total income produced within the frontiers of a nation rather than of the income and output
available for use to the nation. Real GDP per capita is a less desirable measure to use if one is
interested in a surrogate welfare measure for the broad range of development goals of nations.
Real GDP per capita does give information on how output is changing in an economy, but it
does so irrespective of who ultimately receives the income earned from such production. The
GDP per capita measure, then, is not as closely connected to what remains in the hands of the
residents of the nation for current and future consumption as the GNI per capita measure, and
thus GDP per person is a more imperfect measure of a nation‚Ä™s overall well-being.
Which measure is employed will be determined by the use to which the income criterion is
to be put. If one is solely interested in the pace of economic growth and total production for a
country, then the real GDP per capita measure will serve quite nicely. If, however, one wishes
to use the income proxy that best measures what is available for use by a country‚Ä™s residents
and which can concretely contribute to their level of well-being, it makes more sense to use
the real GNI per capita measure as the surrogate yardstick.
Having said all this, a glance back at Table 2.1 shows that for most of the economies
shown in the table, the GDP/GNI gap is not that large. Thus in most cases, using GDP or
GNI per capita will make little substantive difference in evaluating the level of development
using the income criterion. There are, however, exceptions, not shown in the table, such as
Angola, Brunei, Puerto Rico, Republic of the Congo, and Gabon, where the gap is such that
GNI is anywhere from 20 percent to 40 percent smaller than the value of GDP (based on
data in the Penn World Tables). For such countries, using real GDP per capita will be a less
reliable measure for approximating the level of development of those countries than the GNI
per capita measure.
International comparisons of income: purchasing power parity (PPP)
There is a further issue to consider when using income levels as a basis of comparison and
as proxy measures of the level of development of different economies. What exactly does
a comparison of Cambodia‚Ä™s 2006 GNI per capita of $481 with Chile‚Ä™s average income of
$7,892 mean? Of Malaysia‚Ä™s GNI per capita of $5,596 with Mexico‚Ä™s $7,970 per capita
income? Is it legitimate to infer from comparing these figures that one dollar of income in
each economy is worth the same? Does the local equivalent of one US dollar purchase an
equivalent quantity of goods in every single country, so that one could say that the equiva-
lent of US$1,500 of income would provide the same standard of living in Chile, Malaysia,
Mexico, and Cambodia?
The simple answer is, no, it is not the case that the equivalent of one US dollar purchases
the same quantity of goods regardless of the economy. A little introspection perhaps suggests
why this is the case. Would you expect the price of housing, for example, to be the same in
Cambodia with a lower level of income per person than the US? Of a haircut? Of medical
and dental services? Of public transportation?
What is likely is that economies with lower per capita GNI will have lower prices for
48 The Process of Economic Development
some items when these values are converted from the local currency to US dollars than
will a country with a higher average GNI. In other words, the equivalent of US$1,500 will
buy different quantities of goods and services in different economies since the prices of
some ‚Ä“ but not all ‚Ä“ goods and services will vary with the level of average income of an
The GNI and GDP and the GNI per capita measures shown in Tables 2.1 and 2.2 are shown
in US dollar units, but these values are not precisely comparable among economies, for
reasons we just hinted at. These values were calculated by taking each economy‚Ä™s own GNI
and GDP values, calculated by each country‚Ä™s statistical agencies in their own currency units
(pesos for Mexico, rupees for India, for example) and converting these values to US dollars.
How is this conversion to US dollars done? Very simply, the average official exchange rate
between each country‚Ä™s currency and the US dollar is used as the means to arrive at the US
dollar values shown in Table 2.1 (and the real values in Table 2.2).
Let‚Ä™s imagine that the average exchange rate over a year between the Indian rupee and the
US dollar is 1 US dollar = 39.45 rupees. If India‚Ä™s GDP is determined to be 41,525,000,000
rupees, when that is converted to dollars using the official exchange rate, it would be equal to
$1,052,598,226. What such a conversion from rupees to dollars implies is that 1 US dollar in
the US can buy exactly the same bundle of goods and services as can 39.45 rupees in India.
That is, the buying power of 1 dollar is the same as the buying power of 39.45 rupees.
But you already know that this is not likely to be true. This official exchange rate conver-
sion only makes comparable the prices of traded goods, that is, goods that are traded inter-
nationally, such as computers, motor cars, shoes, oranges, and wine. The presumption is
that the prices of traded goods will be quite similar between countries because of the forces
of international competition and the potential for arbitrage that large differences in prices
However, for non-traded goods and non-traded services, which by definition do not enter
into international trade between nations, prices between countries can vary quite substan-
tially. These differences will depend upon conditions internal to each country, particularly
the average level of income, but also local customs, regulations, the degree of competition,
and so on. For non-traded goods and services, like housing, transportation, personal services,
and prepared foods that are location-specific, there are no international forces of competition
or the possibility of arbitrage to bring prices into line between economies. Significant price
differentials for these goods and services between economies can make international GNI
and GDP comparisons like those shown in Table 2.1 based on simply converting domestic
currency measures to a common US dollar measure somewhat deceiving, since a US dollar
does not have the same buying power everywhere. In poor countries, a US dollar, converted
into the local currency, can buy more than a dollar can in the US, since the prices of non-
traded goods and services tend to be lower.
There, is however, another way to compare income between countries that attempts to
overcome the shortcoming of the traditional exchange rate-converted GNI or GDP values.
This is known as the purchasing power parity, or PPP, income measure.
Table 2.4 provides a comparison between the values of GNI per capita calculated at the
official exchange rate (the same as those shown in Table 2.1) and GNI per capita calculated
at PPP values, both in US dollars. The PPP measure makes an adjustment to GNI between
countries similar to the adjustment made to determine real GNI discussed earlier. The prices
of one country, in this case the United States, become the base prices for determining the
purchasing power parity value of GNI per capita in other countries. Thus, Mozambique‚Ä™s
PPP GNI (‚Äúpurchasing power parity GNI‚ÄĚ) per capita is determined as
Measuring economic growth and development 49
Table 2.4 Purchasing power parity (PPP) measure of GNI per capita
GNI per capita at official exchange rate, 2005 PPP GNI per capita, 2005
Algeria 2,730 6,770
Argentina 4,470 13,920
Brazil 3,550 8,230
C√īte d‚Ä™Ivoire 1,490
Kenya 540 1,170
Korea 15,840 21,850
Mozambique 310 1,270
Venezuela 4,820 6,440
Source: World Bank, World Development Indicators 2007: 14‚Ä“16, Table 1.1.
Pi, US √— Qi,M
PPP GNI per capita = Population (2.5)
Qi,M is the output vector of all newly produced final goods and services, i, available for use
by residents of Mozambique and Pi,US is the price vector for these goods and services, i, in
US prices. Effectively, then, what the PPP GNI measure provides is the estimated value of
Mozambique‚Ä™s available output valued at the prices for such goods and services prevailing
in the United States. There is no need to use the exchange rate between the two countries to
find the value of Mozambique‚Ä™s GNI per person. Mozambique‚Ä™s output is valued directly by
multiplying that production by US prices.
Obviously, large differences not only in the prices of non-traded goods and services
between the two countries but of the mix of traded to non-traded goods in total national
output will affect the PPP measure of GNI per capita compared to the value obtained from
the official exchange rate conversion. From Table 2.4, for example, the 2005 per capita PPP
value of income in Mozambique was estimated as $1,270, which is more than three times
50 The Process of Economic Development
greater than the exchange rate-converted GNI per capita value of $310. This PPP per capita
income figure is more meaningful when comparing incomes among economies at any point
in time. The PPP value of Mozambique‚Ä™s income can be interpreted as follows: $1,270 of
income would be required in the US to buy what the equivalent of $310 is able to purchase
in Mozambique with its lower prices for non-traded goods and services compared to the US.
In other words, the equivalent of $310 in Mozambique can buy, roughly, what it would take
$1,270 to buy in the US.
For the less-developed countries in the top part of Table 2.4, all have PPP GNI per capita >
the exchange rate determined value of GNI per capita. For example, Rwanda had a PPP
GNI per capita nearly six times as large as the exchange rate GNI per capita value. Rwanda
is undoubtedly a poor country, but the $230 income per person figure makes the economy
seem poorer than it is. That income buys more in Rwanda than it would in the US or other
more developed economies, and there is also more home production not counted as income,
as you read about in Focus 2.2, that adds to the standard of living but does not appear as
income. For all the less-developed countries shown in Table 2.4, PPP GNI per capita exceeds
the exchange rate GNI value by at least a third and most often by significantly more. The
PPP measure of income gets us a little closer to understanding the levels of income among
On the other hand, look at the comparison between Japan‚Ä™s PPP GNI per capita and the
value of GNI per person found by simply converting per capita GNI in yen to US dollars
using the exchange rate. PPP GNI per capita is substantially less. Why do you think that is?
What does this say about the prices of non-traded goods and services in Japan compared to
those prices in the US (again, we expect the prices of traded goods to be very similar in all
Typically, then, the actual purchasing power of income in lower-income countries tends
to be understated by simply converting local GDP or GNI per capita to US dollars using the
official exchange rates as a result of the lower prices of non-traded goods and services, such
as housing, retail services, local food products, and local transportation in poorer nations.
These prices are lower because the lower income of these countries keeps the prices of
these non-traded goods below what they are in more developed nations. The more developed
economies tend to have PPP GNI per capita values closer to that calculated at the official
exchange rate because of the greater openness to world trade, a mix of production with more
traded goods relative to non-traded goods, and because of their more modern structures of
production, which result in greater efficiency in production in both traded and non-traded
goods and service sectors.
There is an increasing tendency to prefer the PPP measure of income in making compari-
sons among countries over the exchange rate-converted GNI or GDP values. In future,
income comparisons used as a basis of determining relative levels of development will more
and more use the PPP income measure and that will improve the quality of such comparisons
and the meaning we attach to them.
The indicators criterion of development: the Human Development Index
In the 1960s, there emerged from the International Labour Organization, from the World
Bank, and from independent researchers a growing backlash against the use of per capita
income and the rate of economic growth criterion as the exclusive measures of development.
Whether what was proposed as an alternative to the GNI, GDP, or PPP GNI per capita meas-
ures was the basic needs approach or the physical quality of life index (PQLI), or some other
Measuring economic growth and development 51
composite measure, the objection to the use of the economic growth and income standard
was the same: it was far too aggregate and did not capture the distributional inequalities all
too common in many of the poor nations of the world.
The income per capita criterion gave a biased view, it was argued, of the level of progress
achieved by many countries. Income per capita was, in and of itself, an insufficient target for
ultimately achieving society‚Ä™s broader development goals listed earlier in this chapter. The
link between the level of income per capita and the full range of development objectives was
considered much too tenuous and unreliable, particularly in the poorest nations that needed
to make the most progress.
Neither the basic needs nor PQLI methodologies took hold, however; the former perhaps
because of some undeniable theoretical and empirical ambiguity and the latter possibly for
lack of a powerful institutional champion.10 Since 1990, when it was first proposed, a new
measure of development has gained credibility. The Human Development Index, or simply
the HDI, has been calculated and published each year by the United Nations Development
Programme in its annual Human Development Report.
The HDI is a composite index using ‚Äúlongevity, knowledge, and a decent standard of
living,‚ÄĚ as the representative indicators for development. The actual index uses estimates of
life expectancy at birth, the adult literacy rate, school enrolment ratios, and PPP GDP per
capita to calculate an HDI value for each economy.11 The HDI measure of development is
thus broader than the simple income per person yardstick, though income does enter into
the calculation of the HDI. At the same time, the HDI gives direct value to those factors,
particularly education, which help create opportunities for individuals to reach a higher and
more fulfilling standard of living that may not be captured by the income measure alone. As
the UN Development Programme described the issue:
Human development is about much more than the rise and fall of national incomes. It is
about creating an environment in which people can develop their full potential and lead
productive, creative lives in accord with their needs and interests. People are the real
wealth of nations.
(UNDP 2001: 9)
That last sentence is extremely important. Read it again. It is a nation‚Ä™s people that comprise
the wealth of any society, and meeting the needs and desires of those people is the ultimate
purpose of economic growth and development.
The HDI simplifies the comparison among countries by combining the achievement
on a number of different variables into a single number. The value of the HDI index can
vary between 0 and 1, with an HDI score closer to 0 indicating greater distance from the
maximum to be achieved on the aggregate of the factors entering the HDI. An HDI value
closer to 1 indicates greater achievement relative to the maximum attainable on the variables
that comprise the index and thus a higher level of human development.
The HDI may be said to be measuring ‚Äúrelative deprivation,‚ÄĚ that is, it is a gauge as to
how far a country is from reaching, on average, the maximum value of the components
that make up the HDI. Roughly, since this is a deprivation index, one can interpret an
HDI = 0.660 to mean that an economy has, on average, attained 66 percent of what is
The HDI measure was created with the purpose of attempting to take into account the fact
that countries, meaning both governments and individuals, make choices on their spending
and use of resources among alternative uses. The use of these resources affects the range of
52 The Process of Economic Development
choices open to people and their level of well-being, with effects that may not always be
captured in the income per person ranking of nations.
For example, among the less-developed nations, the UNDP found that though one-quarter
of national income was spent via government, less than 10 percent of this share, on average,
was dedicated to identifiable human development expenditures, such as education, health
care, and social security. The largest area of government spending was on the military, the
contribution of which to human development is, at best, controversial (UNDP 1993: 10).
Of course, different nations will allocate their public expenditures in different ways, both
to achieve particular development goals, as well as to accomplish other priorities, such as
defense, that are deemed significant. The impact of these choices, at least partly, will be
captured by the variables included in the HDI, and thus, it is argued, provide a more robust
view of the average level of development of an economy than is possible by simply looking
at income per person.
Table 2.5 shows the value of the HDI for 1990 and 2004 and the HDI ranking for 2004 for
an even broader range of countries than was listed in Table 2.1. What do the numbers show?
Consider, for example, Mexico in the top part of the table for economies with ‚Äúhigh human
development‚ÄĚ in 2004. In 1990, Mexico‚Ä™s HDI value had been 0.766, which meant that
the country was then still ranked among the ‚Äúmedium human development‚ÄĚ economies. By
2004, with an HDI of 0.821, Mexico had joined the ‚Äúhigh human development‚ÄĚ economies,
being fifty-third among the 177 countries. Remember, too, that an HDI = 0.821 means that,
on average, Mexico has attained 82.1 percent of the maximum values possible on the indi-
vidual components of the HDI ‚Ä“ life expectancy, school enrolments, adult literacy, and PPP
GDP per capita. Chile, Costa Rica, and Malaysia also made the jump from medium to high
human development between 1990 and 2004.
By comparison, Niger, the lowest ranked economy in 2004, had an HDI = 0.246, indi-
cating that only about 25 percent of what could be achieved was attained, or, looking at it
from the other side, Niger had an average 75 percent shortfall from the maximum values on
the HDI components.
Two countries ‚Ä“ Kenya and Rwanda ‚Ä“ dropped from the medium human development
ranking in 1990 to the low human development grouping in 2004. In the case of Zimbabwe,
the deterioration in its HDI value from 0.639 to 0.491 was substantial. It will be noted that
all the low human development economies in Table 2.5 ‚Ä“ and the great majority of the thirty-
one economies with low human development in 2004 ‚Ä“ are Sub-Saharan African countries
for which the difficulties of sustaining economic growth and development has been most
critical. In some cases this has been due to civil war or other conflicts; in others, the HIV/
AIDS crisis has had devastating consequences on human development. Much more attention
needs to be given to Sub-Saharan Africa‚Ä™s particular problems if the MDGs are to have any
chance of being attained.
Table 2.5 also provides information in the penultimate column on the difference between
the PPP GDP per capita ranking and the HDI ranking for each country for 2004. What is the
significance of these numbers?
A positive value in that column indicates by how much a country‚Ä™s HDI ranking exceeded
its PPP GDP per capita, or income, ranking among all economies. That value is determined
by taking a country‚Ä™s PPP GDP ranking minus the country‚Ä™s HDI ranking. Countries with
a positive value in the PPP GDP ranking ‚ą’ HDI ranking column thus were ranked higher
among all economies in the HDI ranking of countries than they were in the PPP GDP income
ranking of economies. For such countries, their PPP GDP per capita ranking understated
the country‚Ä™s overall level of development, as more broadly defined by the HDI. Looking
Table 2.5 Human development index (HDI) and GDI, selected countries, 1990 and 2004
HDI ranka GDIc
HDI value PPP GDP ranking
‚ą’ HDI Rankingb
1990 2004 2004 2004
High Human Development (HDI > 0.800 in 2004)
Australia 0.893 0.957 3 11 0.956
0.914 0.949 11 0.942
United States 0.917 0.948 ‚ą’6 0.946
United Kingdom 0.889 0.940 18 ‚ą’5 0.938
0.823 0.916 25 ‚ą’4 ‚Ä“
Korea 0.823 0.912 26 5 0.905
Argentina 0.813 0.863 36 10 0.859
0.859 38 18 0.850
0.793 0.841 48 13 0.831
United Arab Emirates 0.810 0.839 49 ‚ą’25 0.829
0.766 0.821 53 0.812
0.723 0.805 61 ‚ą’44 0.795
Medium Human Development (0.500 < HD < 0.799 in 2004)
Brazil 0.792 69 ‚ą’5 0.789
Venezuela 0.760 0.784 17
0.717 0.784 74 ‚ą’9 0.781
Saudi Arabia 0.774 76 ‚ą’31 0.744
0.628 0.768 81 9 0.765
0.763 84 19 0.761
0.682 0.757 92 ‚ą’22 0.745
0.706 0.755 93 13 0.749
0.719 0.724 104 6 0.721
0.626 0.711 108 0.704
0.618 0.709 109 12
South Africa 0.735 0.653 121 ‚ą’66 0.646
0.549 0.640 123 ‚ą’15 0.615
0.515 0.611 126 ‚ą’9 0.591
0.680 0.570 131 ‚ą’73 0.555
0.463 0.539 134 ‚ą’6 0.513
0.422 0.530 137 0.524
0.528 0.520 140 25 0.519
Low Human Development (HDI < 0.500 in 2004)
Zimbabwe 0.639 0.491 151 ‚ą’18 0.483
Kenya 0.548 0.491 152 0.487
0.339 0.450 158 ‚ą’5 0.449
Nigeria 0.407 0.448 159 ‚ą’1 0.443
C√īte d‚Ä™Ivoire 0.443 0.421 164 ‚ą’15 0.401
Mozambique 0.316 0.390 168 ‚ą’14 0.387
0.314 0.371 170 1 ‚Ä“
Niger 0.246 0.311 177 ‚ą’7 0.292
High Income Economies
Middle Income Economies
Low Income Economies
Source: UNDP 2006: 283‚Ä“91, Tables 1 and 2, 210‚Ä“13, Table 21.
a The highest, or best, ranking was 1 (Norway) in 2004; the lowest, or worst, ranking, was 177 (Niger).
b If positive, the ranking for the country on the HDI is higher than the country‚Ä™s ranking on per capita PPP GDP (PPP GDP rank ‚ą’ HDI
rank > 0); if negative, the HDI ranking for the country is lower than the per capita PPP GDP ranking (PPP GDP rank ‚ą’ HDI rank < 0).
c Gender-related Development Index; adjusts HDI for differences in achievement on the HDI variables between males and females.
54 The Process of Economic Development
at income alone, these economies seemed worse off compared to other economies than they
really were when a broader definition of development, the HDI, was used.
For example, the Philippines (HDI = 0.763) had an HDI ranking that exceeded its per
capita PPP GDP ranking by nineteen places, since its PPP GDP ranking ‚ą’ HDI ranking differ-
ence was a positive 19. (What, then, was the Philippine‚Ä™s per capita PPP GDP ranking? We
can infer its value from the table. Since the PPP GDP ranking ‚ą’ HDI ranking = x ‚ą’ 84 = 19,
then the Philippines PPP GDP ranking must have been 103 in 2004, since 103 ‚ą’ 84 = 19.) In
other words, the Philippines did substantially better on the HDI measure, ranking, at eighty-
fourth, nineteen countries higher among the 177 countries than it had ranked on its PPP GDP,
being one-hundred-and-third among the 177 countries. This was a better showing on the
broader range of development indicators captured in the HDI than would have been expected
from its per capita income ranking alone.
On the other hand, a negative value in the penultimate column of Table 2.5 indicates by
how many positions a country‚Ä™s HDI ranking fell among all countries compared to where
the country ranked when using its PPP GDP per capita only. The value of ‚ÄúPPP GDP
ranking ‚ą’ HDI ranking‚ÄĚ can only be less than 0 if the PPP GDP ranking < HDI ranking.
For those nations with a negative gap, their income ranking tends to overstate the broader
level of development as measured by the HDI.
For example, South Africa‚Ä™s HDI ranking was 66 places below its income ranking (PPP
GDP ranking ‚ą’ HDI ranking = x ‚ą’ 121 = ‚ą’66, which means that the PPP GDP ranking had
to be 55 since 55 ‚ą’ 121 = ‚ą’66). This strongly suggests that had we looked at income alone,
it would have been reasonable to believe that South Africa‚Ä™s population had a higher level
of development, ranking fifty-fifth among the 177 countries, than it actually does, at least as
measured by the additional development indicators included in the HDI, where South Africa
ranked significantly lower, one-hundred-and-twenty-first among the 177 countries. That is
quite a significant drop in ranking, reflecting the fact that average income in South Africa
fails to provide a reasonable proxy measure of the level of development.
Looking down the column of values in the PPP GDP ‚ą’ HDI column, they do not seem to
vary systematically in any immediately obvious way, underscoring the point of those who
have argued that the GNI (or GDP) per person measure alone is an incomplete index of
development and that there is no automatic link between the level of income per capita and
the level of development (at least as measured by the HDI).12
adjustments to the HDI
Just as it is useful to adjust the GDP or GNI income figures so that they provide a more reliable
standard for evaluating the level of development if one is to use that measure, so too are there
some modifications that can be made to the HDI that refine the information it provides.
The Gender-related Development Index (GDI)
The UNDP also calculates a gender-adjusted HDI, called the ‚Äúgender-related development
index‚ÄĚ or GDI, which takes into account differences in the level of attainment of women
and men on the values of the indicators that enter the HDI. The GDI values are shown in the
last column of Table 2.5. In making such a correction for gender differences in life expect-
ancy, education, and income, every country suffers a deterioration in the value of its gender-
adjusted HDI, meaning in no country do women, on average, score higher than or equal to
men on the HDI components (though some are very close). Thus, the GDI, which is simply
Measuring economic growth and development 55
a gender-adjusted HDI value, is less than the average HDI reflecting the lower average level
of attainment of women compared to men on the variables entering the HDI.
Some countries do better than others on gender equality, however, so that the GDI ranking
of all countries is different from the HDI ranking, rising for those nations for which the
average achievement of women is closer to that of men and falling for those nations where
the achievement of women is more distant from that of men as a group. That can be seen, for
example in looking at Saudi Arabia and China in the table. Saudi Arabia has a higher HDI
value than does China and thus ranks higher, seventy-sixth versus China, which is ranked
eighty-first on the HDI. However, looking at the GDI values of each country, we see that
China scores higher than Saudi Arabia and thus China ranks higher on the GDI, sixty-fourth
versus seventy-second for Saudi Arabia (the GDI rankings are not shown in Table 2.5). Still,
no country had a GDI equal to or greater than its HDI. Closing the gaps in education, health,
income, and political participation between men and women is essential for full development
of a nation, and the GDI gives us some indication of the success of countries in achieving
The Human Poverty Index
Another weakness of the HDI is that it does not indicate what is happening to the poorest
members of society, except to the extent that this is reflected in the overall HDI value via,
say, the impact of poverty on average life expectancy. To attempt to capture the conditions
of living of the poor more directly, the UNDP introduced in 1997 a Human Poverty Index
(HPI) that utilizes slightly different variables from either the HDI or the GDI and hence is not
directly comparable. Instead of life expectancy as a variable, the HPI includes the probability
at birth that a child will not survive to age 40; for education, adult illiteracy is included as
a variable; and in place of an income variable, the HPI includes the ‚Äúpercentage of people
without sustainable access to an improved water source‚ÄĚ and the ‚Äúpercentage of children
under five who are underweight‚ÄĚ as variables entering the index. All of these variables are
stated in percentages, and lower values for each is better than higher values since each vari-
able is actually a ‚Äúdeprivation‚ÄĚ indicator rather than an achievement indicator, as is the case
of the HDI or GDI variables (for a sample calculation, see UNDP 2001: 241).
In 2005, for example, Costa Rica had an HPI of 4.4 percent, which means the combined
probability of not living to age 40, of the incidence of illiteracy, of lack of access to improved
water sources, and of underweight children is quite low. Pakistan‚Ä™s HPI was equal to 36.3
percent, meaning a worse performance on these deprivation indicators. Of the 102 less-
developed countries with calculated HPI values, Mali was last, with an HPI = 60.2 percent,
the result of having a 37.3 percent chance at birth of not living to age 40; of having 81.0
percent adult illiteracy; of 50 percent of the population lacking access to improved water
sources; and of having a third of all children below age 5 being underweight (UNDP 2006:
293‚Ä“4). While this index is not as easy to use as the HDI, it does give some useful informa-
tion about the levels of relative deprivation of different countries that can be used in conjunc-
tion with the HDI value.
The HDI is thus an imperfect measure of the well-being of a nation, just as is the income
criterion. Neither measure is capable of capturing all the critical dimensions of development
that might be thought important, but the HDI is broader in some important respects than is a
simple income per person indicator.
For the future, it might be desirable to have an HDI that included some measure capable
of quantifying environmental and sustainability issues that are a part of the Millennium
56 The Process of Economic Development
Development Goals. Such environmental matters are left out of the HDI, except to the extent
they might indirectly affect life expectancy. With the growing awareness of the pressing need
to understand the interrelation between biological and economic systems and the concern
that an environmental threshold may be being reached with continued global economic
expansion, this is a glaring deficiency of the HDI, as it is in the GNI and GDP measures. So,
too, is the absence of any weight in the gross HDI given to the degree of political democracy
and participation. Other considerations and variables may also be important to include in an
adjusted HDI, as has been the case with the GDI.
Comparing the income per capita and HDI measures
Is the effort to construct an HDI for each country worth the effort? Does the HDI provide
information about the level of development of a country that is different from that which can
be obtained from GDP or GNI per capita figures? Is it reasonable to use real GNI or GDP per
capita as a proxy for the level of development, rather than the admittedly more-difficult-to-
estimate HDI? As the UNDP writes (2001: 13):
Rankings by HDI and by GDP per capita can be quite different, showing that countries
do not have to wait for economic prosperity to make progress in human development. ‚Ä¦
Costa Rica and Korea have both made impressive human development gains, reflected
in HDIs of more than 0.800, but Costa Rica has achieved this human outcome with only
half the income of Korea. Pakistan and Viet Nam have similar incomes, but Viet Nam
has done much more in translating that income into human development. ‚Ä¦ So, with
the right policies, countries can advance faster in human development than in economic
In a statistical study comparing GNI per capita and the HDI as means for ranking nations
as to level of development, the question of a divergence between the HDI and income rank-
ings was considered. It was found that there was a high correlation between the GNI per
person ranking and the HDI ranking when the entire sample of countries was considered
(Dietz and Gibson 1994; in that study there were 143 economies). This tends to support the
view of those who argue that per capita income is a reasonable proxy for ranking nations as
to their relative and absolute level of development. However, when the sample was examined
in more detail, this conclusion could be supported only weakly.
Using income per capita as a surrogate for development is most reliable both for the highest-
income nations and for the lowest-income, least-developed nations, with some notable excep-
tions, like Sri Lanka, China, Guyana, and Indonesia in the latter category. For the seventy-two
lower-middle and upper-middle income countries in the study, however, the level of income
per capita turned out to be an unreliable indicator for the level of human development and an
unreliable ranking methodology for relative human development among those nations.
The results of this study strongly suggest that considering both the level and relative position
of a country using GNI per capita or GDP per capita (particularly if the PPP values are used)
and the HDI score is, perhaps, the more prudent way to evaluate the level of development.
Since the link between economic growth and development is neither direct nor consistent
for all countries, tracking progress over time using both the HDI and income provides more
robust information than either per capita income or the HDI value alone. Given that both
income and HDI values now are readily available annually, there is a strong argument for
making use of both, perhaps especially so for the middle-income, less-developed nations.
Measuring economic growth and development 57
The HDI thus provides an alternative measure to the income per person, economic growth
criterion for evaluating the progress of a country in terms of achieving broadly accepted
development goals. The HDI reminds us that though increased income is vital for the
expanded choices it provides individuals and families, it is not the whole of what develop-
ment is about. The unadjusted HDI partly captures the extent to which the spread effects of
growing incomes are filtered through education, health, social security, and other areas of the
economy and how incomes are distributed to expenditures that are both means to, and ends
of, higher levels of development.
The search for any single indicator that will provide all the information that is important
and relevant to evaluating development is perhaps elusive. Any single indicator is likely
to be at best an imperfect measure. Awareness of the weaknesses of whatever measure is
used and an effort by the observer of the development process of any nation to ‚Äúfill in the
gaps‚ÄĚ of coverage inevitable to any single index remains, nonetheless, an essential and some-
what subjective ingredient to the evaluation of progress and to making recommendations for
Economic growth and equity: goals at odds?
If a country places an emphasis on development and the targeting of objectives such as increased
education and better health care that contribute to improvements on the HDI measure or to a
reduction in the numbers of the poor, will this adversely affect economic growth? Conversely,
if a country takes steps aimed at accelerating economic growth, what effect will this have on
the goals of development, as measured by the HDI or by poverty reduction? In other words,
are economic growth and development at odds or are they broadly complementary?
Underlying the insistence of some economists on measuring development by income per
capita and of targeting this as a proximate goal for development is a concern that if policy-
makers place too much emphasis on achieving other, more specific development goals this
may slow the pace of economic growth by taking scarce resources away from investment and
other uses. A slower pace of economic growth may actually make the achievement of these
desired development goals all that more difficult by reducing the material resources avail-
able to improve the level of human welfare. Is there, in fact, such a conflict between pursuing
economic growth and pursuing development, more broadly defined?
An influential study published in 1955 by the late Nobel Prize-winning economist Simon
Kuznets examined the historical relationship between income per capita and income distri-
bution, one broad indicator of equity. While trends in a country‚Ä™s income inequality are an
imperfect indicator of what is happening to the broader goals of development, rising and
high levels of income inequality, if persistent, may be a signal of underlying weaknesses in
an economy‚Ä™s structure in being able to broadly deliver a higher level of development. The
research by Kuznets and others following in his footsteps has had a great influence on how
many analysts think about the relation between economic growth, as measured by rising per
capita income, and the achievement of the broader goals of development.
Kuznets‚Ä™s analysis suggested that at low income levels economic growth and rising
average income tended to create more income inequality as measured by the Gini coefficient.
As income per capita continues to increase, however, a critical threshold level of income was
reached, and further economic growth and even higher average income tended to reduce a
nation‚Ä™s overall income inequality. This relationship between the level of per capita income
and income inequality is referred to as the Kuznets inverted-U hypothesis, from the shape of
the curve shown in Figure 2.1
58 The Process of Economic Development
0 Threshold income
Figure 2.1 The Kuznets curve.
The Kuznets hypothesis is often interpreted to mean that there is a minimum level of income
that a country must achieve before greater equity and higher levels of development can be
attained. Once that threshold level of income is reached, further increases in income contribute
to greater equity, as shown by the falling value of the Gini coefficient after the peak of the curve
is reached at the threshold income level of approximately $1,000 per person in this graph.
Prior to reaching the threshold level of income in Figure 2.1, however, rising income is
associated with increasing inequality, as shown by the rising Gini coefficient value associ-
ated with higher income on the upward sloping portion of the curve. What the Kuznets curve
suggests is that greater income inequality is the ‚Äúcost‚ÄĚ of rising income per capita prior
to reaching the threshold level of income, a cost that is necessary if the threshold level of
income is ever to be reached. After passing the threshold level of income per person, income
inequality will begin to fall as there is additional economic growth and higher levels of per
capita income are reached.
In other words, the Kuznets curve can be interpreted to mean that poorer countries at an
early stage of their economic development can expect a deterioration in income inequality
until the threshold level of income is reached. ‚ÄúThings must get worse before they can get
better and a higher level of development can be attained‚ÄĚ is one way to summarize this inter-
pretation of what is called the Kuznets hypothesis.
This Kuznets inverted-U hypothesis sometimes has been interpreted as something of a
law of economic growth and development. Nations wishing to promote equity and human
development in the wider sense can best do so by increasing income per capita. Initially, at
income levels below the threshold level, rising income per capita makes income inequality
worse, but that is the price that must be paid both to attain higher average income and to
eventually reduce inequality.
There is no apparent necessity to target development goals or poverty reduction per se
if one accepts this view. The short-term loss in equity that accompanies economic growth
before the threshold level of income is reached is the necessary price of progress over the
longer haul. From this interpretation of the Kuznets curve, growth and development are not
rival goals. economic growth promotes development and equity in income over the longer
term, even if there would seem to be a short-term trade-off (see Focus 2.4 for more on the
possible trade-off between inequality and growth).13
Measuring economic growth and development 59
FOCUS 2.4 INEQUALITY AS A CONSTRAINT ON GROWTH
The conventional wisdom as expressed in the Kuznets curve is that there is a trade-off
between economic growth and reducing inequality. Thus an unequal distribution of income
is sometimes believed necessary for rapid economic growth, at least at low income levels
prior to reaching the threshold income level shown in Figure 2.1. If this is so, however, why
do we find in Latin America relatively low rates of economic growth and high inequality,
and in East Asia low inequality and rapid growth?
Differences in the political economy and the policies of the two regions may be part of
the explanation. In the period after 1945, governing elites in East Asia had their legitimacy
threatened by domestic communist insurgents. They thus sought to widen their base
of political support via policies such as land reform, public housing, investment in rural
infrastructure, and, most commonly, widespread high-quality education. In Latin America,
governing elites acted as if they believed they could thrive irrespective of what happened
to those with the lowest income as reflected in the tax, expenditure, and trade policies that
the political elites legislated and that benefited the poor relatively little.
The association of slow economic growth and high inequality in Latin America may in
part be because too much income inequality may act as a constraint on growth by limiting
demand and the size of the market. Gini coefficients in Table 2.3 are higher for Argentina,
Brazil, Chile, and Mexico than they are for Korea or Thailand. By contrast, East Asia‚Ä™s
lower level of income inequality as a result of different policies may have provided signifi-
cant stimulus to economic growth by increasing domestic demand and spending, as more
people had more income to spend.
If this was the case, there is a strong argument that East Asia‚Ä™s investment in educa-
tion has been a key difference from Latin America, as increases in the average level of
schooling had a leveling effect on the income distribution. Rising average levels of educa-
tion in East Asia, of course, contributed to increasing economic growth directly through
the positive effect on productivity. But higher levels of education spread among the popu-
lation also resulted in a larger core of better-educated workers who all earned more, and
this reduced income inequality at the same time.
How significant a constraint on economic growth is ‚Äútoo much‚ÄĚ income inequality? It
appears to be quite substantial. Research suggest that, ceteris paribus, after twenty-five
years, GDP per capita could be 8.2 percent higher in a country with low inequality than in
a country with income inequality one standard deviation higher.
Simulation results suggest that if, in 1960, Brazil had had Korea‚Ä™s lower level of income
inequality, Brazil‚Ä™s growth rate over the following twenty-five years would have been 0.66
percentage points higher each year. This implies that after a quarter-century, GDP per
capita in Brazil would have been 17.2 percent higher than it was with the higher degree of
inequality in that economy, a substantial difference that could have contributed not only
to a higher average standard of living but, perhaps, to the broader goals of development
Source: Birdsall and Sabot 1994
The debate over the Kuznets inverted-U hypothesis and whether it represents the ‚Äútrue‚ÄĚ
relation between rising average income and income inequality has generated a vast and often
complicated literature. The relation Kuznets discovered between income and equity is not, in
fact, a law of economics but rather a statistical relation that does not show causality.14 What
seems to happen is that once nations pass the threshold level of income, government expen-
ditures on health, education, social security, and other social and human capital areas tend to
rise relative to total expenditures in the economy as public revenues rise. Thus, improvement
in equity and on the HDI measure would be expected, as governments are able to focus on
broader development goals, which leads to more economic growth in the future.15
60 The Process of Economic Development
More importantly, the Kuznets hypothesis ‚Ä“ which was found by looking at cross-section
data for a number of countries at one point in time ‚Ä“ cannot be sustained as a ‚Äúlaw‚ÄĚ once the
experience of individual nations is singled out and examined over time rather than at just
one point in time. Some countries have experienced worsening equity along with economic
growth, even after what might reasonably be thought to be the threshold level of income
has been reached (Brazil is an example), while other nations have been able to improve
equity and score higher on the HDI measure at income levels well below Kuznets‚Ä™s threshold
level of income (Sri Lanka, for example). The reason? Specific government policies on, for
example, education, land and wealth distribution, and other social policies can be focused on
the broader development goals early on (Sri Lanka) or taken away from such goals (Brazil),
somewhat independently of the level of current income. There is no hard and fast law that
says that a country cannot have both rising income and more development at low income
levels ‚Ä“ or vice versa.
What the Kuznets inverted-U demonstrates, then, is historically what on average did
happen for a group of nations at one point in time. It does not imply that all countries, espe-
cially late developers, must necessarily tolerate or even promote increasing inequality to
achieve economic growth over time. The particular path which any nation follows in terms
of the relation between its economic growth rate and its success in reaching the broader
goals of development is at least partly a consequence of conscious public policy. Such policy
can be geared toward high economic growth and greater equity and development, or high
growth and slow development. even slow economic growth paths can generate increasing or
decreasing inequality, depending on government policies aimed at confronting the sources of
inequality and the commitment to achieving important development objectives.
The particular mix of economic growth and development that results is at least partly a
public policy choice that a nation‚Ä™s leaders determine, with or without popular consent. This
is not to say economic growth does not matter for equity or development objectives; it does.
Over time, more economic growth and a higher average income level should contribute to
greater development by providing the resources needed to achieve those broader goals. The
reverse is also true, since increased human development can importantly contribute to higher
levels of labor productivity, particularly via increased education and better health care, which
lead to higher economic growth and income, thus creating a virtuous circle as more economic
growth contributes to improvements in development and vice versa.
Countries cannot ignore one side of the development equation ‚Ä“ either economic growth
or development ‚Ä“ for very long without suffering the consequences of a lop-sided policy. As
we shall see in later chapters, some East Asian countries were able to achieve quite substan-
tial progress on both the economic development and equity ledgers simultaneously, and that
would seem to be a path worth emulating (see Focus 2.5 on China).
The differences in the GDP and HDI rankings of economies shown in Table 2.5
suggested the importance of policy decisions by government and society in achieving
development goals, as measured by the HDI, at different levels of income, especially for
the middle-income, less-developed nations. The divergence in the rankings of countries on
their income per person and on their HDI values also confirms that the Kuznets curve is
not strictly a law governing the relation between equity and development and the level of
per capita income, or there would be no, or fewer, non-zero values for the PPP GDP‚Ä“HDI
Depending on a nation‚Ä™s policies, greater equity and progress on development goals can
be achieved even at relatively low levels of income. It is not necessary for countries to await
the threshold level of per capita income shown in Figure 2.1 before progress toward greater
Measuring economic growth and development 61
FOCUS 2.5 CHINA: A NEW TIGER?
Is the Chinese economy another example of an East Asian economy that is on a path that
will bring it to developed country status sooner rather than later?
There certainly is ample evidence of rapid progress. From 1986‚Ä“96, GDP per person rose
by 8.6 percent per year; from 1996‚Ä“2006, income per person grew by 8.2 percent annually.
It is predicted to grow even faster, nearly 10 percent per year until at least 2010 (World Bank
data). What does such rapid growth mean in concrete terms? With compounding, income
per person increased nearly fivefold over the twenty-year period from 1986 to 2006. A glance
back at Table 2.1 shows that GNI per capita grew even faster, since GNI > GDP in China.
If we look at income distribution, there has been a trend toward greater inequality over
this period of rapid growth. In 1998, the richest 20 percent earned 7.9 times what the
poorest 20 percent earned, compared to 12.1 times more income in 2004, as shown in
Table 2.3. The Gini coefficient rose between the two years from 40.3 to 46.9, suggesting
that economic growth has not been equally shared among income groups. This perhaps
reflects the transition China is making from a socialist to a mixed economy, where capi-
talism and markets play a larger role, and where greater inequality may be both more
likely and more functional than it was in a socialist setting. Or this relationship may simply
reflect the movement along a traditionally interpreted Kuznets curve prior to reaching the
threshold level of income, as discussed in this section, when income distribution worsens
before it gets better. Only time will tell.
China‚Ä™s HDI has increased over time, reaching 0.768 in 2004, not far from the 0.800
threshold that would advance the country into the high human development category.
That is quite an accomplishment. On individual indicators, life expectancy is seventy-
two years; more than 90 percent of the population aged fifteen or more is literate; there
is universal primary school enrolment; and PPP income is about $6,000 per person. On
many indicators, then, China has made remarkable progress. For a country that is still
relatively poor, it has achieved a population growth rate that is more like that in developed
economies than in less-developed nations, as we shall see in Chapter 12.
There are concerns about the sustainability of China‚Ä™s growth path and the issue of
the use of prison labor and a grossly undervalued exchange rate that help to make the
economy internationally competitive. There is little doubt that exports of manufactured
goods to world markets have helped to fuel these rapid growth rates. In fact, China‚Ä™s
consumption as a proportion of total GDP fell from more than 60 percent in 2000 to about
50 percent in 2006, as exports grew rapidly. This is where things get tricky.
China‚Ä™s exports and the profits of its large industrial corporations have been a driving force
for rapid economic growth over the recent past. It is this external dynamic, not the internal
economy, that is providing the motor force for much of China‚Ä™s rapid economic growth. If
there is an interruption in trade or a world slowdown in economic expansion, this could have a
severe impact on China‚Ä™s income and growth. This is why concern about the balance between
internal demand, especially for consumption spending, and the external demand from exports
is of such concern. Too much of a good thing could turn into too much of a bad thing.
Source: Anderson 2007
equity and development can be accomplished. It is a matter of policy decisions, and even
after attaining the threshold level of income there still is no guarantee of progress toward
greater equity if governments are not partners to positive change. Meeting development
objectives depends on the nature of state and social policy aimed at attaining greater equity
and a higher level of human development. Development and equity are not mechanically
determined by the level of income per person.
An innovative use of the Kuznets curve is to consider the relation of income levels and
income growth and the level of pollution (see Focus 2.6).
62 The Process of Economic Development
FOCUS 2.6 AN ENVIRONMENTAL KUzNETS CURVE?
Some researchers have suggested there may be an environmental Kuznets curve, similar
to Figure 2.1, but with pollution levels measured on the vertical axis rather than income
inequality. At relatively low income levels, increases in economic growth result in increased
pollution and environmental destruction. However, after a threshold level of income per
person is reached, pollution and adverse environmental effects will be reduced. Why might
such a relationship be expected?
At low income levels, societies typically place a relatively low value on a clean environ-
ment compared to the value of increasing economic growth; thus little attention will be paid
to environmental conditions. Increases in agricultural output might be expected to expand
the volume of toxic wastes created, and if industrialization is just beginning it is quite likely
that air and water pollution will accompany the growth of new factories. Clean and safe
technologies for production may not be available to poor countries at a reasonable cost.
The alternative to paying for cleaner technologies is to absorb the damage caused by
pollution and other environmental degradation as one of the ‚Äúcosts‚ÄĚ of improvements in
average living conditions.
A clean environment is often assumed to be a ‚Äúluxury good‚ÄĚ in the sense that its
income elasticity is greater than one. If this is the case, only at higher income levels will
a clean environment have a value worth preserving. If this is correct, then one would
expect to see an environmental Kuznets curve, at least for some kinds of pollution and
environmental damage (e.g. sulphur dioxide, particulate matter) that accompany indus-
trialization. Other types of environmental damages increase with higher income levels,
e.g. carbon dioxide emissions from vehicles, groundwater contamination, and municipal
waste and garbage, as the discussion of the pollution of affluence earlier in the chapter
in Focus 2.3 suggested.
However, much like the traditional Kuznets curve for income growth, the observation of
an environmental Kuznets curve has been based on what has occurred in the past. Now,
there is better information, better environmental accounting methods, and greater aware-
ness of the global significance of promoting a kind of economic growth that takes into
consideration environmental effects, in poor and rich countries alike. There is increasing
awareness that progress over time needs to be economically, politically, and environmen-
tally sustainable. In other words, it is recognized that steps need to be take to flatten out
or even induce a downturn of the environmental Kuznets curve. The real issue is whether
it is also politically feasible to do so. As the World Bank noted:
The principles of sound environmental policy ‚Ä¦ are well understood. But they are
difficult for national governments to introduce and are even more difficult to translate
into international agreements. National governments may be reluctant to challenge
those who cause environmental damage; they are likely to be rich and influential,
while those who suffer most are often the poor and powerless.
In other words, it is not wise to leave the environment to chance and to purely market
decisions. In many poor countries, it is not possible to simply wait until incomes rise
and hope that the environment will be valued more. Environmental degradation is now a
global problem and though its causes are often local, the international community needs
to work to see that the age of an environmental Kuznets curve is increasingly in the past.
This reality is recognized in the Millennium Development Goals discussed in Chapter 1,
especially Goal 7.
Sources: Dasgupta and M√¤ler 1995: 2384‚Ä“8;
World Bank 1992: 10‚Ä“13, 18, 38‚Ä“41, 43
Measuring economic growth and development 63
Summary and conclusions
We have considered in this chapter how to measure development among diverse economies.
One way is to use income per person as a proxy measure for the long list of objectives
enumerated at the beginning of the chapter that actually comprise ‚Äúbecoming‚ÄĚ developed. It
is necessary, however, to make several adjustments to the total income (GDP or GNI) values
calculated by each country so as to improve the usefulness of income as a simple measure of
the level of development of economies.
1 Each economy‚Ä™s income needs to be converted to a common currency to make compar-
ison among economies possible.
2 Total nominal income needs to be converted to income per capita values by dividing
total income in the common currency by total population to make comparison among
3 To compare economies over time, it is necessary to calculate real income per person by
deflating nominal income values by an appropriate price deflator.
4 The best measure of income per capita for comparing economies over time is purchasing
power parity (PPP) GDP or GNI.
The following chart summarizes the various income measures for GDP, clearly
showing how they differ, as well as how they are similar. The ‚Ä˜i‚Ä™s are the different goods
and services produced.
Calculating GDP for Botswana
Measure Price vector √— Output vector Text reference
Nominal GDP = √— equation 2.2
Pi, current year, Botswana Qi, current year, Botswana
Real GDP, base year 1992 = Equation 2.3
Pi, 1992, Botswana √— Qi, current year, Botswana
PPP GDP = Equation 2.5
Pi, current year, US √— Qi, current year, Botswana
In calculating the different measures of GDP (or GNI), the Qis used in each calcula-
tion are the same; what differs are the prices used to value each country‚Ä™s current output
As an alternative, or as complement to using income as a means to rank economies
as to their level and rate of economic development, the human development index, the
HDI, has gained support. The HDI includes education, life expectancy, and income to
capture more dimensions of development. It still is not used to the extent that income
per person is, but it serves as a reminder that development is not just about income but a
broader range of goals that societies set for themselves, if only implicitly at times.
Questions and exercises
1 In the discussion as to why GNI and GDP may diverge, the focus was on inflows and
outflows of profits and dividends as a result of foreign investments and on workers‚Ä™
remittances. What other inflows of income to and outflows of income from a country can
cause the GNI measure of income to differ from the GDP measure of income?
2 This problem will give you some practice examining the relationship between economic
growth and population growth.
64 The Process of Economic Development
a Over the period 1970‚Ä“80, Tanzania‚Ä™s total GDP grew at 3.0 percent per annum and
by 3.6 percent from 1980‚Ä“93, while population expanded at 3.1 percent over the
earlier period and by 3.2 percent, 1980‚Ä“93. What was the rate of growth of GDP per
capita over both periods?
b Botswana‚Ä™s population grew by 3.5 percent per year, 1970‚Ä“80, and by 3.4 percent,
1980‚Ä“93, while total GDP grew by 14.5 percent per year, 1970‚Ä“80, and by 9.6
percent per annum, 1980‚Ä“93. What was happening to Botswana‚Ä™s GDP per person
over each period?
c Does population growth ‚Äúcause‚ÄĚ slow or fast growth in GDP? What is the connec-
tion between population growth and the increase in income per person suggested by
these two examples?
3 Determine the estimated average income of the poorest 20 percent and of the richest
20 percent of income earners in Rwanda, Malaysia, Korea, Botswana, and Kenya by
applying the income distribution shares in Table 2.3 to GNI per capita in Table 2.1.
Compare these estimated average incomes for the richest and poorest 20 percent with
the per capita GNI values shown in Table 2.1. For which of these economies does the
average income shown in Table 2.1 give a good idea of actual living standards of the
poorest and richest?
4 You are going to draw a scatter diagram, with 2006 GNI per capita (from Table 2.1) on
the horizontal axis and the ratio of income received by the richest 20 percent to the share
received by the poorest 20 percent (from Table 2.3) on the vertical axis.
a Plot the data for ten of the countries in the tables.
b Is there any systematic relation between GNI per capita and the degree of income
inequality? Do countries with low levels of GNI per capita have more or less
inequality than economies with higher levels of GNI per capita? (If you have access
to excel or a simple statistical analysis package, you could run a simple regression
between the two variables to look for a relation.)
5 In 1990, Pakistan‚Ä™s total nominal GDP was $34,050 million. In 2000, total nominal GDP
had increased to $51,920 million.
a What was the total percentage change in nominal GDP between 1990 and 2000?
b What was the percent change, on average, per year of nominal GDP over the
c Now determine real (constant price) GDP in 2000 calculated at 1990 prices, given
that the price index in 1990 = 100 and in 2000 = 215.9.
d What was the percent change in real GDP between 1990 and 2000, both in total and
the per year average?
6 What explains the fact that Japan‚Ä™s purchasing power parity (PPP) level of GNI per
capita is so much lower than its GNI per capita calculated at the official exchange rate
(Table 2.4)? What does that difference indicate about the prices of non-traded goods in
Japan relative to US prices?
7 Explain why using the purchasing power parity measure of GNI per capita is considered
a better measure for comparing development levels between nations than the exchange
rate-converted GNI per person level.
8 Using the methodology for determining the value of the HDI shown in the UNDP Human
Development Report and summarized in Appendix 2B, calculate the value of the HDI
Measuring economic growth and development 65
for a country that interests you and that is not listed in Table 2.5. Does that country‚Ä™s
level of GDP per capita provide a reasonable proxy for the country‚Ä™s level of develop-
ment as measured by the HDI? Are there other countries with a similar level of income
to that of the country you have selected that have substantially different HDI values?
How might you account for those differences?
9 Focus 2.6 discusses the environmental Kuznets curve and suggests that avoiding such an
outcome is more of a political problem than one of know-how. What does the World Bank
mean in the quotation in Focus 2.6 that it tends to be the ‚Äúrich and powerful‚ÄĚ that cause much
of the pollution and the poor who suffer from the effects? In what specific ways do the ‚Äúrich
and powerful‚ÄĚ cause environmental damage? In what ways do the poor bear such costs?
Can you identify instances where it is the poor who contribute to environmental damage?
(Hint: think of clear-cutting of forests for wood for charcoal or for grazing of animals, as
one instance; also, reread Focus 2.3, which considers the ‚Äúpollution of poverty.‚ÄĚ)
10 In the exercise below, you will be calculating the GDP for two countries, the US and
India. To simplify, each country produces only one good and one service, copper and
retail sales. Fill in the blanks to determine the total nominal GDP for each country in its
own currency for 2001. Don‚Ä™t worry about the units in the problem below. Just do the
calculations using the numbers without attaching units. Remember equation 2.2 above
for doing the calculations! Can the GDP values you calculated be compared?
United States India
Quantity Price ($) Value of Quantity Price Value of output
output ($) (rupees)
160 $210/ton __________ R 2,625 __________
4.68 $5,200/ 16
Retail sales __________ R 22,280 __________
(millions of worker
Total GDP in __________ __________
11 Now calculate total nominal GDP for each country valued not in its own currency, but
in US dollars. If the average exchange rate in 2001 was 12.5 rupees = $1, India‚Ä™s total
nominal GDP valued in US$ at the official exchange rates is equal to ___________?
Can US and India GDP be compared now to say which country is ‚Äúbetter off‚ÄĚ?
12 This problem will give you practice in understanding how real GDP is determined. In
2000 in India (the figures in problem 10 above are for 2001), the price of copper was
R 2,500/ton, while the value of retail sales per worker was R 20,200. What is the value
of India‚Ä™s real GDP in 2001, assuming the 2000 prices are the base year prices for
comparison? Remember, real GDP measures only changes in actual physical output,
holding prices constant between years; if you need to, look back at Equation 2.3. (We
could then convert this value to US dollars using the exchange rate as we did in problem
11 and then divide by population to get real GDP per person.)
66 The Process of Economic Development
13 Now, calculate India‚Ä™s 2001 purchasing power parity GDP valuing India‚Ä™s 2001 output
using 2001 US prices (Equation 2.5 from the text). Is this value larger or smaller than
India‚Ä™s total nominal GDP you calculated in problem 11 using the official exchange rate
conversion rate? What does the PPP GDP value you have calculated mean compared to
the value you calculated in problem 11?
14 If income were perfectly equally distributed, the poorest 20 per cent of individuals
or families would receive 20 percent of the economy‚Ä™s total income, the second
poorest 20 percent of individuals or families would receive 20 percent of the economy‚Ä™s
total income, and so on, and the richest 20 percent of individuals or families would
receive 20 percent of the economy‚Ä™s total income. In such a case of perfect equality, each
individual‚Ä™s or family‚Ä™s income would be equal to per capita income. Everyone would
receive the same income. However, in the real world income is distributed unequally
so that a country‚Ä™s per capita income is not necessarily telling us much about the
living standard of all the population. This problem highlights how the average income
measure may not give a full picture of the living standards of an economy if income
distribution is highly skewed.
In Uganda, the poorest 20 percent of the population received 6.2 percent of
total income while the richest 20 percent received 63.5 percent of total income
of the country. Given that income per capita for Uganda was $310 in 2004, what
was the average income per capita of the poorest 20 percent and of the richest 20
percent of the population? What was the average income of the remainder of the
15 Let‚Ä™s consider what the Kuznets curve looks like for an individual country. The data
may not be easy to find, but look for Gini coefficient values, or as an alternative, the
ratio of the income of the richest 20 percent of the population to the income of the
poorest 20 percent as a proxy measure of inequality, for a number of years. The World
Bank website has such data. Then plot the data you find against income per person
for the same years on the vertical axis of a graph. Does it look like the Kuznets curve
in Figure 2.1? If not, what does the curve you have found tell you about the relation
between income inequality and growth in that country? Does the data you plot show a
threshold level of income as in Figure 2.1?
16 Focus 2.5 considered the viability of China‚Ä™s continued economic growth given the
heavy dependence on exports for growth and the relatively small proportion of impetus
to growth from consumption. Find some recent data on China‚Ä™s economic growth and its
exports. Has there been a slowdown in economic expansion? Have exports continued to
grow as rapidly as in the past? You should be able to find some data at the World Bank
1 Further, it is a strongly held belief of many economists that economic growth in capitalist
societies occurs via a trickle-down process. With economic growth and an expansion of soci-
ety‚Ä™s total income, there is assumed to exist a more or less automatic dispersion of the benefits
of this growth to all income classes of society. While it is admitted that the incomes of the
wealthiest in society perhaps grow most rapidly, those at lower income levels are presumed
to benefit also from economic expansion as income ‚Äútrickles down‚ÄĚ the income pyramid.
This may occur via the provision of new and better jobs that result from the increased invest-
ment undertaken by higher-income individuals who finance such ventures, given their higher
Measuring economic growth and development 67
Thus, one view is that income inequality has a functional purpose in capitalist economies
in that it is higher-income individuals who are likely to save a larger portion of their incomes
relative to lower-income recipients. It is from this pool of savings that the loanable funds for
investment arise. Much like a boat, all of whose passengers are lifted together on a rising tide,
it is suggested that greater economic growth benefits all, or certainly the great majority, of
the members of society via the automatic mechanism of trickle-down growth. This, of course,
is only a theory; the important question is whether this process works as described in each
2 Poorer nations tend to have less dependable estimates of their national income for one obvious
reason: collecting data is expensive, and for economies already facing the constraint of limited
financial and human resources, the gathering and evaluation of economic data is likely to be done
in a manner that is less than desirable and certainly less than would be optimal. To develop strate-
gies that contribute to development, however, there is a compelling need for reliable and timely
statistical data concerning the objective reality in less-developed nations. In fact, one is tempted to
state that the effort put into collecting dependable and timely statistics, making such information
available to the public, and in analyzing such data is one measure of a country‚Ä™s commitment to
doing something positive about its future development.
3 Gross national income (GNI) is the same concept as the former gross national product (GNP) termi-
nology. The use of the term GNI, however, is an improvement as it more accurately conveys what
is being measured, i.e. national income available foe use.
4 There are some further caveats worth mentioning when calculating real GNI or real GDP and using
these values to judge an economy‚Ä™s progress over time. The farther apart in time the comparisons
of income are, the less meaningful they are likely to be. For example, some goods and services may
no longer be produced in later years, while new goods and services can enter the production stream.
Thus, price indices and the deflating technique become less reliable for comparing real output over
The issue of the quality of the Qis is not captured by the price index adjustment either. Prices
may rise with quality improvements over time (some couture clothing) or they may fall (as with
computers and HDTVs), so some price changes reflect not inflation or deflation but rather differ-
ences in product quality. These differences will be lost in adjusting for real income with a price
deflator and such improvements in quality, when they occur, must be reintroduced via other
Still, while always being cognizant of these weaknesses in calculating real GDP or GNI figures,
if the years being compared are not too far apart in time, calculations as in equations 2.3 and 2.4
can be taken as reasonable approximations for estimating changes in real output and income over
5 Other deflators might be used, for example, the GDP price deflator. In the case of Mexico, for
example, using the GDP deflator would have resulted in a very slightly greater drop in income per
6 An equivalent and perhaps easier method for calculating the average income of the lowest and
highest quintiles is to remember that the $5,503 GNI per capita figure for Brazil shown in Table
2.1 would be the actual income of all individuals in Brazil only if income were perfectly equally
distributed. However, the richest 20 percent actually received 3.055 times their equality share of
total GNI (their 61.1 percent actual share of total income √· their 20 percent equality share). Thus
the per capita income of the richest 20 percent in Brazil can be calculated as 3.055 √— $5,503 =
$16,812, very close to the figure in the text, the difference being due solely to rounding. For the
poorest 20 percent, their per capita income is but 0.14 their equality share (the actual 2.8 percent
of total income received √· a 20 percent equality share), for a per capita income of 0.14 √— $5,503 =
$770 for the poorest fifth of the population.
7 A Gini coefficient of 0 would indicate perfect equality of income. A Gini coefficient of 100 would
indicate perfect inequality in the distribution of income, i.e. one person or family receiving 100
percent of the economy‚Ä™s total income and everyone else receiving nothing.
8 Any good macroeconomics book will have the details on this problem. On the GPI, for which
calculations have been made for the United States, see Cobb and Halstead (1994).
9 Arbitrage is the process in which goods are purchased in one market to be resold in another at
a higher price and with a known and certain profit. For example, if the price of a bar of Nestl√©‚Ä™s
chocolate in India sells for the equivalent of $1.50, while it sells for the equivalent of $2.50 in the
68 The Process of Economic Development
Philippines, it would be to the advantage of profit-maximizing traders to purchase Nestl√© bars in
India and resell them in the Philippines, as long as the transaction costs of doing so ‚Ä“ transportation,
tariffs, etc. ‚Ä“ are less than $1 per bar of chocolate, since there would be profit to be made on such a
This process of arbitrage tends to bring the prices of the traded goods in line between countries.
How does this happen? In our example, the price of Nestl√©‚Ä™s bars in India would tend to rise with
the increased demand as traders bought in that market in pursuit of the profits of arbitrage, while
the price of chocolate bars would tend to fall in the Philippines as a result of the increased supply
as chocolate bars would be introduced to that market.
Ultimately, an equilibrium price would exist in both countries at which all the opportunities for
arbitrage and the making of a sure profit from trade would have been exhausted. For this reason,
the prices of traded goods are expected to be very similar across economies.
10 See Streeten (1979) and Streeten et al. (1981) on the basic needs approach and Morris (1979) for
the original contribution to the creation of the PQLI measure.
11 The actual calculation of the HDI value for any country is based on that country‚Ä™s deviation from
the maximum values for each component of the index: a maximum life expectancy of eighty-five
years at birth; 100 percent literacy and 100 percent combined (primary and secondary) school
enrolment; and a maximum PPP GDP per capita income of $40,000.
The HDI thus measures the relative position of an economy compared to the maximum (and
minimum) levels of achievement. In other words, the HDI is a measure of how far away a country
is from the current maximum achievable values on the selected variables that enter the HDI (see
any recent edition of the UNDP Human Development Report for the methodology and an example
of how the HDI is determined; Appendix 2B illustrates such a calculation).
12 Until 1994, the HDI was calculated and compared with GNI per capita evaluated at official
exchange rates. Beginning in 1995, however, the HDI has been calculated using a PPP measure of
GDP per capita within the index. It is thus not strictly correct to compare the HDI values calculated
for years after 1992 (the HDI values in the 1995 Human Development Report), with the HDI values
calculated for 1987‚Ä“91. Further changes in the methodology for calculating the HDI are possible in
13 It is perhaps worth remembering an important axiom of economic theory which states that there are
an infinite number of efficient, i.e. Pareto optimal, outcomes for both the production and distribu-
tion of goods and services among the members of any society. No single distribution of income or
resources is ‚Äúbetter‚ÄĚ than another. As can easily be demonstrated with an Edgeworth-Bowley box
diagram, any initial distribution of wealth will, under conditions of perfect competition and free
exchange, generate a locally efficient level of production and distribution of society‚Ä™s goods and
services, with trade resulting in the contract curve being attained.
Thus the distribution of wealth and income amongst society‚Ä™s members is a choice variable
open to economies, since the distribution issue is independent of the issue of efficiency. Any initial
distribution can be efficient. For an excellent exposition of this issue, see Bator‚Ä™s (1957) classic
14 See Anand and Kanbur (1993) for a review of the literature. These authors argue that the best
empirical relation between income growth and equity or development is actually the opposite
of the Kuznets hypothesis! This contradictory result illustrates one of the problems in studying
economics. There often are competing models that purport to explain some particular phenom-
enon, frequently based on extremely complicated mathematical and statistical analyses, done by
equally competent and respected investigators, that nonetheless come to diametrically opposed
and often irreconcilable conclusions. How does one choose between such competing theories
when compelling empirical evidence can be mustered supporting alternative theories? This is an
excellent question for class discussion!
15 It is not just governments that today can affect the level of inequality and reduce poverty, even
at lower income levels than in the past. There are also a growing number of non-governmental
organizations, or NGOs, which operate in the less-developed nations and which often have
as their primary objective the alleviation of poverty. These range from large and relatively
well-financed groups like Oxfam, the International Red Cross and Crescent, CARE, Caritas,
World Vision to small, region- and often country-specific groups, such as the Voluntary Action
Measuring economic growth and development 69
Alkire, Sabrina. 2002. ‚ÄúDimension of Human Development,‚ÄĚ World Development 20: 181‚Ä“205.
Anand, S. and Kanbur, S.M. 1993. ‚ÄúInequality and Development: A Critique,‚ÄĚ Journal of Development
Economics 41: 19‚Ä“43.
Anderson, Jonathan. 2007. ‚ÄúSolving China‚Ä™s Rebalancing Puzzle,‚ÄĚ Finance & Development 44
Bartelmus, Peter. 1994. Environment, Growth and Development. London: Routledge.
Bator, Francis. 1957. ‚ÄúThe Simple Analytics of Welfare Maximization,‚ÄĚ American Economic Review
Birdsall, Nancy and Richard Sabot. 1994. ‚ÄúInequality as a Constraint on Growth in Latin America,‚ÄĚ
Development Policy, Newsletter on Policy Research by the Inter-American Development Bank
Cobb, Clifford and Ted Halstead. 1994. The Genuine Progress Indicator. San Francisco, CA: Re-
defining Progress (September).
Dasgupta, Partha and Karl-G√∂ran M√¤ler. 1995. ‚ÄúPoverty, Institutions, and the Environmental Resource-
Base,‚ÄĚ Chapter 39 in Jere Behrman and T.N. Srinivasan (eds.), Handbook of Development Economics,
vol. IIIA. Amsterdam: Elsevier Science.
Dietz, James L. and Louise Gibson. 1994. ‚ÄúWhat is Development? The Human Development Index, a
New Measure of Progress?,‚ÄĚ mimeo, California State University, Fullerton.
Elliott, Jennifer A. 1994. An Introduction to Sustainable Development. London: Routledge.
Morley, Samuel A. 1995. Poverty and Inequality in Latin America. Baltimore, MD: The Johns Hopkins
Morris, Morris D. 1979. Measuring the Condition of the World‚Ä™s Poor: The Physical Quality of Life
Index. New York: Pergamon Press.
Redclift, Michael. 1987. Sustainable Development: Exploring the Contradictions. London:
Stewart, Frances. 1995. Adjustment and Poverty. London: Routledge.
Streeten, Paul. 1979. ‚ÄúFrom Growth to Basic Needs,‚ÄĚ Finance and Development 16 (September):
‚Ä”‚Ä” et al. (eds.). 1981. First Things First. Oxford: Oxford University Press.
UNDP (United Nations Development Programme). 1993. Human Development Report 1993. Oxford:
Oxford University Press.
‚Ä”‚Ä”. 1995. Human Development Report 1995. Oxford: Oxford University Press.
‚Ä”‚Ä”. 2001. Human Development Report 2001. Oxford: Oxford University Press.
‚Ä”‚Ä”. 2006. Human Development Report 2006. Oxford: Oxford University Press.
US Department of Commerce. 1994. Survey of Current Business 74 (July).
‚Ä”‚Ä”. 1995. Survey of Current Business 75 (August).
WCED (World Commission on Environment and Development). 1987. Our Common Future. Oxford:
Oxford University Press.
World Bank. 1992. World Development Report 1992. Oxford: Oxford University Press.
‚Ä”‚Ä”. 1995. World Development Report 1995. Oxford: Oxford University Press.
‚Ä”‚Ä”. 2000. World Development Report 2000/2001. Oxford: Oxford University Press.
‚Ä”‚Ä”. 2001. World Development Indicators. On-line.
‚Ä”‚Ä”. 2007 World Development Indicators. On-line.
Appendix 2A: Calculating the Gini coefficient
To understand how the Gini coefficient is calculated, it is helpful to look at a Lorenz curve which
provides a graphical representation of a nation‚Ä™s income distribution. Figure 2.1A shows a simple
Lorenz curve drawn from the following hypothetical income distribution figures.
70 The Process of Economic Development
Income distribution, by quintiles
Share of total income Cumulative percent of total income
4% of total GNI 4%
Poorest 20% of families
8% of total GNI 12% (= 4% + 8%)
Second 20% of families
11% of total GNI 23% (= 4% + 8% +11%)
Third 20% of families
18% of total GNI 41%
Fourth 20% of families
59% of total GNI 100%
Richest 20% of families
100% of total GNI
In Figure 2.1, the box measures the percent of population on the horizontal axis and the percent of total
income received by each percent of the population on the vertical axis. The diagonal in Figure 2.1 is a
reference ‚Äúline of equality.‚ÄĚ Any point along it would mean that X percent of the population received
exactly X percent of total income (where X could be any number between 1 and 100). Along the ‚Äúline
of equality,‚ÄĚ for example, ten percent of the population would be receiving 10 percent of society‚Ä™s
total income; 40 percent of the population would be receiving 40 percent of total income; and so on.
The diagonal provides a referent for visually comparing and precisely measuring the dispersion of the
actual income distribution of a nation with what would be a perfectly equal distribution of income
among all members of society.
By plotting the ‚Äúactual‚ÄĚ aggregate values of income received against the quintiles of population for
our hypothetical example, the bowed Lorenz curve in Figure 2.1A can been drawn. Very roughly, the
further away the Lorenz curve is from the line of equality, the greater the degree of income inequality.
From the Lorenz curve diagram, the Gini coefficient can be calculated. It is equal to the area A (the
area between the Lorenz curve and the diagonal ‚Äúline of equality‚ÄĚ) divided by the total area (A + B) of
the triangle below the ‚Äúline of equality.‚ÄĚ Thus the Gini coefficient is equal to A/(A + B).
0 100 %
Figure 2.1A A Lorenz curve of income distribution.
Measuring economic growth and development 71
appendix 2B: Calculating the HDI index
With the 1995 issue of the Human Development Report, the HDI is calculated as a weighted average
of educational attainment, life expectancy at birth, and income. All components of the index measure the
relative distance between a country‚Ä™s achievement and what is possible. Thus,
actual value of xi ‚ą’ minimum value of xi
index value =
maximum value of xi ‚ą’ minimum value of xi
Educational attainment (E) is measured as a combination of two indices, one for adult literacy (minimum
value = 0 percent; maximum value = 100 percent) and of the combined primary, secondary, and tertiary
enrolment ratios (minimum = 0 percent; maximum = 100 percent) with the following weights:
E = ‚…” adult literacy rate (L) + ‚…“ combined enrolment ratio (C)
Life expectancy (L), for purposes of calculating the HDI index, has a minimum value of 25 years and
a maximum value of 85 years.
Income (Y) enters the HDI as a log value of the purchasing power parity (PPP) measure of GDP per
capita. As the UN Development Programme puts it (UNDP Human Development Report 2003: 341),
‚Äúincome serves as a surrogate for all the dimensions of human development not reflected in a long and
healthy life and in knowledge.‚ÄĚ PPP GDP per capita income beyond $40,000 adds nothing further to
the Y index.
The following shows how the HDI for 2001 was calculated for Albania. The basic data needed
to calculate L, E, and Y are: Albania‚Ä™s life expectancy was 73.4 years in 2001; the literacy rate
was 85.3 percent; the combined enrolment rate was 69 percent; and PPP GDP per capita was
The following HDI calculation for Albania is adapted from UNDP, Human Development Report
1 The life expectancy index, L, was determined as follows:
Using the first equation above, and the minimum and maximum values for life expectancy, the
life expectancy index, L, is equal to
73.4 ‚ą’ 25
L= = 0.807
85 ‚Ä“ 25
2 To determine the education index, E, a two-step procedure is necessary, again using the second
equation in this appendix and the specific values for Albania for literacy and the combined enrol-
2a First, the adult literacy index, A, component of educational attainment was computed as
85.3 ‚ą’ 0
A= = 0.853
100 ‚Ä“ 0
2b Second, the combined enrolment index, C, the other component of education attainment, was
determined as being
69 ‚ą’ 0
C= = 0.690
100 ‚Ä“ 0
72 The Process of Economic Development
Thus from 2a and 2b, the educational attainment index, E, was equal to
E = 2/3L + 1/3C = 2/3(0.853) + 1/3(0.690) = 0.798.
3 The income index, Y, for Albania and using Albania‚Ä™s PPP GDP per capita income shown above
was equal to
log(3,680) ‚ą’ log(100)
Y= = 0.602
log(40,000) ‚ą’ log(100)
The value of the HDI is then calculated as a weighted average of L, E, and Y, with each
component having a value of 1/3. Since each of these is a ‚Äúdeprivation‚ÄĚ measure, what is being calcu-
lated is the gap between a particular country‚Ä™s achievement level and what might be attained.
Thus, the HDI for Albania in 2001 is a simple weighted average of L, E, and Y, computed as
HDI = 1/3L + 1/3E + 1/3Y
= 1/3(0.807) + 1/3(0.798) + 1/3(0.602) = 0.735.
3 Development in historical perspective
after reading and studying this chapter, you should better understand:
‚ÄĘ why and how colonialism left a lasting legacy in developing nations;
‚ÄĘ the difference between semi-feudal/semi-capitalist social structures and capitalist
‚ÄĘ the de-industrialization impact of colonialism and the biased nature of colonial
‚ÄĘ the new role credit played in the construction of neocolonial structures in the
‚ÄĘ the nature and importance of the terms of trade;
‚ÄĘ economic dualism and its impact on colonial and post-colonial society;
‚ÄĘ how to apply the concept of path dependency to post-colonial situations;
‚ÄĘ the differential impact of early and mature colonialism;
‚ÄĘ the concept of colonial drain.
economic development demands and entails profound cultural change, including, often,
transformation of the political system, of individual behavior norms, of the culture of work
and production, and most fundamentally, modifications in the manner in which society
confronts, moulds, propels, and adapts itself to the requirements of technological progress
that are the fount of economic growth and human development. Anyone studying the process
of economic development must appreciate the wide-ranging cultural factors at work in any
society. Failure to do so can result in a narrow and mechanistic interpretation of developing
societies and the adoption of incomplete policy prescriptions which will, at best, diminish the
effectiveness of efforts to achieve further progress.
At times, economists and others directly concerned with the process and problems of
economic development have devoted too few of their efforts to understanding the historical
conditions which have led to economic backwardness and underdevelopment. This failure
may well arise from the fact that orthodox economists have generally been trained in the
science of market behavior, with the assumption that human nature consists of, as Adam
Smith maintained, a ‚Äúnatural propensity to truck, barter and exchange.‚ÄĚ Taking such a
perspective too literally, however, can lead to the view that the peculiarities and specificities
74 The Process of Economic Development
of any country‚Ä™s history can be disregarded. The early development economists, to be
discussed in Chapters 5 and 6, did not view history as insignificant, and we would argue
that many of their insights as applied to the less-developed nations were richer as a
The social conditions under which production takes place are often significantly different
in the developing world from the advanced nations. Today the developing world incorpo-
rates, in shifting proportions, mixtures of neo-feudal and peasant social and productive
forces, combined with some of the most advanced components of early twenty-first-century
capitalist production methods (this division is sometimes referred to as economic dualism).
When development economics emerged as a separate discipline in the postwar period, the
primary arena of its application was Western Europe and, to a much lesser degree, Japan.
In both areas, economic policy had achieved tremendous success, as the Marshall Plan
and US military assistance funds were pumped into Europe. Those economies responded
rapidly to this stimulus, quickly regaining and then overtaking their past levels of develop-
ment. In the early 1950s, then, fresh from this experience, the task of promoting economic
development in the newly independent, less-developed nations did not appear daunting to
The relatively easy success of europe and Japan in spurring output and employment
formed the crucible for development thinking regarding the less-developed regions. One
noted text of the 1950s, Benjamin Higgins‚Ä™s Economic Development, recounts an episode
from the postwar period that seemed to justify such optimism. A small Pacific island commu-
nity was overrun by US military personnel, resulting in the rapid transformation of their
culture. Higgins drew a conclusion from this that was widely shared in the 1950s and early
This experience suggests that an almost complete transformation of a society can take
place within a few years if the external ‚Äúshock‚ÄĚ to the society is powerful enough.
(Higgins 1959: 312)
Higgins shared with most development economists of that time a strong presumption that
whatever the nature or magnitude of the social, psychological, political, or historical obsta-
cles inhibiting economic growth and human development, these barriers could be quickly
overcome. The enduring structural distortions of economic dualism, deeply embedded in
society because of historical factors which had given rise to economic retardation, were not,
however, adequately appreciated at the time.
The 1960s were cast as the First Development Decade by the United Nations, and already
Higgins‚Ä™s ‚Äúfew years‚ÄĚ for promoting development had been lengthened to ten, but the impli-
cation was essentially the same: economic backwardness would yield quickly to the exper-
tise of the development specialists and to informed development advice.
Yet, after that First Development Decade had ended, the level of world poverty and despair
had receded only marginally. In some nations, the standard of living had declined or remained
roughly the same. In those nations where great leaps in overall economic performance had
been achieved, such as Brazil, too often aggregate success had been accompanied by a lower
standard of living for a significant portion of the population, as economic inequality wors-
ened even as total output expanded. Few keen observers believed the Brazilian situation
merely reflected the lead-up to the threshold income level suggested by Kuznets‚Ä™s inverted-U
hypothesis, and time has proved them correct on this point. By the 1970s, much of the opti-
mism about how quickly world poverty might be overcome had been muted, but development
Development in historical perspective 75
economists only rarely turned their attention to the study of historical and cultural factors to
try to broaden their understanding of the stubborn persistence of world poverty. In the early
twenty-first century, however, there has been a significant change in perspective, with some
development economists focusing on the question of the legacy of colonial practices and
institutions. Let us briefly turn to a consideration of some of these issues as they relate to
the less-developed nations, most of which, it is worth remembering again, reached political
independence only after the Second World War.
The origins of economic development
Sustained increases in output and income per capita over time are of relatively recent
historical origin. For much of known human history, population and total output tended to
grow at about the same rate, so that per capita income remained roughly constant. Lloyd
Reynolds has called this the period of extensive growth (Reynolds 1986: 7‚Ä“9). What this
meant was that for many centuries, until the early sixteenth century, the trend line of per
capita production and income rose only very slowly, as shown in Figure 3.1. Most produc-
tion was rurally based, there were few large urban settlements, and most people lived
on and from the land, sometimes selling small surpluses in the village marketplace for
other goods. Artisan and industrial-type products, such as textiles, and services, such as
transportation, were also produced in the countryside, but on a very small scale. Besides
agricultural production, families also produced non-agricultural goods, from clothing to
cutlery to farm implements to soap and alcoholic beverages, primarily for family own-
Production methods were relatively simple during this extensive period of growth, and
technology was very primitive. Such technology, as Reynolds (1986: Chapter 1) argues,
was not static, however. Technological methods were constantly adapting to the growing
demands of population growth, to changing land conditions, and to new crops and strains
of seeds. Such technological change tended to be sufficient to keep pace with population
Total Western Offshoots
0 250 500 750 1000 1250 1500 1750 2000 2250
Figure 3.1 Historical growth trend of per capita income.
76 The Process of Economic Development
expansion and to maintain income per person roughly constant over long periods. Annual
aggregate output during the period of extensive growth was subject to periodic ups and
downs, but that was primarily due to exogenous factors, such as the weather, wars, and other
The major ‚Äúturning point‚ÄĚ for world economic progress occurred with the transition from
feudal production and social organization to the emergence of capitalist forms of production
in Europe. Feudalism, an agricultural-based and hierarchical system of production organized
around manors and based on the labor of serfs tied to the land, began to break down in the
Middle Ages, especially in England, where the Industrial Revolution heralded the advent
of the capitalist factory system as a new means of social and economic organization. With
incipient capitalism, the purpose of production changed from survival and stasis to the pursuit
of ever-increasing profit. Capitalist production was based fundamentally on the application
of new knowledge and ever-greater quantities of physical capital that could produce more,
at lower cost, with the same or fewer inputs to production. Industrial capitalism established
the foundation for intensive production methods, which for the first time in history created