AbstractPrivate <strong>tutor<strong>in</strong>g</strong> <strong>is</strong> be<strong>in</strong>g practiced at an alarm<strong>in</strong>g scale <strong>in</strong> Egypt <strong>and</strong> <strong>in</strong> many o<strong>the</strong>rdevelop<strong>in</strong>g countries. None<strong>the</strong>less, <strong>the</strong> literature on <strong>tutor<strong>in</strong>g</strong> <strong>is</strong> still scant. The purpose of th<strong>is</strong>paper <strong>is</strong> to ga<strong>in</strong> an underst<strong>and</strong><strong>in</strong>g of <strong>the</strong> nature <strong>and</strong> determ<strong>in</strong>ants of <strong>tutor<strong>in</strong>g</strong> <strong>in</strong> Egypt, us<strong>in</strong>gmicro-level data, <strong>in</strong> order to <strong>in</strong>vestigate whe<strong>the</strong>r <strong>gender</strong> bias ex<strong>is</strong>ts <strong>in</strong> <strong>tutor<strong>in</strong>g</strong> dec<strong>is</strong>ions. It <strong>is</strong>expected that s<strong>in</strong>ce <strong>gender</strong> d<strong>is</strong>parities are present <strong>in</strong> educational <strong>in</strong>vestments <strong>in</strong> general, <strong>the</strong>ywould be more pronounced <strong>in</strong> optional educational <strong>in</strong>vestments like that of receiv<strong>in</strong>g<strong>tutor<strong>in</strong>g</strong>. It <strong>is</strong> also expected that s<strong>in</strong>ce labor market outcomes are more favorable to boys,parents would be less will<strong>in</strong>g to spend on <strong>tutor<strong>in</strong>g</strong> for girls. Surpr<strong>is</strong><strong>in</strong>gly, however, no <strong>gender</strong>bias aga<strong>in</strong>st girls was detected with respect to <strong>tutor<strong>in</strong>g</strong>. The absence of bias <strong>is</strong> <strong>in</strong> itself asignificant <strong>and</strong> puzzl<strong>in</strong>g f<strong>in</strong>d<strong>in</strong>g. We conclude that <strong>the</strong> education premium <strong>in</strong> <strong>the</strong> marriagemarket may be <strong>the</strong> answer to <strong>the</strong> puzzle.ملخصتتفشى ظاهرة الدروس الخصوصية بصورة خطيرة في مصر و العديد من الدول النامية. و مع ذلك فان آل ماآتب عنها مازال ضئيلا للغاية. يهدف هذا البحث إلى فهم طبيعة و عوامل الدروس الخصوصية في مصر استناداإلى بيانات دقيقة لمعرفة ما إذا آان يوجد انحياز للنوع في الدروس الخصوصية أم لا. و بما أن التمييز على أساسالنوع موجود بالفعل في الاستثمارات التعليمية بصفة عامة فانه من المتوقع أن يكون اآثر وضوحا فيالاستثمارات التعليمية الاختيارية مثل تلقى الدروس الخصوصية. وبما أن متطلبات سوق العمل في صالح الذآورفمن المتوقع ألا ينفق أولياء الأمور على الدروس الخصوصية للإناث. والمثير للدهشة انه لم يتم رصد أي انحيازعلى أساس النوع بخصوص الدروس الخصوصية. و عدم وجود هذا الانحياز يمثل في حد ذاته نتيجة مهمةومحيرة. و نستنتج أن أهمية التعليم في سوق الزواج ربما يكون الإجابة لهذا اللغز.1
1. IntroductionAlthough technically illegal, <strong>private</strong> <strong>tutor<strong>in</strong>g</strong> has become <strong>in</strong>creas<strong>in</strong>gly widespread <strong>in</strong> Egypt<strong>and</strong> <strong>is</strong> no longer limited to diploma years or to students attend<strong>in</strong>g public schools. Despiterecognition of how widespread <strong>tutor<strong>in</strong>g</strong> has become, <strong>the</strong> phenomenon has not yet beenformally studied. There <strong>is</strong> anecdotal evidence that <strong>private</strong> lessons currently constitute aconsiderable part of households’ expenditure. Th<strong>is</strong> implies that <strong>the</strong> estimation of <strong>the</strong> rates ofreturn on education should be rev<strong>is</strong>ed to <strong>in</strong>corporate expenditure on <strong>private</strong> lessons. Inaddition, <strong>tutor<strong>in</strong>g</strong> potentially exacerbates educational (<strong>and</strong> <strong>the</strong>reby <strong>in</strong>come) gaps acrossdifferent <strong>in</strong>come as well as <strong>gender</strong> <strong>group</strong>s. Accord<strong>in</strong>gly, exam<strong>in</strong><strong>in</strong>g <strong>tutor<strong>in</strong>g</strong> determ<strong>in</strong>ants hasimportant policy implications.The purpose of th<strong>is</strong> paper <strong>is</strong> to ga<strong>in</strong> an underst<strong>and</strong><strong>in</strong>g of <strong>the</strong> nature <strong>and</strong> determ<strong>in</strong>ants of both<strong>private</strong> <strong>and</strong> <strong>group</strong> <strong>tutor<strong>in</strong>g</strong> <strong>in</strong> Egypt <strong>in</strong> order to <strong>in</strong>vestigate whe<strong>the</strong>r <strong>gender</strong> bias ex<strong>is</strong>ts <strong>in</strong><strong>tutor<strong>in</strong>g</strong> dec<strong>is</strong>ions, <strong>in</strong> particular with respect to who takes <strong>and</strong> how much to spend on <strong>private</strong>lessons. Gender gaps <strong>in</strong> <strong>tutor<strong>in</strong>g</strong> can possibly reflect general <strong>gender</strong> gaps <strong>in</strong> education. It <strong>is</strong>expected that if d<strong>is</strong>parities are present <strong>in</strong> educational <strong>in</strong>vestments <strong>in</strong> general, it would even bemore pronounced <strong>in</strong> more optional educational <strong>in</strong>vestments like that of receiv<strong>in</strong>g <strong>tutor<strong>in</strong>g</strong>.Women’s education <strong>is</strong> crucial to <strong>the</strong> development of <strong>the</strong> MENA (Middle East <strong>and</strong> NorthAfrica) region. Over <strong>and</strong> above its <strong>in</strong>tr<strong>in</strong>sic value, school<strong>in</strong>g for girls has significant benefitsto society. Education for mo<strong>the</strong>rs has positive effects on child survival <strong>and</strong> child health. Inaddition, <strong>the</strong> l<strong>in</strong>k of women’s education to lower birth rate <strong>and</strong> better maternal health <strong>is</strong> wellestabl<strong>is</strong>hed. Educated mo<strong>the</strong>rs also tend to emphasize <strong>the</strong> education of <strong>the</strong>ir children,especially <strong>the</strong>ir daughters. Equally important, <strong>is</strong> that education enhances women’s ability to<strong>in</strong>fluence dec<strong>is</strong>ion-mak<strong>in</strong>g at <strong>the</strong> household level s<strong>in</strong>ce <strong>in</strong>creased participation for women <strong>in</strong><strong>the</strong> labor force means <strong>in</strong>creased earn<strong>in</strong>g capacity. Education also contributes significantly towomen’s ability to exerc<strong>is</strong>e <strong>the</strong>ir political rights.Despite <strong>the</strong>se benefits, <strong>gender</strong> d<strong>is</strong>parities <strong>in</strong> education 1 have strongly pers<strong>is</strong>ted <strong>in</strong> MENAcountries <strong>and</strong> <strong>in</strong> develop<strong>in</strong>g countries <strong>in</strong> general. 2 Accord<strong>in</strong>g to <strong>the</strong> World DevelopmentReport (1996), <strong>the</strong> MENA region exhibits <strong>the</strong> highest <strong>gender</strong> gap <strong>in</strong> education, after SouthAsia. A considerable body of research has explored <strong>gender</strong> bias <strong>in</strong> school<strong>in</strong>g <strong>in</strong> South Asiancountries, but <strong>the</strong>re has been a shortage <strong>in</strong> empirical research on women’s education <strong>in</strong> <strong>the</strong>case of MENA countries. Th<strong>is</strong> <strong>is</strong> not surpr<strong>is</strong><strong>in</strong>g given <strong>the</strong> scarcity <strong>and</strong> <strong>in</strong>accessibility ofmicro-level data <strong>in</strong> <strong>the</strong>se countries. Ex<strong>is</strong>t<strong>in</strong>g literature usually documents <strong>gender</strong> bias by onlylook<strong>in</strong>g at aggregate level measures of education. For example, studies tend to compareliteracy rates, enrollment rates, <strong>and</strong> mean years of school<strong>in</strong>g by <strong>gender</strong> (for example, NagatEl-Sanabary, 1993). Aggregate level data, however, does not help expla<strong>in</strong> how <strong>in</strong>dividual,family, <strong>and</strong> community factors affect <strong>the</strong> extent of <strong>the</strong> bias <strong>and</strong>, hence, <strong>is</strong> not sufficient to <strong>the</strong>formulation of effective education policies.In <strong>the</strong> case of Egypt, female literacy <strong>and</strong> enrollment rates still lag beh<strong>in</strong>d that of males. For<strong>in</strong>stance, based on Egypt Labor Market Survey of 1998 (ELMS 98), <strong>the</strong> number of girls thathave never been to school <strong>is</strong> 2.3 times more than boys. However, once girls are sent toschool, <strong>the</strong>re <strong>is</strong> no significant <strong>gender</strong> d<strong>is</strong>parity observed <strong>in</strong> dropout rates.1Gender bias <strong>in</strong> o<strong>the</strong>r forms of human capital <strong>in</strong>vestments (e.g., bias <strong>in</strong> mortality rates, <strong>in</strong> health <strong>in</strong>vestment, <strong>and</strong><strong>in</strong> nutrition) <strong>is</strong> also well documented <strong>in</strong> develop<strong>in</strong>g countries. For <strong>in</strong>stance, Sen (1990) shows that females face aconsiderably higher mortality rate compared to males <strong>in</strong> Asia <strong>and</strong> North Africa.. Behrman (1988), on <strong>the</strong> o<strong>the</strong>rh<strong>and</strong>, found that <strong>the</strong>re <strong>is</strong> a pro-male allocation of nutrients <strong>in</strong> rural India.2In develop<strong>in</strong>g countries, <strong>gender</strong> d<strong>is</strong>parities <strong>in</strong> education are present <strong>in</strong> literacy rates, enrolment rates, <strong>and</strong>atta<strong>in</strong>ment levels. For <strong>in</strong>stance, <strong>in</strong> develop<strong>in</strong>g countries, exclud<strong>in</strong>g India <strong>and</strong> Ch<strong>in</strong>a, <strong>the</strong> percentage of boysenrolled was higher than that of girls by 22 <strong>and</strong> 43 percentage po<strong>in</strong>ts for primary <strong>and</strong> secondary school-age kidsrespectively (World Bank 1996).2