (H.K.), and NIH
the consolidation of synaptic plasticity, and the Nature 381, 706 (1996). (G.C.R.E.-D. and H.K.).
protein-synthetic process can regulate dendritic 16. H. Kang, A. A. Welcher, D. Shelton, E. M. Schuman,
structures at the level of single spines. Neuron 19, 653 (1997). Supporting Online Material
17. S. L. Patterson et al., Neuron 32, 123 (2001).
References and Notes 18. Y. Mu, M. M. Poo, Neuron 50, 115 (2006). Materials and Methods
1. L. R. Squire, Memory and Brain (Oxford Univ. Press, New 19. R. Blum, K. W. Kafitz, A. Konnerth, Nature 419, 687 Text
York, 1987). (2002). Figs. S1 to S7
2. R. J. Kelleher III, A. Govindarajan, S. Tonegawa, Neuron 20. H. S. Li, X. Z. Xu, C. Montell, Neuron 24, 261 References
44, 59 (2004). (1999).
3. M. Matsuzaki, N. Honkura, G. C. Ellis-Davies, H. Kasai, 21. J. Noguchi, M. Matsuzaki, G. C. R. Ellis-Davies, H. Kasai, 12 November 2007; accepted 21 January 2008

Downloaded from on August 30, 2011

Nature 429, 761 (2004). Neuron 46, 609 (2005). Published online 28 February 2008;
4. K. Okamoto, T. Nagai, A. Miyawaki, Y. Hayashi, Nat. 22. A. Gartner, V. Staiger, Proc. Natl. Acad. Sci. U.S.A. 99, 10.1126/science.1152864
Neurosci. 7, 1104 (2004). 6386 (2002). Include this information when citing this paper.

summed to create an index of prosocial spending

Spending Money on Others (M = $145.96, SD = 306.06). Entering the per-
sonal and prosocial spending indices simultaneous-
Promotes Happiness ly into a regression predicting general happiness
revealed that personal spending was unrelated to
happiness (standardized regression coefficient b =
Elizabeth W. Dunn,1* Lara B. Aknin,1 Michael I. Norton2
0.02, NS), but higher prosocial spending was
associated with significantly greater happiness
Although much research has examined the effect of income on happiness, we suggest that
(b = 0.11, P < 0.01). When we included income
how people spend their money may be at least as important as how much money they earn.
in this regression, we found that the effects of
Specifically, we hypothesized that spending money on other people may have a more positive
income (b = 0.11, P < 0.01) and prosocial
impact on happiness than spending money on oneself. Providing converging evidence for this
spending (b = 0.10, P < 0.03) were independent
hypothesis, we found that spending more of ones income on others predicted greater
and similar in magnitude, whereas personal
happiness both cross-sectionally (in a nationally representative survey study) and longitudinally
spending remained unrelated to happiness (b =
(in a field study of windfall spending). Finally, participants who were randomly assigned to
0.04, NS). Although the correlational nature
spend money on others experienced greater happiness than those assigned to spend money
of this design precludes causal inferences, this
on themselves.
study provides initial evidence that how people
spend their money may be as important for their
an money buy happiness? A large body choices that thinking about money promotes; the happiness as how much money they earnand

C of cross-sectional survey research has

demonstrated that income has a reliable,
but surprisingly weak, effect on happiness within
mere thought of having money makes people less
likely to help acquaintances, to donate to charity,
or to choose to spend time with others (7), pre-
that spending money on others might represent a
more effective route to happiness than spending
money on oneself (13).
nations (13), particularly once basic needs are cisely the kinds of behaviors that are strongly If this interpretation is correct, then people
met (4). Indeed, although real incomes have associated with happiness (812). At the same who receive an economic windfall should experi-
surged dramatically in recent decades, happiness time, although thinking about money may drive ence greater happiness after receiving the windfall
levels have remained largely flat within devel- people away from prosocial behavior, money can if they spend it on others rather than themselves,
oped countries across time (5). One of the most also provide a powerful vehicle for accomplish- even controlling for happiness before the windfall.
intriguing explanations for this counterintuitive ing such prosocial goals. We suggest that using We tested this prediction by examining the hap-
finding is that people often pour their increased money in this fashioninvesting income in others piness of 16 employees before and after they
wealth into pursuits that provide little in the way rather than oneselfmay have measurable bene- received a profit-sharing bonus from their compa-
of lasting happiness, such as purchasing costly fits for one's own happiness. ny (13). One month before receiving this bonus
consumer goods (6). An emerging challenge, then, As an initial test of the relation between spend- (M = $4918.64, SD = 1816.98), the employees
is to identify whether and how disposable income ing choices and happiness, we asked a nationally reported their general happiness as well as their
might be used to increase happiness. representative sample of 632 Americans (55% annual income. Approximately 6 to 8 weeks after
Ironically, the potential for money to increase female) to rate their general happiness, to report receiving the bonus, participants again reported
happiness may be subverted by the kinds of their annual income, and to estimate how much their general happiness and then reported what
they spent in a typical month on (i) bills and percentage of their bonus they had spent on (i)
Department of Psychology, 2136 West Mall, the University of expenses, (ii) gifts for themselves, (iii) gifts for bills and expenses, (ii) rent or mortgage, (iii) buy-
British Columbia, Vancouver, BC V6T 1Z4, Canada. 2Market-
ing Unit, 189 Morgan Hall, Harvard Business School, Soldiers others, and (iv) donations to charity (13). The ing something for themselves, (iv) buying some-
Field Road, Boston, MA 02163, USA. first two categories were summed to create an thing for someone else, (v) donating to charity,
*To whom correspondence should be addressed. E-mail: index of personal spending [mean (M) = $1713.91, and (vi) other. The first three categories were SD = 1895.65], and the latter two categories were summed to create an index of personal spending SCIENCE VOL 319 21 MARCH 2008 1687

(M = 63.44, SD = 38.20), and the fourth and fifth spending condition (M = 0.19, SD = 0.66). ipants were doubly wrong about the impact of
categories were summed to create an index of Neither the main effect of windfall size (F1,41 = money on happiness; we found that a significant
prosocial spending (M = 12.19, SD = 18.35). 0.09, NS) nor the Windfall Size Spending majority thought that personal spending (n = 69)
Entering Time 1 happiness and our two spend- Direction interaction (F1,41 = 0.12, NS) approached would make them happier than prosocial
ing indices into a regression predicting Time 2 significance. These experimental results provide spending (n = 40) (P < 0.01) and that $20 (n =
happiness revealed that prosocial spending was direct support for our causal argument that spend- 94) would make them happier than $5 (n = 15)
the only significant predictor of happiness at ing money on others promotes happiness more (P < 0.0005). Given that people appear to over-
Time 2 (b = 0.81, P < 0.02). With income in- than spending money on oneself. look the benefits of prosocial spending, policy
cluded as an additional predictor in this regression In moving away from the traditional focus on interventions that promote prosocial spending
(b = 0.03, NS), the effect of prosocial spending income toward an examination of spending encouraging people to invest income in others
remained significant (b = 0.96, P < 0.02). Simi- choices, our perspective dovetails with recent rather than in themselvesmay be worthwhile
larly, the prosocial spending effect was significant theorizing by Lyubomirsky, Sheldon, and Schkade in the service of translating increased national
(b = 0.81, P < 0.03) when controlling for bonus (8) on the architecture of sustainable changes in wealth into increased national happiness.
amount (b = 0.00, NS). Thus, employees who happiness. According to Lyubomirsky et al. (8),
devoted more of their bonus to prosocial spending the historical focus on life circumstances (e.g., References and Notes
experienced greater happiness after receiving the income, gender, and religious affiliation) as pre- 1. E. Diener, R. Biswas-Diener, Soc. Indic. Res. 57, 119
bonus, and the manner in which they spent that dictors of happiness may be somewhat mis- (2002).
bonus was a more important predictor of their placed; because people readily adapt to the stable 2. B. S. Frey, A. Stutzer, Econ. J. 110, 918 (2000).
3. D. Kahneman, A. B. Krueger, D. Schkade, N. Schwarz,

Downloaded from on August 30, 2011

happiness than the size of the bonus itself (13). circumstances of their lives, circumstantial factors A. A. Stone, Science 312, 1908 (2006).
Building on our correlational and longitudinal tend to have rather limited long-term effects on 4. E. Diener, E. Sandvik, L. Seidlitz, M. Diener, Soc. Indic.
evidence that spending on others may promote happiness levels. Thus, intentional activities Res. 28, 195 (1993).
happiness, we next demonstrated the causal im- practices in which people actively and effortfully 5. R. A. Easterlin, J. Econ. Behav. Organ. 27, 35 (1995).
6. R. H. Frank, Daedalus 133, 69 (2004).
pact of prosocial spending, using experimental choose to engagemay represent a more prom- 7. K. D. Vohs, N. L. Mead, M. R. Goode, Science 314, 1154
methodology (13). Participants (N = 46) rated ising route to lasting happiness. Supporting this (2006).
their happiness in the morning and then were premise, our work demonstrates that how people 8. S. Lyubomirsky, K. M. Sheldon, D. Schkade, Rev. Gen. Psychol.
given an envelope that contained either $5 or choose to spend their money is at least as im- 9, 111 (2005).
9. E. Diener, M. E. P. Seligman, Psychol. Sci. 13, 81 (2002).
$20, which they were asked to spend by 5:00 portant as how much money they make. 10. P. A. Thoits, L. N. Hewitt, J. Health Soc. Behav. 42, 115
p.m. that day. Participants randomly assigned to Finally, despite the observable benefits of (2001).
the personal spending condition were instructed prosocial spending, our participants spent rela- 11. J. A. Piliavin, in Flourishing: Positive Psychology and
to spend the money on a bill, an expense, or a gift tively little of their income on prosocial ends; the Life Well-Lived, C. L. M. Keyes, J. Haidt, Eds.
(APA, Washington, DC, 2003), pp. 227247.
for themselves, whereas participants assigned to participants in our national survey, for example,
12. T. Kasser, K. M. Sheldon, J. Happiness Stud. 3, 313 (2002).
the prosocial spending condition were instructed reported devoting more than 10 times as much 13. Materials and methods are available as supporting
to spend the money on a gift for someone else or money for personal as for prosocial spending material on Science Online.
charitable donation. Participants were called after each month. Although personal spending is of 14. This work was funded by a Hampton Research Grant to
5:00 p.m. that day and again reported their hap- necessity likely to exceed prosocial spending for the first author. We thank T. Rogers, S. Shababi and
J. Beretta for data collection; J. Biesanz, I. Dar Nimrod,
piness. We submitted postwindfall happiness to a most North Americans, our findings suggest that D. Gilbert, S. Heine, J. Helliwell, S. Lyubomirksy,
2 (windfall size: $5 versus $20) 2 (spending very minor alterations in spending allocations A. McConnell, V. Savalei, B. Simpson, and T. Wilson for
direction: personal versus prosocial) between- as little as $5 in our final studymay be suf- helpful input; and J. Goldshine for his invaluable assistance.
subjects analysis of covariance (ANCOVA), with ficient to produce nontrivial gains in happiness
prewindfall happiness included as a covariate. on a given day. Why, then, dont people make Supporting Online Material
This analysis revealed a significant main effect these small changes? When we provided descrip-
Materials and Methods
of spending instructions [F1,41 = 4.39, P < 0.04, tions of the four experimental conditions from Figs. S1 to S6
effect size estimate (p2) = 0.10], whereby partic- our final study to a new set of students at the Tables S1 and S2
ipants in the prosocial spending condition (M = same university (N = 109) and asked them to References
0.18, SD = 0.62) reported greater postwindfall select the condition that would make them hap- 25 September 2007; accepted 12 February 2008
happiness than did participants in the personal piest, Fishers Exact Tests revealed that partic- 10.1126/science.1150952

1688 21 MARCH 2008 VOL 319 SCIENCE


ERRATUM Post date 29 May 2009

Reports: Spending money on others promotes happiness by E. W. Dunn et al. (21 March
2008, p. 1687). We recently identified a bug in the statistical package SPSS 14.0 that we
used to analyze our data. The bug causes the first line of data to overwrite the sixth line of
data (see This bug affected the data set for
our longitudinal study (N = 16). Primarily because this error decreased the variance in our
data set, correcting this error alters the critical statistical test such that the effect of prosocial

Downloaded from on August 30, 2011

spending on happiness is reduced from the originally reported effect of b = 0.81, P < 0.02 to
b = 0.55, P < 0.115 (which falls within the 95% confidence interval for the original parame-
ter estimate). The corrected data set is available from the first authors Web site. The other
three studies reported in our paper were unaffected by this bug. SCIENCE ERRATUM POST DATE 29 MAY 2009 1

Race and genomics Telescope astronomy

1144 1149

LETTERS are interested in learning more about agricul-

tural and environmental issues. Job pros-
pects also are good; the Bureau of Labor
edited by Jennifer Sills
Statistics predicts that employment for agri-
cultural and food scientists will be at least
average overall and much higher than aver-
Agriculture Is Not a Dirty Word age in some specialties.
In the long run, does it really matter

Downloaded from on August 30, 2011

ture sciences have been disparaged in the science and education communities, perhaps because call the people who ensure a safe and plentiful
agronomy, soil science, plant pathology, and animal science use a problem-solving approach food supply, clean water, and healthy soil?
rather than simply seeking knowledge. Maybe not, as long as this critical work is
When science research funds are handed outfor example, in the federal stimulus bill funded and accomplished. But as we move
agriculture often gets left off the list. I suspect this is because policy-makers and some scientists into a new era of shared accountability and
see agriculture as synonymous with agribusiness, rather than as a purely scientific disci- responsibility, lets keep in mind that agricul-
pline, and they assume private funding will take care of agriculture-related research tural sciences affect us all, and when agricul-
needs. Agricultural scientists at land-grant institutions do receive some research tural science is thriving, our communities
dollars from noncompetitive sources, but not all research is funded this way. likely are thriving, too. ALLEN S. LEVINE
Adding insult to injury, the major U.S. science journals dont devote spe- College of Food, Agricultural, and Natural Resource
cific sections or editors to agricultural research. Some schools of agricul- Sciences, University of Minnesota, St. Paul, MN 55108
ture have taken the word agriculture out of their names, presumably to 6074, USA, and Minnesota Obesity Center, Minneapolis VA
Medical Center, Minneapolis, MN 55417, USA. E-mail:
attract more students in a country where only 2% of the population farms.
(It hasnt worked: Enrollment in university agricultural science majors
has dropped steadily nationwide since the early 1980s.)
In short, agricultural science has an image problem. Our disciplines 1. J. Vidal, Stem rust fungus threatens global
are not considered relevant and, more disturbing, were not seen as a wheat harvest, Guardian (19 March 2009);
source of solutions to many of the worlds most pressing challenges, even environment/2009/mar/19/
though many of those challenges directly relate to agricultural science.
Thats unfortunate, particularly in a world where people are starving
or eating unsafe food, where climate change will affect every aspect of Schools Anxiously Await
21st-century life, and where new kinds of sustainable fuel are needed.
The urgency of these global issuesall of them related to the agricultural NRC Program Rankings
sciencesamplifies the need for an applied-science approach. FACULTY MEMBERS AT U.S. INSTITUTIONS
Agricultural scientists can do amazing things when they combine their usually have mixed feelings about U.S. News
expertise and have access to the resources they need. Recently, scientists at an and World Reports graduate school rank-
international conference in Mexico announced that they have found a wheat vari- ings. On one hand, we often criticize its one
ety that is resistant to Ug99a strain of stem rust that could affect up to 90% of the size fits all approach and the importance
worlds wheat (1). Although the scientists have not completely eliminated the threat, its assigned to different categories. On the
clearly a breakthrough with enormous implications. other hand, we are keenly aware that the
Other recent signs also point to a renewed interest in and respect for agriculture. When the ranking of our school can affect our ability
first lady plants a vegetable garden on the White House lawn for the first time in half a century, to recruit top graduate students and even

shes sending a strong message: Food is important. Books about eating a sustainable, healthy faculty. It was thus exciting to hear, in mid-
diet top our best-seller lists. The National Gardening Association expects a 19% jump in the 2006, that the National Research Council
number of people growing at least some of their own food this year. Clearly, a growing number (NRC) decided to perform its own evalua-
of Americans are interested in where their food comes from, even on a small scale. tion of graduate programs, the first in more
The 2008 Farm Bill creates the National Institute for Food and Agriculture, which will be than a decade (1).
headed by a distinguished scientist directly appointed by the president. A small thing, perhaps, In January 2007, I and thousands of my col-
but it elevates agriculture to a level of prominence along the lines of health and other sciences. leagues completed a lengthy questionnaire
The farm bill also increases funding for competitive grants in both basic and applied agricul- sent by the NRC. Given that U.S. News pub-
tural research, which will provide opportunities for advanced study. lishes these rankings on an annual basis, the
Enrollment is up 16% since 2005 among college students in the professional associations anticipated NRC release date of December
that specialize in soil and crop sciences and agronomy, which suggests that todays students 2007 seemed achievable.

1140 29 MAY 2009 VOL 324 SCIENCE

Published by AAAS
Elucidating microbial Symbiosis and
consortia calcification in corals

1150 1153

Unfortunately, it was not. The release has made by NRC staff, data were returned to the
been postponed again and again, for more surveyed institutions several times for verifi-
than a year. Currently, they do not specify cation and correction (with changes recorded
even a tentative release date. Time is an for over 400 programs), and queries to inde-
issue, because the rankings are based on pendent sources were checked repeatedly. It
attributes that may change over a period of a is of the greatest importance that the differ-
few years, including funding and publica- ences that we report between different pro-
tions. The longer the delay between data col- grams be real and not based on either error

Downloaded from on August 30, 2011

lection and release of results, the less rele- or misunderstandings. Although the results
vant they become. will never be free of error, we believe that we
The students applying to graduate schools are now at the end of this process.
are most affected by this delay. Unfortu- The section of the report in which we
nately, like fellow students from past years, outline our methodology has been submit-
they will not be able to use the new rankings ted to the rigorous review process of the
to choose a school. NRC and should soon be available to inter-
ZIV BAR-JOSEPH ested readers. The release of the full report,
School of Computer Science, Carnegie Mellon University, with the complete set of data for university
Pittsburgh, PA 15213, USA. E-mail: graduate programs, should follow not too
long thereafter and will be made available to
1. Research Doctorate Programs, Board on Higher faculty and students well before the begin-
Education and Workforce, Assessment of Research ning of the next academic year. We fully
Doctorate Programs ( concur that the reported data and analysis
should be as current as possible and hope
that, working together with the participating
Response universities, a mechanism can be developed
WE ARE IN TOTAL AGREEMENT WITH BAR- to provide updates to the report on a timely
Joseph in our concern over the length of time and prompt basis.
that this study has consumed. JEREMIAH P. OSTRIKER1* AND CHARLOTTE KUH2
The principal product of the study will 1Department of Astrophysical Sciences, Princeton Uni-
be the release of a great deal of data cover- versity, Princeton, NJ 085441001, USA. 2National Acad-
ing graduate education in the United States, emy of Sciences, Washington, DC 20001, USA.
much of it gathered for the first time, which *To whom correspondence should be addressed. E-mail:
will be of use to prospective students, the
faculty in these programs, and the leader-
ship of institutions of higher education 1. J.P.O. is the committee chair for the study. C.K. is the
throughout the United States. The data were study director.
collected from 93,626 individual faculty
members, 222 institutional coordinators
providing information about 5009 pro- White Wolves Can
grams, and many independent private and
public sources of information. We are
Stand the Heat
extraordinarily grateful to the participating T. M. ANDERSON ET AL. REPORT THAT WOLVES
institutions for the care, time, and work that of the northwestern forest in North America
they have put into this project! are dark in comparison to the wolves of
Because data verification is of the utmost the tundra (Molecular and evolutionary
importance in a study of this nature, we have history of melanism in North American gray
taken extraordinary pains to check and to wolves, Reports, 6 March, p. 1339). Anderson
recheck the input information for accuracy et al. suggest in passing that the dark coats
and uniformity. An ever-relevant question might better hide preying wolves in the
has been Did all respondents to a given forest. An alternative adaptive explanation
query address the identical issue? External is that in the summer, a black tundra wolf
checks for consistency and accuracy were could overheat. SCIENCE VOL 324 29 MAY 2009

Published by AAAS

A study by Finch and Western provides color to the social system of the Kenyan
evidence of the temperature-related conse- herders. Some of them divide themselves
quences of coat color among pastoralists into clans according to the preferred color
cattle in Kenya (1). Finch and Western of their cattle, and the preference correlates
showed experimentally that black Boran adaptively with the usual heat stress of the
Zebu cattle in the sun became hotter than clans home ranges.
did white cattle and that across Kenya, as ALEXANDER H. HARCOURT
heat stress increased, the proportion of dark Department of Anthropology, Graduate Group in Ecology,
cattle in herds decreased. They also found University of California, Davis, CA 95616, USA. E-mail:
that herders report more dark than light cat-
tle dying during droughts at low altitude. Reference
Finch and Western related the variation in 1. V. A. Finch, D. Western, Ecology 58, 1384 (1977).


Perspectives: Tipping pointedly colder by L. R. Kump (27 February, p. 1175). In the figure caption, the cooling trajectory
was mistakenly referred to as red (it is blue) and the warming trajectory as blue (it is red). In the third paragraph, references

Downloaded from on August 30, 2011

4 and 5 were reversed (reference 4 is to Lear et al., 2008; reference 5 is to Katz et al.), and in the fifth paragraph, the ref-
erence to Lear et al. (5) should be to reference 3 (Lear et al., 2000). In the ninth paragraph, the citation to references (24)
should be to references (25).
Perspectives: Harvesting ocean wave energy by J. Scruggs and P. Jacob (27 February, p. 1176). For panel A of the fig-
ure, the photo credit should have read Copyright 2007 New York Times Graphics and the caption should have read
Schematic of a direct-drive point absorber developed at Oregon State University. The rated power of the Ocean Power
Technologies array in Oregon is 1.5 MW, rather than 2.0 MW, and the company name for the AWS point absorber is
Teamwork Technology BV.
Reports: Spending money on others promotes happiness by E. W. Dunn et al. (21 March 2008, p. 1687). We recently
identified a bug in the statistical package SPSS 14.0 that we used to analyze our data. The bug causes the first line of data
to overwrite the sixth line of data (see This bug affected the data set for our
longitudinal study (N = 16). Primarily because this error decreased the variance in our data set, correcting this error alters
the critical statistical test such that the effect of prosocial spending on happiness is reduced from the originally reported
effect of b = 0.81, P < 0.02 to b = 0.55, P < 0.115 (which falls within the 95% confidence interval for the original para-
meter estimate). The corrected data set is available from the first authors Web site. The other three studies reported in our
paper were unaffected by this bug.
Reports: Adaptive plasticity in female mate choice dampens sexual selection on male ornaments in the lark bunting by A. S.
Chaine and B. E. Lyon (25 January 2008, p. 459). In the lower right of Fig. 4, N = 384, not 87. The statistics reported are correct.


COMMENT ON A Large Excess in Apparent Solar Oblateness Due to Surface

J. R. Kuhn, M. Emilio, R. Bush
Fivian et al. (Reports, 24 October 2008, p. 560) analyzed data from the Reuven Ramaty High-Energy Solar
Spectroscopic Imager satellite and reported that the Sun is more oblate than previous measurements have sug-
gested. We argue that their threshold-based analysis yields a biased measure of the solar limb shape geometry.
Full text at

RESPONSE TO COMMENT ON A Large Excess in Apparent Solar Oblateness Due to

Surface Magnetism
M. D. Fivian, H. S. Hudson, R. P. Lin, H. J. Zahid
Our analysis of satellite data from the Reuven Ramaty High-Energy Spectroscopic Imager revealed a strong correla-
tion with a magnetic proxy, leading to an anomalous increase of the solar oblateness. We modeled this effect to yield
an oblateness measurement of an assumed underlying nonmagnetic Sun. Kuhn et al. note that this measurement
could be confused with a systematic spatial brightness variation. The existing data on this are ambiguous.
Full text at

Letters to the Editor

Letters (~300 words) discuss material published in Science in the previous 3 months or issues of general
interest. They can be submitted through the Web ( or by regular mail (1200 New
York Ave., NW, Washington, DC 20005, USA). Letters are not acknowledged upon receipt, nor are authors
generally consulted before publication. Whether published in full or in part, letters are subject to editing for
clarity and space. SCIENCE VOL 324 29 MAY 2009

Published by AAAS