Private Tutoring-A Widespread Phenomenon
Private Tutoring-A Widespread Phenomenon
Private Tutoring-A Widespread Phenomenon
tutoring, and then analyzes the associated equity and efficiency issues, with a focus on developing
countries. This paper begins by establishing the importance of private tutoring as an economic
phenomenon in much of the world, developed and developing. It then provides a simple graphical
framework of supply and demand for education with private tutoring, which provides theoretical
guidance to the discussion of equity and efficiency issues later on. The next sections review (i) the
determinants of private tutoring, from both a micro and a macro perspective, and (ii) the effects of
tutoring on student achievement, with special attention to the statistical problems with identifying
these effects, and to recent studies that have confronted those problems. Finally, we use those results
and the theoretical framework to discuss the efficiency and equity implications of private tutoring, and
to explore their implications for policy toward tutoring, as well as to identify the areas that may
demand more research.
2
Japan has been a pioneer in the provision of this type of supplementary education. Private
tutoring has long been a huge commercial industry in Japan, with annual revenues reaching an
estimated US$14 billion by the mid-1990s. Nine private tutoring schools were already listed on the
Japanese stock exchange at that time, and the tutoring sector had become a “crucial component of
Japanese education” (Russell, 1997). Many students use school vacations, including the important
New Year’s holiday, for intensive tutoring programs. To stimulate “school” spirit, several private
tutoring schools—or juku—have even had their students wear a kind of white headbands once worn
in battle by samurai warriors (Rohlen, 1980). The proportion of college students who have spent
additional years after high school graduation to cram for college entrance examinations, often at
specialized private tutoring classes, averages about 30 percent. For those who end up enrolling at the
most highly ranked schools, the share may exceed 60 percent (Ono, 2007). 4
But recent research has made it clear that Japan is not unique in supporting a large and vibrant
private tutoring industry. Tutoring is now widespread in many parts of the world, including the
developing countries on which we focus in this paper. Table 1, which is largely based on Table II.1 in
Dang (2007a), shows evidence on the extent of private tutoring in selected countries with a focus on
developing countries. Although the studies cited there vary somewhat in methodologies and
populations surveyed, certain patterns are clear.
First, while the incidence appears to be highest in East Asian countries, private tutoring is now an
important phenomenon in many countries of different size, level of economic development, political
institutions, or geographical locations. In some cases, spending on private tutoring approaches the
level of spending on the formal public system. In Korea, households spent 2.9 percent of GDP on
private tutoring in 1998, almost equaling the 3.4 percent of GDP spent on education by the public
sector (Kim and Lee, 2004). A similar situation happens in Turkey, where the corresponding figures
are 1.44 percent of GDP for private tutoring and 2 percent for public education expenditures (Tansel
and Bircan, 2006).
Second, private tutoring is an important phenomenon not only for upper-secondary students
preparing for university exams, but also for students at the primary and lower secondary levels, and
sometimes (as in Japan) even among upper-secondary graduates.
Third, the private tutoring industry appears to be growing in many countries, both in absolute
terms and relative to the formal education sector. Table 1 includes evidence of growth in terms of the
percentage of students taking tutoring (in Kenya and Mauritius) and the number of private tutoring
firms catering to them (in Turkey and Canada). 5 Evidence on tutoring expenditures, where it is
available, also supports the notion that the sector is growing, as in Korea, where household spending
as a share of GDP on private tutoring rose continuously from 0.7 percent of GDP in 1977 to 1.2 in
1990 and 2.9 in 1998. Finally, anecdotal reports also suggest an expansion in tutoring elsewhere; for
3
example, in both low-income countries such as Vietnam (Dang, 2007a) and high-income countries
such as the US (Fuchs, 2002; Borja, 2005), some households have reportedly begun sending their
children to private tutoring to give them an edge as early as preschool.
4
same household can consume a larger amount of education at Q*2 since the supply curve of education
in this case is not constrained by the vertical part rising at point A.
This setup, which we refer to as the “standard framework”, underlies the discussion below on the
determinants and welfare consequences of tutoring. The framework incorporates certain assumptions
that will not always be valid, and so we return to those below. One assumption is that the market for
private tutoring is competitive, and indeed that households are allowed to choose whether to purchase
tutoring services. A second is that public schooling reaches a strict capacity constraint after a certain
point, which is likely a better description of the short run than of the long run. In section 6, we will
explore how relaxing these assumptions would change our analytical and policy conclusions.