While not denying the usefulness of different concepts like life satisfaction and subjective well-being, this chapter argues that happiness should be preferred in most cases, particularly with respect to what individuals and the society should really be interested in ultimately. Life satisfaction is more liable to a shift in the aspiration level, reducing the comparability of the resulting indices (e.g. Keller 2019). Life satisfaction and/or preference may also differ from happiness due to a concern for the happiness of others. [In the next chapter, a moral philosophical argument in favour of happiness as the only rational ultimate objective is given. All proposed qualifications to this principle can be explained by the effects on the happiness in the future or of others (hence really no qualification) or that their apparent acceptability is due to our imperfect rationality. In Chap. 6, simple ways to improve the accuracy and interpersonal and intertemporal comparability of happiness measurement include using happiness instead of life satisfaction (or other concepts), pinning down the dividing line of the zero amount of net happiness, using an interpersonally valid unit based on the just perceivable increment of happiness, and the complementary use of this method for small samples and the traditional methods for large samples.]

As defined above, our concept of happiness (or SWB or welfare) is subjective (‘hedonic’ in the philosophical sense), rather than attitudinal, as is the concept of life satisfaction. In other words, it is what one actually feels good (and minus the bad feelings to get ‘net happiness’), irrespective of what one regards. Some authors define happiness or SWB as inclusive of the attitudinal aspect.Footnote 1 In my view, it is less confusing to call such attitudinal concepts of ‘happiness’ as life satisfaction. Even with the appropriate terminology, there is still the question of which one is more appropriate or important.

We do not have to choose only one of the two and give up the other. Different concepts may serve different purposes and be appropriate or important in different issues. For example, a political party concerned with election victory may be more concerned with people's preference than their welfare, and may thus be more interested in their life satisfaction than their actual feeling of happiness. A statesman or philosopher concerned with people's true welfare may find happiness more relevant than life satisfaction. Also, since we should be concerned with long-term welfare than just the short-term ones, we should also recognize that life satisfaction now may affect happiness in the future. If we abstract from this consideration of effects on future values, or take both concepts (happiness vs. life satisfaction) as both the long-term or a-temporal ones, there are at least two important considerations that make happiness the intrinsically more important concept than life satisfaction. Happiness should be preferred in most cases and particularly with respect to what individuals and the society should really be interested in ultimately. This is related to normative valuation and different persons may have different views. It is difficult if not impossible to have full agreement here. Nevertheless, the views expressed here may be persuasive to some readers. In particular, the two problems of using life satisfaction are discussed in this chapter and the point that it is happiness that is of intrinsic value ultimately is argued in the next chapter.

Since happiness is the ultimate objective in life for most people (and argued to be the only thing of intrinsic value ultimately in the next chapter), life satisfaction is very closely related to happiness. This is supported by the fact that surveys give very similar results whether happiness or life satisfaction is used. However, life satisfaction may yet differ from happiness. Here, we are not concerned with the practical difficulties both from the researcher side and from the side of the subjects in measuring and in forming judgements regarding happiness and life satisfaction, especially the later. It is well-known that such judgments ‘are constructions drawn on the spot on the basis of currently available information and circumstances, and thus they are highly unstable and sensitive to changes in the context of inquiry’ (Alexandrova 2005, p. 303; also Stundziene 2019). As summarized by Schwarz and Strack (1999), such reports vary with the order of the questions asked, the time of inquiry, the mood of the subject, etc. This unreliability is probably the main reason why Kahneman (1999) prefers the use of ‘objective’ (a somewhat misleading term since happiness itself is subjective by nature; better understood as ‘objectively measured’) happiness, measured by the temporal integral of moment-based happiness reports. Here, especially for this and the next paragraphs, the practical problems of reporting and measurement inaccuracy are abstracted away. (For a meta-study of reliability, see Vassar 2008.) Also, Haybron (2007) argue convincingly that ‘our attitudes toward our lives can reflect various virtues and vices, such as gratitude, fortitude, ambition, pride, complacency, smugness, softness, low self-regard, etc.’ (p.107) and are rather arbitrarily affected by the norm and perspective taken. Even in the absence of these difficulties, happiness and life satisfaction in themselves may still differ.Footnote 2

For simplicity, consider a simply hypothetical example of 1,000 individuals. (Like Einstein’s thought experiments, such examples need not be realistic. In fact, deliberate exaggeration from reality is made to drive home the point.) All individuals believe that the only ultimately valuable thing is happiness. However, they are not self-centred and care also for the happiness of other individuals. Thus, they do not just pursue their own happiness but also try to do things that can increase the happiness of others. Evolutionary biology suggests that we are probably so programmed as our sociability is a trait that increases our fitness for survival and reproduction. In fact, even the gene that gives those who possess it a high in helping others has been found; see Bachner-Melman et al., 2005.Footnote 3 This, however, does not negate the importance of upbringing and social influences. On some insights on happiness issues from the evolutionary biological perspective, see Ng 2015. Then it is hypothetically possible for the following extreme case (exaggerated to emphasize the point) to happen.

Each individual sacrifices much time, effort, and happiness to do something believed to be good for the society. Due to ignorance, unlucky events, etc., their admirable effort does not pay off. They all end up really unhappy (negative affective feelings more than offset positive ones in aggregate) despite some positive feelings of doing something good for the society. If anyone of them is asked how happy they are, each will say fairly unhappy. However, if asked for life satisfaction, each may say reasonably satisfied, because each believe that what she has done for the society makes her life worthwhile. She is so much satisfied with doing something good for the society that this offsets her own unhappiness. This feeling itself is likely to increase her happiness, but not by enough to make the net happiness positive.

For example, suppose that A, one of these individuals, believes that her good work increases the happiness of each and every other individual by 10 (what unit happiness is measured in is irrelevant to the point being made here; the measurability and interpersonal comparability of happiness are discussed in Chap. 5), giving a total contribution of 9990 to others. This belief increases her net happiness from minus 100 to minus 30. Though she is still unhappy in her own subjective feelings, she thinks that her life is worthy as she has contributed 9990 to the happiness of others. If asked about life satisfaction, she may well say that she is satisfied, though she also says that her happiness is negative. If all the 1,000 individuals are in somewhat similar situations like A, we may get a high degree of life satisfaction and low happiness. Since happiness is really all these individuals ultimately value, the index of life satisfaction may well be misleading in such cases where the two diverge significantly from each other. In this example, the divergence is partly due to the existence of altruism. Other things being equal, the higher the degree of altruism, the larger is the potential divergence between happiness and life satisfaction.

There is another problem with the concept and measurement of life satisfaction. Though this problem also applies to those of happiness and subjective well-being, the extent of the problem is more serious for life satisfaction. Consider the finding that the average index of life satisfaction for a country such as the U.S. has remained largely unchanged over the last seventy years or so. Can we really be confident that happiness has not increased? Consider a popular method of obtaining the index of life satisfaction. A subject is asked to rate her own index of life satisfaction from the range 0–10, with 0 signifies the least satisfied life and 10 the most satisfied life, taking everything into account. The average index of a country may have remained at say 7. However, it is possible that people fifty years ago were more moderate in their aspiration not only in terms of objective things like income or consumption levels, but also more moderate in terms of subjective happiness. For simplicity, suppose we can use an interpersonal and intertemporal comparable unit of happiness (on which see Chap. 5). Suppose that an average person fifty years ago enjoyed a net happiness level of 700 units and rated herself a life satisfaction index of 7. Now, suppose that an average person enjoys a net happiness level of 1,400 units but still rates herself a life satisfaction index of 7, since her aspired level of happiness is much higher. If so, then an unchanged life satisfaction index may actually hide a doubling in net happiness level. (For some evidence of such a shifting standard and the discussion of related issues, see Hagerty 2003 and Diener and Lucas 2001.)

The above problem may also exist even if the concept of happiness or subjective well-being is used instead, at least for most methods of measurement used currently, including the 0–10 or 0–100 self-anchoring scale. Even if subjects are asked to tick either one of say: very happy, pretty happy, not too happy, and unhappy, the same problem exists. Thus, it may be the case that, people now typically report themselves as ‘pretty happy’ if their net happiness level is within say the range of 600 to 800 units, while people fifty years ago typically report themselves as ‘very happy’ for the same range. However, it is likely that, using the concept of life satisfaction makes this problem of changing subjective aspiration more pronounced. This is so because ‘satisfaction’ is more a concept of relative gratification in relation to the aspiration level. Happiness and subjective well-being are less so, though not completely.

Let me illustrate the point by reporting on the actual situation of a person I know best, myself. For simplicity and to isolate the current issue from the previous issue of the effect of contribution to the happiness of others in affecting one’s life satisfaction, let us abstract away any effect on others. If asked to rate my happiness and life satisfaction levels now within the scale of 0–10, I will probably rate both as 9 and tick the box ‘very happy’. If I am also asked now to rate my happiness and life satisfaction levels four decades ago, I will give 6 to happiness level but 8 to life satisfaction. Since I am the same person who experienced my happiness and life satisfaction both now and four decades ago, subject to some imperfection in recollection, I can compare these levels cardinally.Footnote 4 (Despite my age, I still have a good memory; I can still recite many poems, some of many hundred words each.) Thus, I can confidently say that my (net) happiness level now is at least four times that of four decades ago. I may well be inaccurate in my memory but this does not affect the argument here. A change in the correct multiple to 3 or 6 does not change the point to be made. Taking the mid point 5 to be a level of zero net happiness, putting my net happiness level four decades ago as 6 and the current level as 9 provides a roughly correct reflection of the fourfold difference. [ (9–5) = 4 (6–5).] While I am also more satisfied with life now than four decades ago, the increase is certainly much less than doubled, not to mention a three or fourfold increase. This difference between the changes in happiness and in life satisfaction is mainly because four decades ago I was also fairly satisfied; not having experienced a much higher level of happiness, I was fairly satisfied with a net happiness level I now describe as 6. The value of 8 is a good description of my level of life satisfaction then compared to the value of 9 now. But this small increase from 8 to 9 in life satisfaction hides the actual larger than three or fourfold increase in happiness.

Though I now put my net happiness level four decades ago as 6 and my level of life satisfaction then as 8, if I were asked four decades ago for reports on the situation at that time, I would probably have reported 8 for both happiness and life satisfaction. I now describe that lower happiness level as 6 only in comparison to my current much higher happiness level. If we normalize the amount of my (net) happiness four decades ago as 100 (in some subjective unit, not out of 10), my happiness amount now is 400. Suppose my happiness level were to decrease back to 100 five years from now in 2025. If someone asks me in 2025, I will probably report my happiness level as 6 and life satisfaction as also 6. Having experienced the high happiness level at 400, the same level of happiness of 100 that would led me to report a life satisfaction of 8 four decades ago, will in 2025 lead me to report a life satisfaction of only 6. A more important point is this. If my happiness level will be 150 in 2025, I will probably report in 2025 that my happiness level as 6.5 and my life satisfaction level as also 6.5. The crucial comparison now is: Would I prefer:

  • X. A life like me four decades ago with a happiness amount of 100 (reported at that time as 8, but reported now as 6) and a life satisfaction of 8; or

  • Y. A life like me in 2025 with a happiness amount of 150 (reported as 6.5) and a life satisfaction of 6.5?

It is absolutely clear to me that I will have not the slightest hesitation in choosing Y, due to its 50% higher amount of happiness (150 over the figure of 100 in X), despite its lower figure of life satisfaction (6.5 in Y vs. 8 in X). It is true that, being less satisfied with life in Y than in X should itself reduces the happiness in Y somewhat. However, this effect should have already been taken into account in the happiness figure of 150 that should be inclusive of all affective feelings, including the happy or unhappy feeling in evaluating the life satisfaction. Happiness is the net sum total of all such affective feelings that are valuable to the individual. Thus, the happiness index is more appropriate than the life satisfaction index. (More on this in Chap. 5.)Footnote 5

In the above example concerning myself, the happiness and life satisfaction indices reported contemporarily are the same (8 and 8 four decades ago; 9 and 9 now; etc.). However, cases where the two diverge contemporarily may also be possible. Consider this likely possible though hypothetical example. Consider a fairly happy (amount of happiness = 100) and ambitious young man who reported a happiness level of 7 and life satisfaction of 6 (as his ambition for much higher achievements had far from being realized). Twenty years later, he has experienced much real-life problems and has also come to know many miseries of others, etc. His own happiness amount drops from 100 to 50. He then reports his happiness as 6 but his life satisfaction as 7. His life satisfaction index goes up despite a drop in the actual happiness amount and the reported level because of a much lower level of ambition. The crucial question is: If you have the chance to live only one of these two periods of his life, which one would you want? The earlier one with higher happiness and lower life satisfaction, or the later one with lower happiness but higher life satisfaction? I believe that most people, myself included, will choose the former. The answer to this question largely depends on whether one takes happiness or life satisfaction to be intrinsically valuable, ultimately speaking. The next chapter argues for happiness as the only intrinsic value. Also, by taking psychological happiness in the sense of feeling good instead of life satisfaction or ‘attitudinal’ happiness as of the ultimate value, many controversies in moral philosophy may be resolved; an example is discussed in Appendix A.