Is the First Advantage Luck? Option Luck and Brute Luck, Ex Ante and Ex Post, and the Evidence on Initial Advantages and Fairness Judgments in the Academic Literature

Where does the first advantage come from?

Do Those Who Have Receive More? Mechanisms, Causal Identification, and Limits of Cumulative Advantage and the Matthew Effect in the Academic Literature examined the process by which those who were initially advantaged keep accumulating advantages (cumulative advantage). Its conclusion was that the process in which randomly assigned early success increases later success has been confirmed, but the evidence does not show that gaps necessarily widen. However, that note did not ask what the starting advantage consists of. In experiments, researchers assign the advantage at random, so the starting point is luck by design. In real life, the starting point mixes the family into which a person is born, the month of birth, the state of the economy in the year of graduation, lotteries, and gambles that the person chose.

The Japanese terms jizenteki un and jigoteki un (literally “ex ante luck” and “ex post luck”) may seem suitable for thinking about this mixture. They express the intuition that luck already determined before a choice differs from luck that falls on a person as a result after the choice. However, a single pair of terms does not necessarily correspond to this intuition. Each field has built its own distinctions of luck, and even when the same words “ex ante” and “ex post” are used, they separate different things.

This note first examines what the terms for kinds of luck refer to in each field, then identifies what becomes an initial advantage in empirical studies, and finally examines how people judge differences due to luck in experiments and surveys. The answer reached at the end is that most initial advantages derive from luck that the person did not choose, but that the type of luck affects both whether the advantage persists and whether people regard it as unfair.

Scope and method

• Target: peer-reviewed papers and academic encyclopedia entries on the philosophy of luck egalitarianism and moral luck, the ex ante and ex post perspectives in the economics of equality of opportunity, factors that can become initial advantages (birth order, relative age, economic conditions at graduation, and lottery-based natural experiments), and people’s judgments about inequality due to luck (experiments and surveys).
• Collection: candidates were collected across academic databases (Crossref, OpenAlex, and Semantic Scholar), and Japanese sources were collected from CiNii Research and J-STAGE. Duplicates, retracted papers, and predatory journals were removed. Books (Roemer’s Equality of Opportunity and Frank’s general book on luck and success) were excluded. Arneson (1989), Cohen (1989), Kahn (2010), Fong (2001), and Gromet et al. (2015) were excluded because neither the full text nor the abstract could be obtained.
• Verification: for 29 of the 35 sources, the full text (the published version, an author version, a working paper version, or an encyclopedia entry) was obtained, and the numbers and claims were checked against the relevant passages. For the sources read in working paper versions (Fleurbaey and Peragine, Ramos and Van de gaer, Fleurbaey 2010, Black et al., Schwandt and von Wachter, Oyer, Bleakley and Ferrie, Cesarini et al., Deming et al., Alesina and Angeletos, and Möllerström et al.), only claims consistent with the abstract of the published version were used, and the version is stated for each number. The definitions in Dworkin (1981), Lippert-Rasmussen (2001), and Williams and Nagel (1976), whose full texts could not be obtained, were treated within the range quoted by secondary sources whose full texts were read (Otsuka 2002, Vallentyne 2002, Abe 2018, and entries of the Stanford Encyclopedia of Philosophy). For the remaining 3 sources (Bedard and Dhuey 2006, Genda et al. 2010, and Almås et al. 2020), claims were attributed to the authors within the range stated in the abstracts.
• Presentation: each section separates the definition, what was measured (identification), the magnitude and persistence of the effect, and its limits.

Where are the terms “ex ante luck” and “ex post luck” used?

We first checked whether the terms exist. A CiNii Research search of titles, abstracts, and subject headings returned 0 records for jizenteki un (“ex ante luck”), jigoteki un (“ex post luck”), and jizen no kikai no byodo (“ex ante equality of opportunity”). In the English literature read for this note, no paper defined ex ante luck and ex post luck as two types of luck. This pair of terms is, at least, not an established academic term.

The established Japanese terms are translations of Dworkin’s pair. Abe (2018) translated option luck as sentakuteki un and brute luck as genzentaru un, and developed the argument after translating Dworkin’s definitions (pp. 298–299). Nishiguchi (2012) translated option luck as sentakuteki un and brute luck as hisentakuteki un (“non-chosen luck”); the English title of the paper is “Brute luck and equal treatment.” The word jigoteki (“ex post”) appears in the opening of Abe (2018) as an adverb, in the statement that the influence of bad luck cannot be identified ex post (p. 291). Here, “ex post” refers to the problem of whether the influence of luck can be distinguished afterward.

Which distinction comes close to the intuition of luck before a choice and luck after a choice? At least three distinctions come close, and each draws a different line.

Luck in gambles that were chosen and luck that was not chosen

The most widely used distinction is that of Dworkin (1981). According to Dworkin’s definition as quoted by Otsuka (2002), option luck is “a matter of how deliberate and calculated gambles turn out—whether someone gains or loses through accepting an isolated risk he or she should have anticipated and might have declined,” and brute luck is “a matter of how risks fall out that are not in that sense deliberate gambles” (Otsuka 2002, p. 40). The same definition is quoted by Vallentyne (2002, p. 533) and by Lippert-Rasmussen’s entry in the Stanford Encyclopedia of Philosophy (as page 73 of the reprint in Dworkin 2000). In the entry’s example, suddenly going blind because of a genetic condition is bad brute luck, whereas buying a lottery ticket and winning is good option luck.

This distinction determines the type of luck by whether the person accepted the risk, not by the nature of the event. For this reason, insurance can move the line. As quoted by Otsuka (2002, p. 43), Dworkin claimed that insurance, so far as it is available, provides a link between brute and option luck, because the decision to buy or reject catastrophe insurance is a calculated gamble. Even if going blind is itself brute luck, ending up badly off as a result reflects option luck if the person declined insurance. Otsuka criticized this claim and argued that an equal opportunity to insure does not make luck unproblematic.

Option luck is itself further divided. According to Abe (2018, p. 299), Lippert-Rasmussen (2001, pp. 555–556) distinguished “gambles in the strict sense,” which are made because the person prefers gambling itself, from “quasi-gambles,” which are made to obtain the expected value. The argument is that redistributing from the lucky to the unlucky in casino gambling would remove the point of the choice, but the same does not hold for quasi-gambles. Lefranc et al. (2009, footnote 9) noted that Lippert-Rasmussen (2001) emphasized the strong informational requirements of option luck, which presupposes that agents share similar subjective and objective probabilities of outcomes.

Dworkin’s distinction draws one line for thinking about the starting point of cumulative advantage. An advantage gained through a gamble that the person accepted is option luck, and an advantage gained through a risk that the person did not accept is brute luck. However, this distinction has no time axis that indicates when the luck occurred.

Luck before and after the point at which a person can be held responsible

Vallentyne (2002) introduced a time axis. Vallentyne first argued that defining brute luck as what is not a foreseeable outcome of a person’s choices is inadequate, and examined a view based on whether the outcome was reasonably avoidable (pp. 531–533). He then wrote that brute luck includes luck in initial opportunities (one’s initial genetic endowment and initial social position) and some kinds of outcome luck (such as whether an uninfluenceable possibility of a lightning strike is realized) (p. 538). Vallentyne’s claim is that justice requires equalizing the value of initial opportunities for advantage rather than always equalizing brute luck advantage (p. 529). Equality of initial opportunities calls for equality in the value of the initial decision trees that each agent confronts, whereas equality of brute luck also calls for compensating inequalities in later brute outcome luck (p. 539).

Lefranc et al. (2009) summarized this distinction as a pair of initial brute luck and later brute luck. Initial brute luck is the set of factors that influence lifetime prospects up to the moment when individuals can be considered responsible for their choices, and later brute luck denotes luck after that moment. On when to place this moment, Vallentyne named two candidates, the onset of sentience and the onset of full moral standing (for example, adulthood), and left the issue open (p. 540).

This distinction comes closest to the intuition of “luck before a choice.” The family into which a person is born and genetic endowment correspond to initial brute luck. The state of the economy in the year of graduation is luck that the person did not choose and that occurs after the point of responsibility, so it corresponds to later brute luck. In Vallentyne’s position, inequalities due to later brute luck do not necessarily require compensation if initial opportunities are equal and the luck falls evenly on people. Lefranc et al. (2009) wrote that equalizing initial prospects does not prevent individuals from ending up, ex post, with different outcomes as a result of brute luck, but ensures that later brute luck is ex ante even-handed, that is, uncorrelated with circumstances.

The natural lottery of Rawls (1999) refers to the part of initial brute luck that concerns the distribution of natural abilities. Rawls wrote that a system that permits distributive shares of income and wealth to be determined by the natural distribution of abilities and talents leaves the distribution to the outcome of the natural lottery, and that this outcome is arbitrary from a moral perspective (p. 64). On the same page, Rawls wrote that even the willingness to make an effort and to try depends on happy family and social circumstances. In contrast to Dworkin’s distinction, which places effort on the side of choice, Rawls pointed out that the capacity for choice itself is affected by initial luck.

Four kinds of moral luck

The philosophy of moral luck divided luck from a different question. The question is whether actions that depend on factors beyond a person’s control may be morally assessed. According to Nelkin’s entry in the Stanford Encyclopedia of Philosophy, Nagel identified four kinds of luck.

• Resultant luck: luck in the way things turn out. An example is a pair of drivers with the same intentions and plans, only one of whom hits a child who appears in the road.
• Circumstantial luck: luck in the circumstances in which one finds oneself. People who collaborated with the Nazis because they lived in Germany in the 1930s might have led exemplary lives if their companies had transferred them to Argentina in 1929.
• Constitutive luck: luck in who one is, or in the traits and dispositions that one has. Genes, caregivers, peers, and other environmental influences make people who they are, and these are beyond their control.
• Causal luck: luck in “how one is determined by antecedent circumstances,” as Nagel put it (quoted as Nagel 1979, p. 60). This is essentially the problem of free will, and some have viewed it as fully captured by the combination of constitutive and circumstantial luck.

Measured against the intuition behind “ex ante luck” and “ex post luck,” constitutive luck and causal luck are luck that precedes an action, and resultant luck is luck determined as the result after an action. Circumstantial luck includes both a part determined before an action, such as the era and place of birth, and a part concerning the situations encountered during an action.

However, the distinctions of moral luck serve to examine the attribution of responsibility and blame, and their purpose differs from that of the distinctions by Dworkin and Vallentyne, which examine distributive fairness. For the same constitutive luck, the moral luck debate asks whether a person may be blamed for actions performed because of his or her character, whereas luck egalitarianism asks whether disadvantages due to differences in endowment should be compensated. Lefranc et al. (2009) called differences in genetically inherited talent “genetic luck” and placed it as a concept corresponding to constitutive luck and to Rawls’s natural lottery.

In economics, ex ante and ex post distinguish the points at which fairness is judged

In the economic literature on equality of opportunity, ex ante and ex post are not names of types of luck. According to Fleurbaey and Peragine (2009 working paper version), the ex post approach focuses on outcome inequalities among individuals who exert the same effort, and therefore requires identifying individual effort. The ex ante approach focuses on differences between the outcome prospects of classes of individuals with identical circumstances. The authors noted that the distinction between an ex ante situation, in which circumstances are determined but effort is not yet, and an ex post situation, in which all variables are determined, does not always correspond to a real time sequence and is a convenience for interpreting ethical principles. They then showed that researchers motivated by ex post equality of opportunity may be led to believe that ex ante equality of opportunity is another natural embodiment of the same idea, and that it is not. Ramos and Van de gaer (2012 working paper version) defined ex post compensation and ex ante compensation as axioms and showed with an example that the two cannot in general be satisfied together.

In this literature, luck is treated as a third factor alongside effort and circumstances. Lefranc et al. (2009) divided the determinants of outcomes into circumstances, whose effect should be compensated, effort, which is a legitimate source of inequality, and luck, whose impact on outcomes should be even-handed, and illustrated luck with four ideal types. These are differences in family social connections (social background luck, Rawls’s social lottery), differences in genetically inherited talent (genetic luck), later brute luck from a lottery that a person could not escape (exemplified by the Vietnam draft lottery), and informed option luck, which corresponds to Dworkin’s option luck (pp. 1190–1191). The authors concluded that almost all authors agree that social background luck should be fully compensated, but views diverge on genetic luck and option luck.

On measurement, ex ante prospects cannot be observed. Lefranc et al. (2009) wrote that all the conceptions of equality of opportunity equalize the distribution of outcomes before individuals know their degree of luck, that is, the ex ante distribution, but this distribution is never observable empirically, and what can be recovered is the ex post distribution among individuals of a given type (§3.3). When individuals of the same type are exposed to a common shock (children of farmers hurt by bad weather in the same year), the ex post distribution may differ from the ex ante prospects. The authors acknowledged that, whatever cautions one takes, it is an ex post view that is tested empirically.

In welfare economics, the same conflict had already been discussed for the evaluation of risky social situations. Fleurbaey (2010, working paper version) contrasted ex-ante egalitarianism, which applies inequality aversion to the distribution of individual expected utilities, with ex-post egalitarianism, which evaluates prospects by the expected value of an inequality-averse social welfare function. Ex-ante egalitarianism has the difficulty that, if climate change will harm either the high or the low latitudes but it is unknown which, it judges the change harmless as long as individual expected utilities are not diminished. Ex-post egalitarianism has the difficulty that it may not respect individual ex ante preferences even when no inequality arises in any possible consequence.

How the three distinctions correspond

The following table relates the intuition behind “ex ante luck” and “ex post luck” to the three distinctions.

The table shows that the intuition, which appears to rest on a single axis of “before or after a choice,” actually mixes two axes. One is whether the person chose (option luck and brute luck). The other is when it occurred (initial and later). The state of the economy in the year of graduation is not chosen by the person but occurs after the point of responsibility. Starting a business as a gamble is chosen by the person and also occurs later. The month of birth is not chosen by the person and is determined initially. To classify initial advantages, these two axes need to be set up separately.

What becomes an initial advantage?

The empirical literature uses luck in the opposite direction from philosophy. Natural experiments in economics look for random assignments that the person did not choose (month of birth, year of graduation, and lotteries) and use them to identify causal effects. The advantages identified in this way are, by definition, advantages due to brute luck. The following subsections separate initial brute luck from later brute luck and examine how large the advantages are and how long they last.

Birth order and month of birth

Birth order is a typical case of initial brute luck. Black, Devereux, and Salvanes (2005) used data on the entire population of Norway to separate the effects of family size and birth order on children’s education. The negative correlation between family size and education became negligible when birth order was controlled or when twin births were used as an instrument. In contrast, birth order had a significant and large negative effect on education, and birth order effects were also found for adult earnings, employment, and teenage childbearing, particularly among women (abstract and text of the working paper version). Children born later were disadvantaged in education.

The effect of the month of birth was examined as the relative age effect in Do Those Who Have Receive More? Mechanisms, Causal Identification, and Limits of Cumulative Advantage and the Matthew Effect in the Academic Literature. As shown there, the youngest children in a school cohort score lower than the oldest (abstract of Bedard and Dhuey 2006), and in Norway the earnings effect of school starting age had essentially disappeared by age 30. For Japan, Kawaguchi (2011) used the Japanese school system, in which April 1 is the cutoff date. According to the author version, students born in April to June scored 0.11 to 0.22 standard deviations higher than students born in January to March on tests in grades 4 and 8. Among men born in 1968, eventual years of education were 13.20 for those born in April to June and 13.03 for those born in January to March, and men born in January to March earned 3.9% less per year than men born in April to June. Among women, there was a difference in years of education (12.96 and 12.88), but no difference in annual earnings was found. In Japan, the initial luck of the month of birth reached men’s earnings through years of education.

The state of the economy in the year of entering the labor market

The state of the economy in the year of graduation is a typical case of later brute luck. People can choose the timing of further education and job entry to some extent, but they cannot choose the year in which the economy deteriorates.

Oreopoulos, von Wachter, and Heisz (2012) examined the effect of the unemployment rate at graduation using large longitudinal data on Canadian college graduates. A typical recession (a rise in unemployment rates by 5 percentage points) implied an initial loss in earnings of about 9%, which halved within 5 years and faded to zero by 10 years. These reductions added up to a loss of about 5% of cumulative earnings over 10 years. The effect differed between individuals: graduates with the lowest predicted earnings based on college and major suffered a loss of 8% of cumulative earnings, almost double that of the median graduate and more than four times that of the top graduate. Recovery from the loss occurred through moving from the lower-paying employers that graduates first entered to better employers. More advantaged graduates moved to better firms quickly, whereas cyclical downgrading persisted for less advantaged graduates.

The Japanese results show even greater persistence. According to the abstract of Genda, Kondo, and Ohta (2010), comparable labor force surveys in Japan and the United States showed a persistent, strong negative effect on earnings of entering the labor market in a recession for less educated Japanese men, in contrast to no long-term effect for less educated American men. A substantial part of the effect for less educated Japanese men was attributed to the decreased probability of regular employment. The authors proposed the mechanism that, in a dual labor market, people who graduate in a recession are more likely to be trapped in the secondary sector.

The abstract of Schwandt and von Wachter (2019), which examined labor market entrants in the United States from 1976 to 2015, also reported persistent earnings and wage reductions, especially for less advantaged entrants, that increases in government support only partly offset. The abstract of Oyer (2008) states that stock market conditions while MBA students are in school have a large effect on whether they go directly to Wall Street upon graduation. It also states that starting on Wall Street immediately causes a person to be more likely to work there later and to earn substantially more on average, and concludes that investment bankers are largely “made” by circumstance rather than “born.”

Later brute luck works through sorting at the point of entry. Sorting into the first employer, into regular or nonregular employment, and into an industry determines the subsequent path. The effect of this sorting persists longer for groups that were already disadvantaged.

Wealth allocated by lottery and schools allocated by lottery

Lotteries are situations in which random assignment occurs in real life without being created by researchers. The results of lotteries that allocate wealth showed that the initial advantage did not persist.

Bleakley and Ferrie (2016) used the Cherokee Land Lottery held in Georgia in 1832. Nearly every adult white male in Georgia took part, and winners received close to the median level of wealth, independent of their personal characteristics. When Bleakley and Ferrie followed the descendants of participants for nearly fifty years, winners had slightly more children but did not send them to school more. Sons of winners had no better adult outcomes in wealth, income, or literacy than sons of nonwinners, and winners’ grandchildren did not have higher literacy or school attendance (abstract of the published version). Bleakley and Ferrie concluded that, in this environment, family financial resources play only a limited role in the formation of human capital in the next generations, and that other factors that persist through family lines may play a more important role.

Cesarini, Lindqvist, Östling, and Wallace (2016) examined the same question for the present day using administrative data on Swedish lottery players. According to the abstract of the published version, wealth had no detectable effect on adult mortality or health care utilization. For children, wealth increased health care utilization in the years following the lottery and may also have reduced obesity risk, but the effects on most other outcomes, including scholastic performance and skills, could usually be bounded to a tight interval around zero. Cesarini et al. concluded that, in affluent countries with extensive social safety nets, causal effects of wealth are not a major source of the wealth-mortality gradient or of the observed relationships between child developmental outcomes and household income.

The results of lotteries that allocate schools differ from these. According to the abstract of Deming, Hastings, Kane, and Staiger (2014), students who won the school choice lottery in Charlotte-Mecklenburg in the United States and attended their first-choice school showed a significant overall increase in college enrollment and degree completion. The effects were strongly predicted by gains in school quality and were concentrated among girls.

The difference between the wealth lotteries and the school lottery can be read as a difference in the route through which luck was allocated. Family financial resources, at least in these environments, were not readily converted into the human capital of the next generation. In contrast, sorting into a school, like the state of the economy in the year of graduation, changed the subsequent path.

Randomly assigned early success and models

Experiments in which researchers randomly assign early success (van de Rijt et al. 2014 and others) were examined in Do Those Who Have Receive More? Mechanisms, Causal Identification, and Limits of Cumulative Advantage and the Matthew Effect in the Academic Literature. Giving a single donation increased subsequent donations from 1.11 to 2.49, but increasing the number of donations given produced little further differentiation. From the recipient’s point of view, the advantage in these experiments is brute luck that the recipient did not accept as a gamble.

Pluchino, Biondo, and Rapisarda (2018) started from the observation that talent follows a normal distribution whereas wealth follows a power law, and examined the roles of talent and luck in an agent-based model. In the model, some degree of talent was necessary for success, but the most talented people almost never reached the highest peaks of success and were overtaken by mediocre but noticeably luckier individuals. This is a result of a model and does not measure what proportion of real-world success is due to luck.

How do people judge differences due to luck?

Philosophy separated whether compensation is needed according to the type of luck, but whether people actually judge in this way is a separate question. A series of studies, mainly by researchers in Norway, measured this question experimentally.

Cappelen, Hole, Sørensen, and Tungodden (2007) estimated the distribution of fairness views in a dictator game in which production was determined by the amount invested (the person’s choice) and a randomly assigned rate of return (luck). The estimated shares were 43.5% strict egalitarians, who regard all inequalities in outcomes as unfair, 38.1% liberal egalitarians, who regard only inequalities due to choices as fair, and 18.4% libertarians, who regard it as fair for each person to receive what he or she produced, including inequalities due to luck. People’s fairness views differed according to whether they regarded inequalities due to luck as fair.

Cappelen, Konow, Sørensen, and Tungodden (2013) focused on risk-taking and measured the conflict between the ex ante and ex post perspectives directly. Participants chose between a safe alternative and a risky alternative, and were then paired with another participant who had faced the same choice and asked to distribute the total earnings of the pair. The two members of a pair had equal ex ante opportunities, but their ex post earnings could differ. The authors defined the ex ante fairness view, which focuses on initial opportunities, as providing a fairness argument for no redistribution of gains and losses from risk-taking when opportunities are equal, and the ex post fairness view, which focuses on outcomes, as considering it fair to eliminate all inequalities resulting from risk-taking. In the estimates, the ex ante fairness view was the most frequent and accounted for the behavior of around 40% of participants (42.7% among third-party spectators). The view that equalizes all ex post inequalities accounted for about 30%, and the remainder was accounted for by choice egalitarianism, which accepts differences due to choices but eliminates differences due to luck. When a lucky risk-taker was paired with an unlucky risk-taker, spectators split earnings equally in more than 40% of the situations, but when an unlucky risk-taker was paired with a participant who had chosen the safe alternative, they did so in fewer than 20%. Even when the value of the safe alternative was changed so that risk was difficult to avoid, the unlucky risk-taker received, on average, about one-third of the total earnings. Many people judged that part of the ex post inequality due to luck should be redistributed even when ex ante opportunities were equal.

Möllerström, Reme, and Sørensen (2014 working paper version) generated controllable luck (option luck) and uncontrollable luck (brute luck) independently and asked third parties to redistribute. Under luck egalitarianism, losses due to brute luck should be compensated regardless of how the person dealt with option luck. However, many spectators made compensation for bad brute luck conditional on whether the person had avoided exposure to option luck. They did so even when the person’s choice did not affect the outcome. Möllerström et al. called this norm “choice compensation,” and the estimated share of luck egalitarians was about 1% (the abstract of the published version reports this conditioning). People’s judgments looked at the person’s behavior rather than at the cause of the luck.

Fairness views change with age. Almås, Cappelen, Sørensen, and Tungodden (2010) asked Norwegian students from grade 5 to grade 13 to perform a task in which earnings were determined by production (the person’s achievement) and a randomly assigned price (luck), and then to choose a distribution (author version). The share of strict egalitarians decreased from 63.6% in grade 5 to 22.4% in grade 13, and the share of meritocrats, who regard inequalities due to achievement as fair, increased from 5.4% to 42.8%. The importance of achievement increased with age, but no similar developmental trend was found in the importance of luck (the price). As children entered adolescence, they increasingly viewed inequalities reflecting achievements as fair, but not inequalities reflecting luck.

Countries also differ. According to the abstract of Almås, Cappelen, and Tungodden (2020), in a large-scale experiment in which Americans and Norwegians made distributive choices in identical environments, the two groups differed significantly in fairness views but not in the importance assigned to efficiency. Almås et al. also provided causal evidence suggesting that fairness considerations are more fundamental for inequality acceptance than efficiency considerations.

Beliefs about luck are also connected to social institutions. Alesina and Angeletos (2003 working paper version) showed in the World Values Survey that people who believe that luck determines income are more likely to identify as leftist. They then built a model with multiple equilibria. In the model, a society that believes that effort determines income chooses low redistribution and low taxes, so that effort is high, the role of luck is limited, and the beliefs are self-fulfilled. A society that believes that luck, birth, connections, or corruption determine wealth taxes heavily, and these beliefs are also self-sustained. The authors’ claim is that beliefs not only reflect facts but can also create facts through institutions.

Successful people are less likely to attribute their success to luck. Deffains, Espinosa, and Thöni (2016) randomly assigned participants to a hard or an easy task, without telling them which, and divided them into successful and unsuccessful groups. Although success was determined exogenously, successful participants were more likely to attribute their success to effort rather than luck and chose less redistribution, whereas unsuccessful participants tended to attribute their failure to external factors and chose more redistribution. Deffains et al. concluded that this self-serving bias contributes to a polarization of views on redistribution. At the societal level, Mijs (2021) used 25 years of International Social Survey Programme data to show that the more unequal a society, the more likely its citizens are to explain success in meritocratic terms and the less important they deem non-meritocratic factors such as family wealth and connections.

Criticisms of luck egalitarianism and limits of identification

The very idea of separating compensation by the type of luck has met strong criticism. Anderson (1999) named the position of Arneson, Cohen, Dworkin, Nagel, Rakowski, Roemer, and others luck egalitarianism, and argued that it fails the most fundamental test that any egalitarian theory must meet, namely that its principles express equal respect and concern for all citizens (pp. 288–290). The first failure is that it excludes some citizens from the social conditions of freedom on the ground that losing them was their own fault. In Anderson’s example, when an uninsured driver negligently makes an illegal turn and causes an accident, the logic of luck egalitarianism leaves him to die by the side of the road (the “abandonment of negligent victims,” from p. 295). The second is that it bases compensation on the fact that some people are inferior to others in the worth of their lives and talents, so that compensation expresses pity and stigmatizes the unfortunate. The third is that, in attempting to hold people responsible, it requires demeaning and intrusive judgments of people’s capacities. Anderson proposed democratic equality instead, according to which the point of equality is not to eliminate the impact of brute luck but to end oppression and to create a community in which people stand in relations of equality to others.

Equality of initial opportunities also has limits. Lippert-Rasmussen’s entry in the Stanford Encyclopedia of Philosophy gives the example of a caste society in which babies are assigned starting positions by a fair lottery. This society may well realize initial equality of opportunity, yet it does not seem just, and it is far from clear that the lottery reduces its injustice at all. Even when initial advantages are allocated by a fair lottery, the size of the gap between advantage and disadvantage remains unaddressed.

The empirical side also has limits in distinguishing luck. Abe (2018, p. 291) argued that the extent to which choice and bad luck each influenced a disadvantage cannot be identified ex post, and examined a hypothetical insurance scheme on that premise. As described above, Lefranc et al. (2009) acknowledged that only the ex post distribution is observable. Natural experiments partly avoid this limit by identifying luck as random assignment. However, natural experiments can identify only luck that is assigned from outside, such as the month of birth, the year of graduation, and lotteries, and most luck, such as genetic endowment and family environment, cannot be observed in the form of random assignment.

What can be said about the first advantage?

Returning to the opening question, “ex ante luck” and “ex post luck” are not established academic terms, and the intuition behind them mixes two axes. One is whether the person chose (option luck and brute luck, Dworkin), and the other is when the luck occurred (initial brute luck and later brute luck, Vallentyne). The ex ante and ex post perspectives in economics distinguish the points at which fairness is judged, and the two perspectives cannot in general be satisfied together (Fleurbaey and Peragine 2013).

The initial advantages identified in empirical studies are, by definition, due to brute luck. Their persistence differed according to the route through which luck was allocated. Luck that passed through sorting at the point of entry (the first employer or regular employment depending on the economy in the year of graduation, and schools allocated by lottery) persisted longer for groups that were already disadvantaged (Oreopoulos et al. 2012, Genda et al. 2010, Deming et al. 2014). Luck in wealth allocated to families was hardly converted into the human capital of children and grandchildren (Bleakley and Ferrie 2016, Cesarini et al. 2016). The initial luck of birth order and month of birth sometimes reached earnings through education (Black et al. 2005, Kawaguchi 2011).

People’s judgments broadly correspond to the philosophical distinctions but do not coincide with them. Many people distinguished differences due to choices from differences due to luck (Cappelen et al. 2007, 2013, Almås et al. 2010), and judged that part of the differences due to luck should be redistributed even when ex ante opportunities were equal (Cappelen et al. 2013). However, many people made compensation for brute luck conditional on the person’s behavior (Möllerström et al.), and successful people tended to attribute exogenously determined success to effort (Deffains et al. 2016).

Based on the above, this note proposes the following organization. When discussing initial advantages, luck needs to be separated along the two axes of choice and timing, and the route through which that luck became an advantage (resources, sorting, or recognition) also needs to be separated. The “process” shown in the literature on cumulative advantage concerns this last route, and which kinds of luck tend to become the starting point of accumulation depends on the sorting mechanisms.

This organization would need revision if the following observations accumulated. Such observations would be multiple results, in present-day environments with thin social safety nets, showing that wealth randomly allocated to families persistently raises the education or earnings of children and grandchildren. In that case, the conclusion that luck in wealth is hardly converted into human capital would be a property of specific periods and countries.

Gaps

• Sources of the terms “ex ante luck” and “ex post luck”: the terms were not found in a search of titles and abstracts of Japanese academic literature. The possibility that they come from general books, lectures, or the terminology of particular authors was not examined in this note.
• Attributions to luck and support for redistribution: Fong (2001) and Gromet et al. (2015) were excluded because neither the full text nor the abstract could be obtained. Peer-reviewed studies showing that attributing outcomes to luck does not increase support for redistribution, or that the effect varies with political ideology, could not be checked in full text in this review.
• Underestimation of luck by successful people: the experiment of Deffains et al. (2016) comes close, but no peer-reviewed study was found that directly measured whether successful people in real life underestimate the share of luck in their own success.
• The genetic lottery: the treatment of genetic endowment as initial brute luck was discussed in Is the Ability to Make an Effort Also a Talent? The Heritability of Grit, Self-Control, and Motivation to Learn, and How Far These Traits Can Be Changed in relation to cumulative advantage and heritability. Peer-reviewed studies incorporating genetic evidence into the context of luck egalitarianism were not collected in this review.
• Attributions of luck and interpersonal relationships: how attributing achievements to luck, effort, or talent affects views of and burdens on high performers was discussed in Why Is the Stress of Outperformers Hard to Voice? Evidence on Displaying Competence, Being Envied, Hiding Success, and Invisible Effort.

Unverified items

• No major claim remains unverified.
• Three sources could not be obtained in full text and were treated within the range quoted by secondary sources (Dworkin 1981 through Otsuka 2002, Vallentyne 2002, and Lippert-Rasmussen’s encyclopedia entry; Lippert-Rasmussen 2001 through Abe 2018 and Lefranc et al. 2009; Williams and Nagel 1976 through Nelkin’s encyclopedia entry). In the text, the quoting source is stated in each case.
• Three sources were treated within the range of their abstracts (Bedard and Dhuey 2006, Genda et al. 2010, and Almås et al. 2020). In the text, these claims are attributed to the authors with “according to the abstract.” For Schwandt and von Wachter 2019, Oyer 2008, and Deming et al. 2014, the full text of the working paper version was obtained, but only claims within the range of the published abstract were used, and the text states “according to the abstract.”
• The numbers from Kawaguchi (2011) and Almås et al. (2010) are from the author versions, and the estimates from Möllerström et al. (2015) are from the working paper version (2014); they were not checked against the published versions.

References

All sources were accessed on 2026-10-09.

Luck egalitarianism and the philosophy of luck

• Dworkin, R. (1981). What is equality? Part 2: Equality of resources. Philosophy & Public Affairs, 10(4), 283–345. https://www.jstor.org/stable/2265047
• Anderson, E. S. (1999). What is the point of equality? Ethics, 109(2), 287–337. https://doi.org/10.1086/233897
• Lippert-Rasmussen, K. (2001). Egalitarianism, option luck, and responsibility. Ethics, 111(3), 548–579. https://doi.org/10.1086/233526
• Vallentyne, P. (2002). Brute luck, option luck, and equality of initial opportunities. Ethics, 112(3), 529–557. https://doi.org/10.1086/339275
• Otsuka, M. (2002). Luck, insurance, and equality. Ethics, 113(1), 40–54. https://doi.org/10.1086/341322
• Rawls, J. (1999). A theory of justice (Rev. ed.). Belknap Press of Harvard University Press. https://archive.org/details/rawls99
• Williams, B., & Nagel, T. (1976). Moral luck. Aristotelian Society Supplementary Volume, 50(1), 115–152. https://doi.org/10.1093/aristoteliansupp/50.1.115
• Nelkin, D. K. (2025). Moral luck. In E. N. Zalta & U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/moral-luck/
• Lippert-Rasmussen, K. (2023). Justice and bad luck. In E. N. Zalta & U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/justice-bad-luck/
• Abe, T. (2018). Luck egalitarianism, the harshness objection, and hypothetical insurance: How should we deal with bad luck that coexists with choice? (Un no byodo shugi, kakokusei hihan, kaso hoken: Sentaku to heizon suru fuun ni ika ni taisho subeki ka). Seiji Shiso Kenkyu, 18, 291–318. (in Japanese) https://doi.org/10.11326/jjpt.18.0_291
• Nishiguchi, M. (2012). Brute luck and equal treatment: By the clue of the equality of resources approach by Ronald Dworkin (Hisentakuteki un to byodo naru shogu). Ningen Kankeigaku Kenkyu, 11, 13–23. (in Japanese) https://lib.sugiyama-u.repo.nii.ac.jp/records/2144

Equality of opportunity, ex ante and ex post

• Fleurbaey, M., & Peragine, V. (2013). Ex ante versus ex post equality of opportunity. Economica, 80(317), 118–130. https://doi.org/10.1111/j.1468-0335.2012.00941.x
• Lefranc, A., Pistolesi, N., & Trannoy, A. (2009). Equality of opportunity and luck: Definitions and testable conditions, with an application to income in France. Journal of Public Economics, 93(11–12), 1189–1207. https://doi.org/10.1016/j.jpubeco.2009.07.008
• Ramos, X., & Van de gaer, D. (2016). Approaches to inequality of opportunity: Principles, measures and evidence. Journal of Economic Surveys, 30(5), 855–883. https://doi.org/10.1111/joes.12121
• Fleurbaey, M. (2010). Assessing risky social situations. Journal of Political Economy, 118(4), 649–680. https://doi.org/10.1086/656513

Evidence on initial advantages

• Black, S. E., Devereux, P. J., & Salvanes, K. G. (2005). The more the merrier? The effect of family size and birth order on children’s education. Quarterly Journal of Economics, 120(2), 669–700. https://doi.org/10.1162/0033553053970179
• Bedard, K., & Dhuey, E. (2006). The persistence of early childhood maturity: International evidence of long-run age effects. Quarterly Journal of Economics, 121(4), 1437–1472. https://doi.org/10.1093/qje/121.4.1437
• Kawaguchi, D. (2011). Actual age at school entry, educational outcomes, and earnings. Journal of the Japanese and International Economies, 25(2), 64–80. https://doi.org/10.1016/j.jjie.2009.02.002
• Oreopoulos, P., von Wachter, T., & Heisz, A. (2012). The short- and long-term career effects of graduating in a recession. American Economic Journal: Applied Economics, 4(1), 1–29. https://doi.org/10.1257/app.4.1.1
• Genda, Y., Kondo, A., & Ohta, S. (2010). Long-term effects of a recession at labor market entry in Japan and the United States. Journal of Human Resources, 45(1), 157–196. https://doi.org/10.3368/jhr.45.1.157
• Schwandt, H., & von Wachter, T. (2019). Unlucky cohorts: Estimating the long-term effects of entering the labor market in a recession in large cross-sectional data sets. Journal of Labor Economics, 37(S1), S161–S198. https://doi.org/10.1086/701046
• Oyer, P. (2008). The making of an investment banker: Stock market shocks, career choice, and lifetime income. Journal of Finance, 63(6), 2601–2628. https://doi.org/10.1111/j.1540-6261.2008.01409.x
• Bleakley, H., & Ferrie, J. (2016). Shocking behavior: Random wealth in antebellum Georgia and human capital across generations. Quarterly Journal of Economics, 131(3), 1455–1495. https://doi.org/10.1093/qje/qjw014
• Cesarini, D., Lindqvist, E., Östling, R., & Wallace, B. (2016). Wealth, health, and child development: Evidence from administrative data on Swedish lottery players. Quarterly Journal of Economics, 131(2), 687–738. https://doi.org/10.1093/qje/qjw001
• Deming, D. J., Hastings, J. S., Kane, T. J., & Staiger, D. O. (2014). School choice, school quality, and postsecondary attainment. American Economic Review, 104(3), 991–1013. https://doi.org/10.1257/aer.104.3.991
• van de Rijt, A., Kang, S. M., Restivo, M., & Patil, A. (2014). Field experiments of success-breeds-success dynamics. Proceedings of the National Academy of Sciences, 111(19), 6934–6939. https://doi.org/10.1073/pnas.1316836111
• Pluchino, A., Biondo, A. E., & Rapisarda, A. (2018). Talent versus luck: The role of randomness in success and failure. Advances in Complex Systems, 21(3–4), 1850014. https://doi.org/10.1142/S0219525918500145

Judgments about differences due to luck

• Cappelen, A. W., Hole, A. D., Sørensen, E. Ø., & Tungodden, B. (2007). The pluralism of fairness ideals: An experimental approach. American Economic Review, 97(3), 818–827. https://doi.org/10.1257/aer.97.3.818
• Cappelen, A. W., Konow, J., Sørensen, E. Ø., & Tungodden, B. (2013). Just luck: An experimental study of risk-taking and fairness. American Economic Review, 103(4), 1398–1413. https://doi.org/10.1257/aer.103.4.1398
• Möllerström, J., Reme, B.-A., & Sørensen, E. Ø. (2015). Luck, choice and responsibility: An experimental study of fairness views. Journal of Public Economics, 131, 33–40. https://doi.org/10.1016/j.jpubeco.2015.08.010
• Almås, I., Cappelen, A. W., Sørensen, E. Ø., & Tungodden, B. (2010). Fairness and the development of inequality acceptance. Science, 328(5982), 1176–1178. https://doi.org/10.1126/science.1187300
• Almås, I., Cappelen, A. W., & Tungodden, B. (2020). Cutthroat capitalism versus cuddly socialism: Are Americans more meritocratic and efficiency-seeking than Scandinavians? Journal of Political Economy, 128(5), 1753–1788. https://doi.org/10.1086/705551
• Alesina, A., & Angeletos, G.-M. (2005). Fairness and redistribution. American Economic Review, 95(4), 960–980. https://doi.org/10.1257/0002828054825655
• Deffains, B., Espinosa, R., & Thöni, C. (2016). Political self-serving bias and redistribution. Journal of Public Economics, 134, 67–74. https://doi.org/10.1016/j.jpubeco.2016.01.002
• Mijs, J. J. B. (2021). The paradox of inequality: Income inequality and belief in meritocracy go hand in hand. Socio-Economic Review, 19(1), 7–35. https://doi.org/10.1093/ser/mwy051