Men enter the tournament more than womenbecause they are less risk averse. As tournaments involve uncer-tain payoffs, potential gender differences in risk attitudes are

Explanation 2: Men enter the tournament more than womenbecause they are more overconfident. Psychologists often find thatwhile both men and women are overconfident about their relativeperformance, men tend to be more overconfident than women(e.g., Lichtenstein, Fischhoff, and Phillips [1982], Beyer [1990],and Beyer and Bowden [1997]).

Nurture as well as nature may cause women to be relativelymore reluctant to perform in a competition. First, we tend to raisegirls and boys differently. Parents, teachers, and peers encouragegender-typed activities in children while cross-gender activitiesare discouraged. While boys are encouraged to be assertive, girlsare encouraged to show empathy and be egalitarian [Ruble, Mar-tin, and Berenbaum 2006]. Second, nature may also cause agender difference in preferences for competition. E

Explanation 1: Men enter the tournament more than womenbecause they like to compete.

To assess whether general factors, such as overconfidence,risk, and feedback aversion by themselves cause a gap in choicesof compensation scheme, we also determine if, absent the thrill orfear of performing in a competition, a gender gap in choice ofcompensation scheme still occurs. We find that combined suchfactors cause men and women of equal performance to selectdifferent compensation schemes. This difference appears to belargely explained by gender differences in overconfidence, whilerisk and feedback aversion seem to play a negligible role

Despite there being no gender difference in performance un-der either compensation scheme, we find that twice as many menas women select the tournament. While 73 percent of men preferthe tournament, this choice is only made by 35 percent of thewomen. This gender gap persists when we compare the choices ofmen and women of equal performance. Compared to payoff-max-imizing choices, low-ability men enter the tournament too much,and high-ability women do not enter it enough

Women may not select into top level jobs because they donot enjoy the responsibilities associated with a managerial posi-tion. Or they may avoid these jobs because they tend to have longwork hours, which may conflict with the desire or necessity forchild rearing. Second, discrimination or anticipated discrimina-tion may cause women and men with equal abilities to holddifferent occupations. 1 Gender differences in preferences for com-petition may be an additional explanation for differences in labormarket outcome; in particular, it may help explain the absence ofwomen in top-level and very competitive positions.2

While boys spend most of their time atcompetitive games, girls select activities where there is no winnerand no clear end point. This difference increases through puberty,and by adulthood more men than women describe themselves ascompetitive (see Campbell [2002] for a review of the literature).The objective of this paper is to investigate whether men andwomen differ in their preferences for competition and how suchdifferences impact economic outcome

articipants in a laboratory experimentsolve a real task, first under a noncompetitive piece rate and then a competitivetournament incentive scheme. Although there are no gender differences in per-formance, men select the tournament twice as much as women when choosingtheir compensation scheme for the next performance. While 73 percent of the menselect the tournament, only 35 percent of the women make this choice.

There are a host of ways to intervene to improve women’s and girls’ well-being incontexts where cultural norms deny them equal opportunities. These include laws grantingequal rights, affirmative action policies, subsidies to encourage investment in girls, programsthat impart skills or target resources to women, and many others

Self-reported behavior change continues to be larger for boys than girls in the mediumrun. Our interpretation of this pattern in the short run was that girls were more constrainedby external factors in translating attitudes to behavior, but now another contributing factoris that attitude change is less sustained for them

Thus, the program seems to have increased girls’ intention to go to college by raisingalready-high aspirations and convincing already-engaged parents to support their daughter’sgoals, rather than by converting girls and families who started out lower on these dimensions.

Meanwhile, as in the short run, there is a sizable (0.23 standard deviations) and statis-tically significant effect on self-reported behavior, pooled for boys and girls. Note that weupdated the elements in the behavior index between the two endlines, so the effect size isnot as directly comparable over time as the effect on attitudes is

We continue to find no effect on girls’ stated aspirations (Table 7, column 2). It ispossible an effect on this outcome could have emerged between the endlines, for example ifgirls’ aspirations in the control group became more tempered over time

The control group mean of 0.33 in the secondendline indicates that attitudes became more progressive between the endlines, absent theintervention. Thus, the 11% fade-out in the treatment effect is not because the treatmentgroup held less progressive attitudes at the second endline than first endline; rather, theirattitudes improved slightly less between the two waves than control group’s attitudes did

We next discuss results from the second endline survey, conducted 2 to 2.5 years afterthe program ended

The intervention made personal attitudes about female employment more progressiveby 13 percentage points (Table 5, column 1). It also increased by 5 percentage points theperception that others in the community hold that gender-progressive view (column 2), whichis consistent with other findings that signals from institutions (Breakthrough in this case)can be effective in changing subjective perceptions of norms

Overall,there is very little evidence that parental support for gender equality either facilitates orhinders the average success of the intervention.2

For behavior, we find that the program had a significantly smaller impact for girls thanboys (interaction coefficient of -0.13, p < 0.01), although the net effect for girls is still positiveand significant. One interpretation of this finding is that boys and girls can adopt gender-equal attitudes with relatively equal ease, but girls face more constraints on translating theirattitudes into behavior.

As shown in Table 3, the main effect of having a below-median social desirability scoreon gender attitudes is negative and significant, suggesting some upward shading of responsesoverall for the sample. We view this pattern as reassuring, a validation that the scoreseems to capture the propensity to shade responses. Importantly, there is no more of thisshading up in the treatment group than the control group; the interaction term is small andinsignificant. In other words, the treatment effect remains large and significant when wefocus on the subsample with lower susceptibility to experimenter demand effects.

The third primary outcome is self-reported behaviors influenced by gender attitudes.As reported in column 3 of Table 2, behavior became more aligned with gender-progressivenorms by 0.20 standard deviations

Turning to our second outcome, Table 2 shows that the program did not affect girls’aspirations. The average effect is 0.03 standard deviations and not significantly differentfrom zero.

We find that the intervention converted 15.5% of gender-regressive views into sup-port for gender equality. This persuasion rate is calculated by stacking all of the variablesincluded in the attitudes index, which are coded as binary values. In the treatment group,on average 29.9% of views are gender-regressive, compared to 35.4% in the control group

We find that the intervention made gender attitudes more progressive: Column 1 showsthat students in treatment schools have a 0.18 standard deviation higher attitude index thanthose in control schools (p < 0.01).

Thus, β1 represents the average effect of the intervention on the outcome. The outcomes areconstructed so that a higher value represents more gender progressiveness, so the hypothesisis β1 > 0

The intervention is hypothesized to make participants’ attitudes less discriminatoryagainst females, raise girls’ aspirations, and increase gender-equitable behavior.

Gender attitudes are balanced between the treatment and control group at baseline (seeTable 1).16 They are also quite regressive. For example, about 80% of boys and 60% of girlsbelieve that a woman’s most important role is being a good homemaker. The pattern thatgirls are less likely than boys to endorse gender-discriminatory positions is seen for each ofthe attitude questions (see Appendix Table 4)

For the sample overall, girls’ aspirations are negatively correlated with attrition; girls with high aspi-rations were more likely to still be enrolled in school, which made them easier to track. Appendix Table2 details the reasons for attrition, which include permanent or long-term migration, death or poor health,refusal to participate by the student or parent, not being available at the time of their appointment, and ourinability to track the respondent

We pre-specified three primary outcomes for the first endline: gender attitudes, girls’aspirations, and self-reported gender-related behavior.

There isa high illiteracy rate for mothers, reflecting the low level of female schooling in the parents’generation. A small proportion (29%) of mothers work outside the home, which is consistentwith low female labor force participation rates in India

There was a high rate of parentalconsent; 84% of forms were returned. The rate of returned consent forms does not vary bygender or by village-level proxies for gender norms, namely the child sex ratio and femaleemployment rate from the 2011 Census. Anecdotally, lack of consent was usually due to thestudent losing or forgetting the form

The sample size of 314 schools was chosen to be able to measure the short- and medium-run effects of the program on gender attitudes, aspirations, and behavior, as well as long-termeffects on educational attainment, occupational choice, marriage, and fertility that mightemerge up to ten years after the program ended.

o complement the in-class material and encouragefurther reflection, the facilitators gave out some homework assignments such as writing storiesand recording observations, and they encouraged students to talk to their family membersabout what they were learning. The sessions did not displace one specific subject like mathor history; they crowded out a roughly even mix of material in other subjects. While theprogram could have harmed other learning by displacing instructional time, it is also possiblethat the discuss-and-debate style strengthened students’ critical thinking and speaking skills,conferring value to them beyond the gender focus.

For example, it conveyed that equal oppor-tunity for education is a universal human right. The hypothesis is that this rights-basedmessage would increase how much participants value girls having access to higher education;a boy would get disutility from seeing his sister denied the same chance to attend collegethat he has. As an example of a pragmatic argument for girls’ education, the interventioninformed participants that outcomes for children improve when their mother is more edu-cated. The hypothesis is that this information causes belief updating; girls want to stay inschool longer, and both boys and girls will want to educate their daughters down the road

The objective of the program, which was named Taaron ki Toli, or Legion of Stars, was tocreate awareness of gender-based discrimination, change dominant gendered perceptions andpromote gender-equitable attitudes, raise girls’ aspirations, and provide tools to participantsto translate attitude change and greater aspirations into behavior change. By changingfundamental gender attitudes, the program aimed to ultimately influence a wide range ofbehaviors related to female education, mobility, work, marriage and fertility, for both femaleparticipants and male participants’ female family members (e.g., their future wives).

We find that respondents with a high propensity for social desirability bias express moresupport for gender equality overall for the sample, but importantly, this pattern is not moretrue for the treatment group than the control group. In other words, the positive treatmenteffect on self-reported attitudes and behavior is similar in magnitude for respondents with alow and high propensity for social desirability bias.3 We view this analysis as an importantcheck on the validity of our results, and think that this approach could be useful in a widearray of studies in which experimenter demand is a concern

We examine heterogeneity of the treatment effects by two pre-specified characteristics,the parents’ gender attitudes and the student’s gender. We find no evidence of heterogeneitybased on parents’ attitudes, but we find important differences in treatment effects betweenboys and girls. At the first endline, the effect size on attitudes is similar for boys and girls,but by the second endline, the effect is larger for boys. In addition, behavior change is morepronounced for boys in both the short and medium run. For example, boys report doingmore chores, but girls do not report doing fewer, and boys have a larger increase in howmuch they encourage their older sisters to pursue college. With chores, the smaller effectfor girls can be explained by the asymmetry in the outcome; greater gender equality meansthat girls seek a lower burden, which might not be granted, while boys can voluntarily helpout more. But the fact that we also see heterogeneity by gender in support given to sisterssuggests a broader phenomenon of girls facing more external constraints on their behavior.

The intervention also affected self-reported gender-related behaviors. An index ofgender-equal behaviors increased by 0.20 to 0.23 standard deviations, in both the shortrun and medium run.

We continue to find a large effect on attitudes — 0.16 standard deviations — in themedium run (five years after baseline). The treatment group’s attitudes are more progressiveat the second endline survey than the first one; the 11% fade-out is because the controlgroup’s attitudes improved even more over the intervening two years.

We focus primarily on how the program changed participants’ gender attitudes, that is,their views about what is right and wrong or desirable and undesirable, such as whether itis wrong for women to work outside the home and whether it would be good to have morewomen in politics. A second outcome is girls’ aspirations for their education and career;for this outcome, we hypothesize an effect only on girls. Our third primary outcome isself-reported behaviors influenced by gender norms. It will only be once the participantsbecome older that we can assess impacts on major outcomes they have control over, suchas their employment and childbearing. Nonetheless, we view effects on self-reported day-to-day behaviors as adolescents to be informative about whether reshaping gender attitudestranslates into behavior change. We focus on behaviors that adolescents likely have somesay over, such as chores done at home and interaction with opposite-gender peers

4,000 students.

While boys and girls start sec-ondary school at the same rate, only 0.80 girls enroll in tertiary schooling for every boy(World Bank, 2011). Early marriage and childbearing are common, and many women havelimited agency and physical mobility and face barriers to labor force participation (Fieldet al., 2010; Klasen and Pieters, 2015; Afridi et al., 2018; Calvi, 2020). India’s sex ratioamong children age 0 to 6 years is 1.09 boys per girl, reflective of the widespread practiceof sex-selective abortion; Haryana’s sex ratio of 1.20 is the most male-skewed among Indianstates (Sen, 1990; Jha et al., 2006; Govt. of India, 2011)

The reason for targeting secondary school students specifically is that adolescence isa critical time in the development of morality and formation of identity, when people areyoung enough to still have malleable attitudes but mature enough to reflect on complex

A few sessions taught communication skills to helpstudents convey their views to others and be able to, say, persuade their parents to permitthem to marry at a later age. The program’s messaging combined a human-rights case forgender equity with pragmatic reasons to value women, such as their economic contributions

Discussiontopics included gender stereotypes, gender roles at home, girls’ education, women’s employ-ment outside the home, and harassment

We evaluate a school-based program in the state of Haryana, India,for seventh to tenth graders. The program centered around classroom discussions aboutgender equality in secondary schools, with a 45-minute session held every three weeks fortwo and a half school years. The sessions taught facts and endorsed gender equality and,as importantly, prompted students to reflect on their own and society’s views

Studies suggestthat even centuries-old norms might be amenable to change. For example, reserved seatsfor female politicians reduce gender-biased attitudes in India (Beaman et al., 2009), andtelevision programming can change fertility preferences (Jensen and Oster, 2009; La Ferraraet al., 2012). These findings are part of a broader literature on how people’s preferences areshaped (Becker and Mulligan, 1997; Bowles, 1998).

Gender inequality exists in every society but is especially severe in many developingcountries: Women and girls have fewer educational opportunities, less autonomy over mar-riage and fertility, and more restrictions on labor force participation (Duflo, 2012; Jayachan-dran, 2015).

elf-reported behavior became morealigned with progressive gender norms, particularly among boys.

Using a randomized controlled trial, we find that the program made attitudes0.18 standard deviations more supportive of gender equality, or, equivalently, converted16% of participants’ regressive views

This paper evaluates a school-based interventionin India that engaged adolescents in classroom discussions about gender equality fortwo and a half years with the goal of eroding their support for restrictive gendernorms

Societal norms about gender roles contribute to the economic disadvantages facingwomen in many developing countries

There islittle agreement, however, about what deter-mines institutions and government attitudes to-wards economic progress, making it difficult toisolate exogenous sources of variation in insti-tutions to estimate their effect on performance.In this paper we argued that differences in co-lonial experience could be a source of exoge-nous differences in institutions.

The data support the overidentifying restric-tions implied by our approach.

McArthur and Sachs (2001) reportthat in these regressions the institution variableis still significant, but geography/health are alsosignificant. In contrast to McArthur and Sachs’results, we find that only institutions are signi

La Porta et al. (1999) argue for the impor-tance of colonial origin (identity of the maincolonizing country) as a determinant of currentinstitutions

impact of institutions on incomeper capita is 0.94.

Figure 3 illustrates the relationship between the(potential) settler mortality rates and the index ofinstitutions. We use the logarithm of the settlermortality rates, since there are no theoretical rea-sons to prefer the level as a determinant of insti-tutions rather than the log, and using the logensures that the extreme African mortality rates donot play a disproportionate role. As it happens,there is an almost linear relationship between thelog settler mortality and our measure of institu-tions. This relationship shows that ex-colonieswhere Europeans faced higher mortality rateshave substantially worse institutions today.

The standard measure is annualized deathsper thousand mean strength. This measurereports the death rate among 1,000 soldierswhere each death is replaced with a new soldier

make us believe that settler mortality is aplausible instrument for institutional develop-ment: these diseases affected European settle-ment patterns and the type of institutions theyset up, but had little effect on the health andeconomy of indigenous people.1

In this subsection, we give a brief overviewof the sources of mortality facing potential set-tlers. Malaria (particularly Plasmodium falcipo-rum) and yellow fever were the major sourcesof European mortality in the colonies.

To control forthis, in columns (3)–(6), we add latitude as aregressor (we follow the literature in using theabsolute value measure of latitude, i.e., distancefrom the equator, scaled between 0 and 1)

Table 2 reports ordinary least-squares (OLS)regressions of log per capita income on theprotection against expropriation variable in avariety of samples

We use a variety of variables to capture in-stitutional differences. Our main variable, re-ported in the second row, is an index ofprotection against expropriation.

The gains to an extractive strategy maydepend on the size of the ruling elite. Whenthis elite is small, each member would havea larger share of the revenues, so the elitemay have a greater incentive to be extrac-tive.

etting up institutions that place restrictionson government power and enforce propertyrights is costly

There are a number of economic mechanismsthat will lead to institutional persistence of thistype.

Young emphasizes that the extractive institu-tions set up by the colonialists persisted longafter the colonial regime ended. He writes “al-though we commonly described the indepen-dent polities as ‘new states,’ in reality they weresuccessors to the colonial regime, inheriting itsstructures, its quotidian routines and practices,and its more hidden normative theories of gov-ernance” (1994 p. 283).

Soon afterthe conquest, the Spanish crown granted rightsto land and labor (the encomienda)

An interesting example of the awareness ofthe disease environment comes from the Pil-grim fathers. They decided to migrate to theUnited States rather than Guyana because of thehigh mortality rates in Guyana (see Crosby,1986 pp. 143– 44).

We hypothesize that settler mortality affectedsettlements; settlements affected early institu-tions; and early institutions persisted andformed the basis of current institutions.

Interestingly, our resultsshow that distance from the equator does nothave an independent effect on economic perfor-mance, validating the use of this variable as aninstrument in the work by Hall and Jones(1999).

In contrast to this approach whichfocuses on the identity of the colonizer, weemphasize the conditions in the colonies. Spe-cifically, in our theory—and in the data—it isnot the identity of the colonizer or legal originthat matters, but whether European colonialistscould safely settle in a particular location:where they could not settle, they created worseinstitution

We then useoveridentification tests to detect whether settlermortality has a direct effect on current perfor-mance

We find thatnone of these overturn our results; the estimateschange remarkably little when we include con-trols for the identity of the main colonizer, legalorigin, climate, religion, geography, natural re-sources, soil quality, and measures of ethnolin-guistic fragmentation. Furthermore, the resultsare also robust to the inclusion of controls forthe current disease environment

Interestingly, we show that once the effect ofinstitutions on economic performance is con-trolled for, neither distance from the equator northe dummy for Africa is significant. These re-sults suggest that Africa is poorer than the restof the world not because of pure geographicor cultural factors, but because of worseinstitutions

The major concernwith this exclusion restriction is that the mor-tality rates of settlers could be correlated withthe current disease environment, which mayhave a direct effect on economic performance

The exclusion restriction implied by our in-strumental variable regression is that, condi-tional on the controls included in the regression,the mortality rates of European settlers morethan 100 years ago have no effect on GDP percapita today, other than their effect throughinstitutional development

(potential) settlermortality rates were a major determinant ofsettlements; settlements were a major determi-nant of early institutions (in practice, institu-tions in 1900); and there is a strong correlationbetween early institutions and institutions to-day.

The regression showsthat mortality rates faced by the settlers morethan 100 years ago explains over 25 percentof the variation in current institutions

It shows astrong negative relationship. Colonies whereEuropeans faced higher mortality rates are to-day substantially poorer than colonies that werehealthy for Europeans.

We use data on the mortality rates of soldiers,bishops, and sailors stationed in the colonies be-tween the seventeenth and nineteenth centuries,largely based on the work of the historian PhilipD. Curtin.

At the other extreme, many Europeans mi-grated and settled in a number of colonies,creating what the historian Alfred Crosby(1986) calls “Neo-Europes.

At one extreme, European powers setup “extractive states,” exemplified by the Bel-gian colonization of the Congo. These institu-tions did not introduce much protection forprivate property, nor did they provide checksand balances against government expropria-tion.

To estimate the impact of institutions on eco-nomic performance, we need a source of exog-enous variation in institutions. In this paper, wepropose a theory of institutional differencesamong countries colonized by Europeans,1 andexploit this theory to derive a possible source ofexogenous variation. Our theory rests on threepremises:

Nevertheless, we lack reliable estimates ofthe effect of institutions on economic perfor-mance. It is quite likely that rich economieschoose or can afford better institutions.

Countries with better “institutions,”more secure property rights, and less distor-tionary policies will invest more in physicaland human capital, and will use these factorsmore efficiently to achieve a greater level ofincome

What are the fundamental causes of thelarge differences in income per capita acrosscountries

We exploit differences in European mortality rates to estimate the effect of institu-tions on economic performance.

evertheless, the perspectivedeveloped in this article applies to these cases as well: institu-tional problems are important in a variety of instances, and,in most of these, the source of institutional problems and thedifficulty of institutional reform lie in the fact that any majorchange creates winners and losers, and the potential losersare often powerful enough to resist change.

That there is no naturalgravitation toward good institutions is illustrated by the atti-tudes of the landed elites and the emperors in Austria-Hungary and in Russia during the nineteenth century. Theseelite groups blocked industrialization and even the introduc-tion of railways and protected the old regime because theyrealized capitalist growth and industrialization would reducetheir power and their privileges.

Recent empirical and theo-retical research suggests that the answer is no: there are nocompelling reasons to think that societies will naturally grav-itate toward good institutions. Institutions not only affect theeconomic prospects of nations but are also central to the dis-tribution of income among individuals and groups in soci-ety—in other words, institutions not only affect the size ofthe social pie, but also how it is distributed.

. Contrasting with thispattern, in colonies where therewas little to be extracted, where most of the land was empty,where the disease environment was favorable, Europeans set-tled in large numbers and developed laws and institutions toensure that they themselves were protected, in both theirpolitical and their economic lives. In these colonies, the insti-tutions were therefore much more conducive to investmentand economic growth

The chart here shows a strongnegative relationship between urbanization rates in 1500 andincome per capita today. That is, the former Europeancolonies that are relatively rich today are those that werepoor before the Europeans arrived

The historical evidence shows no evidence of the persis-tence suggested by the geography hypothesis. On the con-trary, there is a remarkable reversal of fortune in economicprosperity. Societies like the Mughals in India and the Aztecsand the Incas in America that were among the richest civi-lizations in 1500 are among the poorer societies of today.

Therefore, if geogra-phy is the key factor determining the economic potential of anarea or a country, the places that were rich before the arrival ofthe Europeans should have remained rich after the coloniza-tion experience and, in fact, should still be rich today. In otherwords, since the key determinant of prosperity remains thesame, we should see a high degree of persistence in economicoutcomes. If, on the other hand, it is institutions that are cen-tral, then those places where good institutions were intro-duced or developed should be richer than those in whichEuropeans introduced or maintained extractive institutions toplunder resources or exploit the non-European population

we need to find a sourceof exogenous variation in institutions—in other words, anatural experiment where institutions change for reasonsunrelated to potential omitted factors (and geographic fac-tors remain constant, as they almost always do)

nce again, however,this correlation between institutions andeconomic development could reflect omit-ted factors or reverse causality

t is true there is a correlationbetween geography and prosperity. Butcorrelation does not prove causation.

Good institutions have three key charac-teristics: enforcement of property rights fora broad cross section of society, so that a variety of individu-als have incentives to invest and take part in economic life;constraints on the actions of elites, politicians, and otherpowerful groups, so that these people cannot expropriate theincomes and investments of others or create a highly unevenplaying field; and some degree of equal opportunity fo

The two main candidates to explain thefundamental causes of differences in pros-perity between countries are geographyand institutions. The geography hypothesis,which has a large following both in thepopular imagination and in academia,maintains that the geography, climate, andecology of a society shape both its technol-ogy and the incentives of its inhabitant

Butthese are only proximate causes of poverty, begging the ques-tion of why these places don’t have better markets, betterhuman capital, more investments, and better machinery andtechnology. There must be some fundamental causes leadingto these outcomes, and via these channels, to dire poverty.

Average per capitaincome in sub-Saharan Africa, for example, is lessthan one-twentieth that in the United States.

I argue that although much of the relationship between development and gender inequality canbe explained by the process of development, society-specific factors are also at play. Manycountries that are poor today have cultural features that exacerbate favoritism toward males.Being poor is insufficient to explain parents’ strong desire to have a son in China and India, forexample

This implies that giving women the right tovote makes a difference. And indeed, Miller(2008) shows that the introduction of womensuffrage in the United States was associatedwith a decline in infant mortality. In mostcountries however, women and men nowhave the right to vote, although they do notget elected. Thus, the next step to increasewomen’s political power is to directly affecttheir representation through quotas or reser-vation policies.

In an experiment, they provided836 married women in Lusaka, Zambia, witha voucher guaranteeing free and immediateaccess to a range of modern contraceptivesthrough a private appointment with a family-planning nurse. Some women received thevoucher in private, while others received thevoucher in the presence of their husbands.This made a huge difference: comparedto cases where husbands were involved,women who were seen alone were 23 per-cent more likely to visit a family-planningnurse, 38

A first dimension of empowerment (thatmentioned by Wolfensohn) is education.Education can have a range of benefits withinthe households: if more educated womencommand higher outside wages, and it is eas-ier for them to get a job, then investing morein educating women, rather than in men,may indeed have a more positive impact onchild health than when the investments arespread evenly if women’s bargaining powerin the household leads to better outcomesfor children.

If poor households are less likelyto spend money on a girl’s illness than on aboy’s illness, then improved access to healthservices, through either health insurance forthe entire family or free medical care for thepoor, would disproportionately help girls,even if parents do not change their behaviortoward them

Even in the countries where the prefer-ence for boys is strongest, the evidence thatgirls systematically receive less care thanboys under normal circumstances is not asclear-cut as one might guess. It is, of course,very difficult to observe whether, for exam-ple, girls are given less to eat than boys, sincehouseholds under observation are likely tochange their behavior. Also, when asked tokeep diaries of how much each member con-sumes over a given period, households maymisreport the portions given to boys andgirls. Fi

This essay arguesthat empowering women does indeedchange society’s choices in important ways,although the usual depiction of womenas always making the best decisions forlong-term development is somewhat exag-gerated. The conclusion here is a more bal-anced, somewhat more pessimistic pictureof the potential for women’s empowermentand economic development to mutuallyreinforce each other than that offered bythe more strident voices on either side ofthe debate.

This paper reviews the evidence on bothsides of the empowerment–developmentrelationship. It first shows that povertyand lack of opportunity breed inequalitybetween men and women, so that wheneconomic development reduces poverty,the condition of women improves on twocounts: first, when poverty is reduced, thecondition of everyone, including women,improves, and second, gender inequalitydeclines as poverty declines, so the condi-tion of women improves more than that ofmen with development.

Women empowerment and economic development are closely related: in one direction,development alone can play a major role in driving down inequality between menand women; in the other direction, empowering women may benefit development.

Communism and large historical migrations (most notablymass European migrations during the 17th to the 20th centuries)are other recent historical episodes that may have had a largeimpact on beliefs about gender. Communist regimes typically im-plemented policies to eliminate gender differences in the economyand promote female participation outside the home. For manyparts of the world, the adoption of the plough may be associatedwith the diffusion of other aspects of Western European societies.In column 5, we include an indicator variable that equals one ifthe country was under a communist regime in the post-WWIIperiod. We also control for the fraction of each country’s popula-tion in 2000 with ancestors from Western Europe.

Since these areas today tend to be richer and more proneto equal gender-role attitudes, this would tend to bias the OLSestimates toward zero

The database provides information on whether societies trad-itionally used the plough. Ethnicities are classified into one of threemutually exclusive categories: (1) the plough was absent, (2) theplough existed at the time the group was observed, but it was notaboriginal, and (3) the plough was aboriginal and found in the so-ciety prior to contact.

Within this framework, Boserup’shypothesis suggests that in societies that engaged in plough agri-culture, cultural beliefs about gender inequality were relativelybeneficial

Nunn and Wantchekon (2011) showthat Africa’s slave trade generated a culture of distrust that con-tinues to persist today. Nisbett and Cohen (1996) and Grosjean(2011a) show that the culture of honor in the US South has itsorigins in a tradition of herding among the Scots-Irish. The find-ings of our paper add to this line of enquiry by providing add-itional evidence that shows that historical factors—namely,differences in traditional farming practices—have shaped theevolution of norms and beliefs about the appropriate role ofwomen in societ

we check that the esti-mates are robust to flexibly controlling for geographic covariates,including overall agricultural suitability, temperature, precipita-tion, soil depth, and slope.

These cultural beliefs tendto persist even if the economy moves out of agriculture, affectingthe participation of women in activities performed outside thehome, such as market employment, entrepreneurship, or partici-pation in politics

Plough cultivation, by contrast,is much more capital intensive, using the plough to prepare thesoil. Unlike the hoe or digging stick, the plough requires signifi-cant upper body strength, grip strength, and bursts of power,which are needed to either pull the plough or control theanimal that pulls it. Because of these requirements, whenplough agriculture is practiced, men have an advantage in farm-ing relative to women.

In some societies, the dominant belief is thatwomen should be allowed to participate freely, and equally tomales, in employment outside the home. In others, there is thevery different view that the appropriate place for women is withinthe home, and they are discouraged from participating in activ-ities outside the domestic sphere.

We test the hypothesis that traditional agricultural practices influenced thehistorical gender division of labor and the evolution of gender norms. We findthat, consistent with existing hypotheses, the descendants of societies that trad-itionally practiced plough agriculture today have less equal gender norms, mea-sured using reported gender-role attitudes and female participation in theworkplace, politics, and entrepreneurial activities

For historical reasons, land is the one asset the poor tend to own. But landrecords in developing countries are often incomplete and many people do not havetitles to their land.

Indeed, one reason why many of the poor respond so well to microcredit is notnecessarily because it offers them credit, but because once you take a loan and buysomething with it, you have a disciplined way to save—namely, by paying down theloan.

These high interest rates seem not to occur directly because of high rates ofdefault, but rather as a result of the high costs of contract enforcement. While delayin repayment of informal loans is frequent, default is actually rare

Temporary migration is rarely documented in surveys, but in the Udaipursurvey, which had questions about this activity, 60 percent of the poorest house-holds report that someone from their family had lived outside for a part of the yearto obtain work. For 58 percent of the families, the head of the household hadmigrated.

Among our 13 countries, food typicallyrepresents from 56 to 78 percent of consumption among rural households, and56 to 74 percent in urban areas.

A common image of the extremely poor is that they have few real choices tomake. Indeed, some people surely work as hard as they can—which may not beparticularly hard, because they are underfed and weak and earn barely enough tocover their basic needs, which they always try to fulfill in the least expensive way.Historically, poverty lines in many countries were originally set to capture thisdefinition of poverty—the budget needed to buy a certain amount of calories, plussome other indispensable purchases (such as housing). A “poor” person wasessentially defined as someone without enough to eat.

The typical extremely poor family tends to be large, at least by the standards oftoday’s high-income countries. The number of family members varies betweenabout six and twelve, with a median value (across the different countries) ofbetween seven and eight, compared to 2.5 in the 2000 U.S. census, for example

The way in which we identify the poor does raise questions. Purchasing powerparity exchange rates, which are essential to compute a “uniform” poverty line, havebeen criticized as inadequate, infrequently updated, and inapplicable to the con-sumption of the extremely poor (Deaton, 2004, 2006). Prices are typically higher inurban than in rural areas, and even in rural areas, the poor may pay different pricesthan everyone else. Also, reporting periods vary significantly from survey to surveyand this has been shown to affect systematically what people report

From each of these surveys we identified the extremely poor as those living inhouseholds where the consumption per capita is less than $1.08 per person per day,as well as the merely “poor” defined as those who live under $2.16 a day using 1993purchasing power parity (PPP) as benchmark.

Building soundinvisible infrastructure will require working with individuals within the state whohave the power and incentives to implement reform, and at the same time ensuringthat poor citizens remain the ultimate principal

Given the myriad vested interests in any society, governments—especially inthe presence of resource constraints—will often only respond to clear demandsfrom citizens. The poor are more likely to be sidelined both in economic develop-ment and in democratic processes.

The head of Rockefeller’s programand governmental partners centralized control, worked through the public schoolsystem, and broadened the set of implementing actors to include schoolteachers,community leaders, and priests. Palmer (2003) argues that this use of the publicschool network was central to the program’s success: by the end of Rockefeller Foun-dation’s operations, more of the foundation’s funds were being channeled throughthe country’s School Health Department, rather than directly, to hookworm relief.The campaign contributed to centralizing and modernizing health care in CostaRica and paved the way for the foundation of a national health ministry in 1927, thethird in Latin America.

However, vertical public health programs may not contribute to the strength-ening of domestic “horizontal” primary healthcare systems (Oliveira-Cruz, Kurowski,and Mills 2003). In some cases, vertical programs, or more generally, deliveringsocial services through a cadre of nonstate actors, may even weaken public servicedelivery by diverting civil servants, funding, and political interest away from statestructures and into parallel systems.

Now consider foreign aid, or official development assistance. The total volumeof aid has increased substantially over time, rising nearly fivefold between 1960 and2016, from about $32 billion to $158 billion in 2016—both in constant 2016 US dollars(OECD 2018).

n sum, while private players can often play a substantive role in providing invis-ible infrastructure, ensuring that those services reach the poor will require that thestate remains as an active regulator

But other components—such as access to health care, education, and financialservices—are often provided by a range of players

North (1990) famously characterized institu-tions as “humanly devised constraints that structure political, economic and socialinteractions,” setting the rules of the game for an economy. He argued that well-func-tioning institutions enable prosperity by ensuring secure property rights, facilitatingcomplex contractual arrangements, aligning norms to allow markets to functioneffectively, and so on.

But poverty can have a long half-lifein the presence of inequality. In India, which in 2013 contained the largest share ofthe world’s extreme poor, over 100 billionaires lived alongside 210.4 million peoplein extreme poverty in 2013. This imbalance arises from unequal growth: India’s top10 percent of incomes captured 66 percent of growth between 1980 and 2016, whilethe bottom 50 percent captured only 11 percent (Alvaredo et al. 2018)

Ultimately, it is democracy, done right, that bestallows citizens to demand what they need to end poverty

A key part of the task at hand is to ensure that aid policies strengthen domesticinstitutions rather than undermine them.

In advocating for investment in invisible infrastructure, we emphasize that thedomestic state is the inevitable regulator, if not always the provider, of these servicesand institutions for the poor. First, the state is the only body with the mandate toprovide certain critical institutions, like property rights and a monopoly of violence.Second, even where for-profit businesses and nongovernmental organizations are best-placed to provide specific services, such as micro-credit, the state alone can regulatethe provision of these services to the poor.

Continued growth may ultimately lift up thoseremaining millions, but it may do so much more slowly than is necessary. Endingpoverty by 2030 in this second group of countries will require not just growthof the economy, but redistribution of new domestic resources to the poorest.Such redistribution must come in the form of services and institutions that thepoor need for economic mobility. Because these countries receive relatively littleforeign aid, domestic states will bear most of the responsibility for providing theseservices to the poor.

Low-incomefragile countries are often trapped in cycles of erratic growth and misdirected aid,while high-poverty middle-income countries typify a global trend of falling cross-country inequality accompanied by greater within-country inequality (Hammar andWaldenström 2017)

The international development community should recalibrate the allo-cation of resources to increase accountability and state capacity

Historically, the quest to reduce poverty has relied on two levers: economicgrowth (the idea that “a rising tide lifts all boats”) and the intentional redistributionof resources to the poor, either by the domestic state or foreign aid. I

he share of the world’s population living below the global extreme povertyline ($1.90 in consumption per day) has plunged dramatically in recentdecades, from 42 percent in 1981 to 11 percent in 2013

It is, therefore,particularly important to build into the concept of development the possibility ofpersistent incompleteness in ranking. Seeing the agreed ranking as the intersec-tion of the partly divergent valuation functions must, of necessity, entail this.

The dependence of the concept of developmenton evaluation becomes a problem to the extent that (1) the valuation functionsaccepted by different people differ from each other, and (2) the process of changeinvolved in development alters the valuations of the people involved. These twoproblems may be called respectively "value-heterogeneity" and "value-endogene-ity

The important thing to note here is that the problem of evaluation neednot be seen in an all-or-nothing way. It is possible to extend the partial order bynarrowing the ranges of weights, and how far one can go on the basis ofagreement on evaluation will depend contingently on the nature of the exercise inquestion

One way of introducing the importanceof freedom in the determination of well-being is to see well-being as a functionnot only of the actual functioning achievement, but also of the capability setfrom which that n-tuple of functionings is chosen. In this way of formallycharacterizing the problem, the list of functionings need not include "choosing"as such, but the value of choosing will be reflected in the evaluation by makingthat evaluation depend both on the chosen n-tuple of functionings, and on thenature and the range of the capability set itsel

One of the functionings that may be thought to be particularly important inassessing the nature of development is the freedom to choose.

The achievement of functionings depends not only on the commodities ownedb y the person in question, but also on the availability of public goods, and thepossibility of using private goods freely provided by the state. Such achievementsas being healthy, being well-nourished, being literate, etc. would depend naturallyalso on the public provisions of health services, medical facilities, educationalarrangements, and so on.

Insofar as development is concerned with the achievement of a better life, thefocus of development analysis has to include the nature of the life that peoplesucceed in living. This incorporates, of course, the length of the life itself, andthus life expectancy data have an immediate relevance to the living standard andthrough that to the concept of development, But the nature of the life that peoplesucceed in living in each period is also a matter of importance.

A second source of difference between growth and development relates to thequestion of externality and non-marketability. The G N P captures only thosemeans of well-being that happen to be transacted in the market, and this leavesout benefits and costs that do not have a price-tag attached to them.

•Life expectancy is, of course, a very limited measure of what has been called" t h e quality of life". Indeed, in terms of what it directly measures, life ex-pectancy is more an index of the "quantity" of life rather than of its quality.

To illustratethe problem, figures for GNP per head and life expectanc

The close link between economic development and economic growth is simulta-neously a matter of importance as well as a source of considerable confusion.There can scarcely be any doubt that, given other things, an expansion ofopulence must make a contribution to the living conditions of the people inquestion. I

It is, however, arguable that for one reason or another, agood deal of standard economics has tended to move away from broad issues ofpoverty, misery and well-being, and from the fulfilment of basic needs andenhancing the quality of life. Development economists have felt it necessary toemphasize and justify their involvement with these-rather "old-fashioned"-problems, even though the relevance of these problems is by no means confinedto development economics.

The fact that in the early writings in economics there was this noticeablecongruence of development economics and economics in general is a matter ofsome interest, especially in the context of investigating the nature of "the conceptof development". Interest in development problems has, traditionally, providedone of the deepest motivations for the pursuit of economics in general, and thisbroad basis of development economics has to be borne in mind when investiga

Sir William Petty in 1676.

The company changed the inscription on its watches to read, “Built in Detroit” followed by thephrase “Swiss and imported parts” to comply with the ruling. Said Kartsotis, “Shinola will alwaysstrive to do as much as it can in America with the benefit of an American workforce. We would like toassure our customers that our high quality products will continue to be as American as possible whileremaining a competitive alternative to products that are not supporting the same level of investmentback into the U.S.

Contemporary artist Rebekah Modrak created an art piece in protest of the company called“RETHINK SHINOLA” (see Exhibit 7) in which she questioned the motivations for and extent ofShinola’s commitment to Detroit and to its workers, engaging politically-charged issues of commerce,consumerism, labor exploitation, white privilege, power, and racism. Within Shinola’s brand narrativesabout the company’s role in Detroit’s economic revitalization, she revealed power dynamics fueled byrace and patriarchy and the exploitation of Shinola’s employees and Detroit’s misfortunes for thecompany’s gain. She explained, “Shinola sells themselves as the intrepid white savior civilizing the‘wild’ environment (their words)

The Detroit mythos didn’t always transfer to international locales. Said a resident ofStockholm, “For me, [Detroit] doesn’t really mean anything because I don’t connect clocks with Detroitreally. If it was a car, maybe.”116 When other Stockholm residents were asked what they associatedwith Detroit, they replied, “the car industry” or “bankruptcy.”1

Fun. This new collectionfrom Shinola arrives at the intersection of quality craftsmanship and youthful design, featuringwatches that are built with high-grade components, clever engineering and vibrant colorways.

A new sub-branded timepiece line called Detrola (a portmanteau of Detroit and Shinola) waslaunched in 2019 with a manufacturer’s suggested retail price of $395, about $100 cheaper than thebrand’s flagship watches.102 The colorful Detrola watches were named after aspects of Detroit’s sportsand cultural scene, such as the “The Wolverine,” representing the mascot of the University of Michigan,and “The Detroit Tiger,” representing the city’s professional Major League Baseball team.

Knowledge, rather than money, was the cultural currencyconsumers needed to access these brands, as one needed to be “in the know” to recognize theseinconspicuous brands and to know the story behind their making.

Shinola was an example of what pundits were calling “new luxury”.88 New luxury brands tendedto be handcrafted, artisanal, slow, and analog like their traditional luxury brethren, such as brands likeHermès or other established luxury brands like those in the LVMH portfolio, but came withouthundreds of years of heritage and tradition and often without inaccessible price points

“It’s a very exciting time for the brand to be launching a campaign with the message‘Let’s roll up our sleeves

What’s brilliant about the ad is how perfectly it captures this moment of Americanzeitgeist and aligns it with the brand. At a political level, our national conversationis focused on the fate of the American working class. As last week’s New Hampshireprimary results showed, it doesn’t matter which side of the political spectrum you’reon. There was record-breaking turnout at the polls—and each of the winners, both[Republican Donald] Trump and [Independent Bernie] Sanders, portray themselvesas champions of the little guy; as outsiders who are ready to fight the establishment;and...as advocates for a centuries-old tradition of American exceptionalism

The company’s marketing campaigns included branded videos, print ads in The New York Times,content marketing partnerships with magazines, and direct mail catalogs. Digital marketing used of allof this content and the company employed geo-location targeting to deliver it to people while the

However, for craft products, such as fine wines or mechanical watches,value has historically been embedded in the making activity, and relies on traditional knowledge andspecialized skill.60 These industries often remain embedded locally as they draw from a quicklydisappearing pool of resources (skilled craftspeople, supporting institutions, raw materials, andspecialist suppliers). Industries such as Swiss watches, champagne, cognac, and Italian leather goodsdraw an aura of authenticity from continued connection to their place of making and from their highlyspecialized and meticulous makers

Artisanal craft’s renaissance was driven by the enjoyment and meaning humans gained from hardwork and meticulous production. As the contemporary economy emphasized newness and thereby adisposal of the old, craft objects increased in attractiveness because of their perceived authenticity. Asa result, makers gained social standing, particularly in economies where machine-led manufacturinghad declined.58 Many brands, Shinola included, communicated a connection to the past and a sense ofcontinued tradition through references to real practices and references to the spirit of making and craft.

In its brand narratives,the company stressed that an average of 100 staff (approximately 20% of its employees) worked inmanufacturing in the firm’s in-house factory, on the fifth floor of the historic Argonaut Building.52Throughout its retail stores and on its e-commerce website, images of Shinola craftspeople areaccompanied by descriptions about their particular skills and a count of the number of people involvedin the production of any one item (e.g., we learn 30 people are involved in the production of a singlewatch)

By 2015, the company’s sales were evenly split among its tencompany-owned stores, its wholesale distribution partnerships with retailers, and its e-commercewebsite, each of which was responsible for about one-third of its revenues.

Kartsotis acquired the rights to use the brand, paired it with alightning bolt logo, which the company claimed signified “the beauty of industry, our respect for craft,and a commitment to quality,”43 (see Exhibit 1) and added shoe polish to his product line forauthenticity, despite the fact that the new Shinola did not sell shoes

Detroit was initially chosen as the place to locate because the initial strategy wassimply to create 100 jobs in a place that really needed them...It was both a philanthropicimpulse and a desire to experience the sport of trying to create a business model thatwas...unlikely to return the fruits that doing something less complex and more provenwould have brought us financiall

After analyzing the results of consumerresearch on country-of-origin effects that indicated that consumers were willing to spend $5 on a penmade in China, $10 on the same pen made in the U.S., and $15 for one made in Detroit,

Post-bankruptcy, a narrative of revitalization emerged, with claims that Detroit was now the“Comeback City” and a hub for startups.26 The city’s revitalization ushered in the arrival ofindependent coffee shops, public art installations, new hotels, and start-ups.27 Furthermore, public-private partnerships built a new tramline, stadium, and revitalized museums.28 The city today isregularly balancing its budget; the exodus of residents from the city has slowed, and some areas areshowing signs of growth.29 However, assessments of this revitalization are mixed, with manysuggesting they reflect Detroit’s enduring racial divides. The creative sector is overwhelmingly white,30and 70% of the regeneration activities are carried out by whites in a few select neighborhoods whichaccount for just 7.2 (out of 137) square miles within the city limits and are those with the lowest levelsof poverty,31 leaving many neighborhoods abandoned and impoverished.

Detroit had the lowest fiscal capacity but the greatest community need amongthe nation’s largest 50 cities.20 Things were tough, as a writer at Surface described, “The enfeebled autoindustry, hemorrhaging profits and furloughing employees, faced a long post-bailout road back.”21

Billboard Hot 100 charts duringthe 1960’s.12 Diana Ross and the Supremes, Stevie Wonder, Marvin Gaye, the Four Tops, and TheJackson 5 became some of the label’s earliest and most successful artist

This new identity was a far cry from Detroit’s heyday. During its “first globalization” period of1870-1914, Detroit became an industrial powerhouse and an engine of modernity, driven by theemergence of mass production.

As theycontinued to expand the Shinola brand into international markets, the brand’s story, rooted in Detroit’simagery, history, and lore, might have to be reconsidered to appeal to non-American audiences

By 2018, 35-45% of the company’s sales were coming from the state of Michigan,6where the company maintained six stores. With aggressive goals to grow to 75-100 stores,7 consumersand store employees, many of whom had never been to Detroit, would need to become familiar withthe Shinola brand narrative.

he yearhad been a difficult one for the company and its former CEO, Tom Lewand, had been forced to lay off30 members of its workforce in July.3 The 500+ remaining employees supported revenues that werereported to be north of $100 million (as of 2015)4 that came from sales in Shinola’s thirty-threecompany-owned stores and numerous wholesale accounts in the U.S. and abroad, includingBloomingdale’s, Nordstrom’s, Saks 5th Avenue, and other upscale department stores and specialtyjewelers.

It is a city filled with the rich history of the industrial age, the pride of American manufacturing,and of the soulful sounds of Motown music. It is also a place that more recently has becomesynonymous with rustbelt urban decay, characterized by abandoned factories, commercial buildings,and homes, soaring unemployment and homelessness, and a sense of despair since the city’sdeclaration of bankruptcy in 2013, which led to its other nickname, “America’s Warzone”.2 It was asurprising place to find the headquarters of a new luxury brand.

is isthe city that made this country, with its steel, with its skill, with its labor. Reestablishing trades that haven’t beenseen in this country in a generation or more, it’s a tall order to return to full, but it’s what hundreds of employeeswork toward every day here in Detroit

As soon as it relaunched, Basu kept track of Monginis’s performance. People were enthusiastic aboutMonginis and soon, its revenue was close to what Mio Amore stores were making in similar localities.There was considerable press coverage of the return of Monginis.

Consumers were initially apprehensive, believing that a foreign company had taken over Monginis. Salesslowly grew by approximately 2.5 per cent without any price increase. Customer perceptions of the brandchange revealed that the change was welcome, but many expressed higher expectations from Mio Amore.23Like its competitors, Mio Amore operated on wafer-thin margins. Thus, a 3–4 per cent growth in businessmade a big difference. Basu said, “It was better to stay grounded so that there were more people togetherand there was no fear of being docked off from the top” (see Exhibits 6, 7, 8, and 9).

Basu believed that if the quality was retained, dropping the Monginis brand would have little effect on thebusiness. Nevertheless, the market surveys suggested otherwise—that is, Monginis had a very high consumerpresence and recall and, thus, any change would likely impact these perceptions. To find a new brand name,SFPL worked with its marketing communications consultant, Mayurakshi, which offered five suggestions:Fresh Bake, MonerGinis, Mio Amore, Goldilocks, and I Bake. These names were put to focus groupdiscussions to gain a better understanding of how consumers perceived them. Mio Amore and Goldilockswere serious contenders. Further, a sample survey of 600 respondents found that Mio Amore was associatedwith gifts, perfume, birthday cake, processed packaged food, and confectionery savouries. Basu commented:

Monginis was positioned as an economy brand, providing value for money and good-quality bakeryproducts. Khorakiwala, MFPL’s chairman and managing director, said, “We offer good value for moneyand score on quality. We are a middle-class bakery serving first-class products and we, the owners, consumeand serve our own products at internal functions and meetings.”

SFPL preferred to work with dealers who owned their property rather than those who rented. Basu said, “Aproperty owner had a stake in making the business work unlike a tenant who could easily close shop andmove out of the area.” The average store size was 150–250 square feet (14–23 square metres). The storeswere air-conditioned, clean, attractively furnished, and well lit, the first of their kind in the region. SFPLadopted a new tagline—“Your friendly neighbourhood cake shop.”

Basu preferred neighbourhood shopping areas(i.e., bazaars) for Monginis dealers because such shop owners generally built personal connections withtheir regular customers. Basu said, “In these bazaars, the quality of the product and the reputation of theperson selling it were as important as the store ambience or other aspects of presentation.”

The SFPL philosophy in terms of manufacturing for its own brand was essentiallyto keep the margins low when the products were meant for the masses. Prices were kept the same, and thequality of the raw materials was not compromised. No information-sharing took place among the variousclient companies through SFPL, in line with Basu’s high ethical standards.

In addition to its manufacturing facilities indifferent states in eastern India, the group also had 70 refrigerated vans outfitted with the Global PositioningSystem (GPS) in Kolkata to supply freshly baked short-shelf-life Mio Amore bakery items. It was one ofthe largest fleets of refrigerated vans owned by any firm in eastern India. Basu said

The company also invested in state-of-the-art equipment, which had been fabricated locally under Basu’ssupervision. This equipment functioned as efficiently as imported equipment but was made at a much lowercost. As SFPL experienced steep growth, it opened two more factories in the suburbs of Kolkata to produceindustrial bakery products. At the same time, the company also opened a factory in Siliguri for its retailproducts, to capitalize on the government’s concession to entrepreneurs who set up factories in the northernpart of West Bengal

SFPL sold its products through dealer stores. After the agreement with MFPL, SFPL focused ondealership agreements with local stores to sell Monginis products, especially the fresh and short-shelf-lifeproducts. For long-shelf-life packaged products, SFPL rolled out a retail distribution network, as SFPL wasfirst of its kind in the territory and did not have a distribution network in place. Other bakery chains includedMio Amore, Monginis, Kathleen Confectioners, and The Sugarr & Spice, which catered to the masses; semi-premium bakeries such as Cakes, The French Loaf, and Kookie Jar; and premium bakeries such as TheBelgian Bakery and Flurys Bakery. The market share of the three types of bakeries was approximately 85 percent for the bakery chains, 10 per cent for semi-premium bakeries, and 5 per cent for premium bakeries

Kookie Jar catered to the elite of Kolkata. The clientele included industrialistssuch as the Goenkas, the Mittals, the Ambanis, and the Dabur Burmans,18 and Satyajit Ray, the greatfilmmaker. The sisters believed in exclusivity, which they practised by visiting restaurants and patisseriesabroad, namely in Paris, Austria, and Vienna. Kookie Jar followed international quality standards byprocuring imported Belgian dark chocolate, food colours, and decorative glitters; even the ribbon aroundthe cake box was often imported.

t offered cakes for birthdays, farewells, and wedding anniversaries, and wasone of the first to introduce cakes customized to various shapes and cartoon characters. Some of its storeswere big enough to accommodate customers who could sit and eat in the store while other stores had limitedspace that was only for standing and eating.