Content uploaded by David Dawe
Author content
All content in this area was uploaded by David Dawe on Mar 07, 2018
Content may be subject to copyright.
Food Policy 27 (2002) 355–370 www.elsevier.com/locate/foodpol
The changing structure of the world rice
market, 1950–2000
D. Dawe
∗
International Rice Research Institute, Los Ban
˜os, Laguna, Philippines
Received 1 August 2001; received in revised form 1 June 2002; accepted 28 July 2002
Abstract
The world rice market has been unstable for much of its post-world war II history, with
prices volatile and the availability of supplies uncertain. These characteristics, exemplified by
the world food crisis of the mid-1970s, influenced domestic price and production policies in
a number of Asian countries. However, the structure of the world rice market has evolved
and changed during the past 50 years. This paper identifies three distinct phases in the history
of the market based on trends in the level and stability of production and the trade orientation
of major exporters. The level and stability of production increased steadily over the entire
period. Exporters have generally been active in the world market, with the period 1965–1981
being an important exception. Since the mid-1980s, prices have been low and quite stable,
and the patterns identified in the paper suggest that prices will remain so in the future. The
policy implication is that Asian rice importers can afford to rely more on the world market
than was warranted in the past.
2002 Elsevier Science Ltd. All rights reserved.
Introduction
Nearly 70% of the world’s poor live in Asia (UNDP, 1997), a continent where
rice is by far the dominant staple food. The world rice market has traditionally been
a particularly unstable commodity market (Falcon and Monke, 1979–1980; Monke
and Pearson, 1991; Siamwalla and Haykin, 1983), where the availability of supplies
is uncertain and prices are highly variable. There are several characteristics that made
∗
Present address: DAPO Box 7777, Metro Manila, Philippines. Tel.: +632-845-0563; fax: +632-
891-1292.
E-mail address: d.dawe@cgiar.org (D. Dawe).
0306-9192/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved.
PII: S0306 -9192(02)00038-6
356 D. Dawe / Food Policy 27 (2002) 355–370
the world rice market more unstable than world markets for other grains such as
wheat and maize: the geographic concentration of rice production, a thin and frag-
mented world market with high transaction costs in trading, low domestic price elas-
ticities of demand, and relatively low world stockholdings (Jayne, 1993).
Price variability on the world rice market has profoundly affected rice policies
throughout Asia via at least two main channels. First, governments pursued policies
such as expansion of irrigation and implementation of fertilizer and credit subsidies
in order to increase domestic rice production and approach or attain self-sufficiency,
thus reducing reliance on the world market. Secondly, most rice producing countries
have pursued price stabilization policies in order to mitigate the impact of changing
world prices on domestic prices when some exposure to the world market was too
costly to avoid (Dawe, 2001). However, the world rice market has undergone contin-
ual evolution and change during the past half-century, and it is not the same world
market today that it was in the past. These changes have affected both the level and
the variability of world rice prices.
One notable feature of its evolution has been the low level of world rice prices
from the mid-1980s to the present. The transition to this new period of lower prices
was not gradual, but was instead concentrated within a period of 4 years, 1982–
1985. From 1950 to 1981, world rice prices averaged about US$934/ton (constant
year 2001 prices), with no distinct trend over time (Fig. 1). Then, from 1981 to
1985, world prices plunged precipitously, falling by 62% in just 4 years. From 1985
to 1998, prices averaged about US$355/ton, without any distinct time trend during
this latter period. Most recently, prices have plunged once more, falling 48% between
1998 and 2001 to a new low of US$173/ton.
Just as the level of prices has changed considerably during the past 50 years, so
has the variability of prices. Three distinct phases can be identified. From 1950 to
1964, prices were relatively stable, as measured by either the root mean square error
Fig. 1. Inflation-adjusted world market rice prices, 1950–2001, 100Bs, FOB Bangkok. Dashed lines
indicate time trends for 1950–1981 and 1985–1998.
357D. Dawe / Food Policy 27 (2002) 355–370
(RMSE) from a regression of real price versus a time trend, or the average absolute
value of the percentage price change from year to year (Table 1). From 1965 to
1981, prices were substantially more variable. This period includes the world food
crisis of 1973–1975, an event that was important in shaping the attitude of Asian
policymakers toward instability. Finally, from 1985 to 1998, world prices were rela-
tively stable once again. (Price stability in a very short time span is not easy to
measure in a meaningful manner, so the transition period from 1982 to 1984 is
ignored in terms of measuring price variability, as is the recent decline from 1999
to 2001.) The conclusion that rice prices have been more stable during the past 20
years (albeit punctuated with two sharp, and apparently permanent, declines) is con-
sistent with the conclusion in Sarris (2000).
Thus, to summarize, the pre-Green Revolution period from 1950 to 1964 was
characterized by high and stable prices. The Green Revolution period from 1965 to
1981, when modern fertilizer-responsive varieties were adopted in many countries,
was a period of high and unstable prices (1965–1981).
1
The years 1982–1984 marked
a short transition to a post-Green Revolution regime of low and stable prices from
1985 to 1998 (see Table 2 for a summary of the characteristics of each period). Most
recently, from 1999 to 2001, prices have plunged once again.
The main objective of this paper is to explain these shifts in the behavior of prices
in terms of technological changes and political disturbances that have affected rice
production and trade. For example, why were prices relatively unstable in the middle
of the past half-century, yet relatively stable at other times? Perhaps more important,
can these factors provide insights into the future of the world rice market in terms
of the level and stability of prices?
Table 1
Stability of world market rice prices
Period Average absolute value of annual RMSE (from regression of real price
price changes (%) vs. time trend) normalized by average
price
1950–1964 7 0.078
1965–1981 24 0.341
1985–1998 11 0.111
Source of raw data: IMF (2002).
1
When the words ‘unstable’and ‘instability’are used in this paper, the meaning is synonymous with
a high level of variability from trend. Similarly, the words ‘stable’and ‘stability’are synonymous with
a low amount of variability from trend.
358 D. Dawe / Food Policy 27 (2002) 355–370
Table 2
Characteristics of the world rice economy, selected periods
Period 1950–1964 1965–1981 1985–1998
Pre-green revolution Green revolution Post-green revolution
Production structure
Level (per capita) Low Medium High
Variability High Medium Low
Trade structure
Number of commercially oriented Many Few Many
exporters
Prices
Level High High Low
Variability Low High Low
Average yield, world (ton ha
ⴚ1
)1.88 2.42 3.54
Modern variety adoption as percentage of planted area, end of period
Bangladesh 0 22 61
India 0 48 74
Indonesia 0 61 81
Myanmar 0 53 61
Philippines 0 79 89
Thailand 0 13 16
Vietnam 0 17 89
Irrigated area as percentage of planted area, end of period
Bangladesh 5 13 32
India 37 42 50
Myanmar 14 18 30
Philippines 30 48 65
Thailand 26 23 20
Vietnam –41 52
Sources of information or data: production structure, trade structure; prices: discussion in text; yield:
FAOStat on-line electronic database (2002); modern variety adoption, irrigated area: IRRI World Rice
Statistics electronic database (2001); a dash (–) indicates data not available.
Trends in the level and stability of Asian rice production
Trends in the level of per capita production
Trends in the level and stability of Asian rice production go a long way toward
explaining the trends in world rice prices noted above. The most striking example
is that the plunge in world prices from 1982 to 1984 coincided with a sharp increase
in per capita rice production in Asia (Fig. 2). During those 3 years, per capita pro-
duction reached a new level of roughly 160 kg paddy (104 kg milled rice) per capita,
and this level has been approximately maintained ever since. The proximate cause
of the price decline was the achievement of rice self-sufficiency by Indonesia in
1984, which in the decade prior to that was by far the world’s largest rice importer.
From 1981 to 1984, rice production in Indonesia increased 16% in a span of just 3
359D. Dawe / Food Policy 27 (2002) 355–370
Fig. 2. Per capita Asian rice production, 1951–2000. Note: Asia includes the following 17 countries:
Bangladesh, Cambodia, China, India, Indonesia, Japan, Korea (North), Korea (South), Laos, Malaysia,
Myanmar, Nepal, Pakistan, Philippines, Sri Lanka, Thailand, and Vietnam.
years, and Indonesia’s abrupt exit from the world rice market undoubtedly had both
real and psychological effects on world prices. But, rice production surged in many
other countries at this same time, and it was the combined growth in many countries
that allowed world prices to stay at this new low level for the next 15 years, even
when Indonesia returned to the world market in the 1990s. For example, production
in China increased rapidly during this period due to the economic reforms begun in
1978 (Huang and Rozelle, 1996; Zhang and Carter, 1997). India and Vietnam also
saw rapid surges in production during this time. These four countries are some of
the world’s leading rice producers, so the impact on Asian production was substantial
(Table 3). At the same time that Asian rice production surged, the Thai baht was
also devalued by nearly 25% during this time, from about 20.70 at the beginning of
1981 to about 27.10 by the end of 1984. The lower value of the baht raised the
profitability of rice production in Thailand and exports surged from an average of
2.4 million ton during 1978–1980 to 4.4 million ton during 1984–1986.
Although the magnitude of the surge in per capita production from 1982 to 1984
was unprecedented, per capita production had been increasing steadily during the
previous three decades. Why did these increases not have a similar depressing effect
on world prices? The most likely reason is that Asian countries were much poorer
in this earlier period, which meant that the income elasticity of demand for rice was
still positive (Timmer et al., 1983). Thus, growth in rice production had to keep pace
not only with population growth, but also with income growth. In other words,
increased per capita production was necessary to keep world prices constant in real
terms. As per capita incomes in Asia reached higher levels, however, the income
elasticity of demand for rice declined to zero in many countries, and even became
negative for some. This is reflected in declining trends in per capita rice consumption
for Japan, South Korea, Malaysia and Thailand. In China, Indonesia, and the Philip-
360 D. Dawe / Food Policy 27 (2002) 355–370
Table 3
Leading rice producing countries, selected periods, with average level of production during each period
(in millions of tons of paddy)
Country 1950–1964 1965–1981 1985–1998
China 73.5 120.4 184.8
India 44.8 64.7 111.1
Indonesia 11.3 21.3 45.5
Bangladesh 12.6 17.6 26.1
Vietnam 7.6 10.3 21.3
Thailand 9.3 14.3 20.5
Myanmar 6.4 9.2 15.4
Japan 14.2 16.0 12.9
Brazil 4.2 7.5 9.7
Philippines 3.4 5.8 9.7
Korea 4.0 6.2 7.4
World 215.0 331.0 521.6
As percent of world total (%):
Top 5 73 73 75
Top 10 87 87 88
Sources of raw data: Palacpac (1977) for 1950–1960, FAOStat on-line electronic database (2002) for
1961–1998.
pines, per capita consumption has been approximately constant for at least the past
decade. Thus, constant levels of per capita rice production are now sufficient to keep
prices constant in real terms. This is consistent with the IMPACT model of the
International Food Policy Research Institute (IFPRI), which projects that world rice
prices will remain constant provided production growth roughly equals population
growth between now and 2020 (Rosegrant et al., 1995).
While trends in per capita rice production explain to some extent the sharp drop
in prices from the early to mid-1980s, this is probably not the whole story. World
market prices for wheat and maize also declined sharply in the 1980s, as did many
other commodity prices. In the case of maize, prices declined substantially even
though world production growth slowed substantially in the 1980s. Pindyck and
Rotemberg (1990) found evidence for excess co-movement of several commodity
prices, even after accounting for the effects of macroeconomic variables, and a simi-
lar phenomenon may occur for rice, wheat, and maize. Yet rice prices declined more
than wheat and maize prices during this period. Comparing the periods 1957–1981
and 1982–2000, the average ratio of rice to maize prices was 18% lower in the latter
period, while the ratio of rice to wheat prices declined by 25%. These declines in
the ratios of rice prices to prices for other grains suggest an important role for the
surge in per capita rice production.
Trends in the stability of per capita production
At the same time that the level of per capita production has increased during the
past half-century, it has also become more stable. The magnitude of year to year
361D. Dawe / Food Policy 27 (2002) 355–370
fluctuations in per capita production has been markedly lower in the past 15 years
than it was previously (Table 4). Prior to 1985, fluctuations in per capita production
of ⬎±3% were relatively common, occurring 22 times in the 29 years from 1952
to 1980. Since then, fluctuations of this magnitude have occurred just four times in
20 years. The average absolute value of annual changes in per capita production was
4.4% from 1952 to 1964, 3.7% from 1965 to 1981, and just 1.6% from 1985 to
1998. Other measures show a similar pattern.
This improvement in the stability of per capita production is most likely due to
two major technological influences, irrigation and pest control. In the initial phases
of the Green Revolution, irrigated area expanded rapidly due to large government
investments (Barker et al., 1985). This growth slowed substantially in the 1980s as
world rice prices declined (Hayami and Kikuchi, 1978). Furthermore, at the same
time that lower rice prices reduced the benefits of constructing new irrigation sys-
tems, the cost of irrigation projects began to rise as the easiest areas to irrigate had
already been exploited (Rosegrant and Svendsen, 1993). Despite these factors, how-
ever, the proportion of rice grown under irrigated conditions was higher in the mid-
1990s than in the early 1980s (Huke and Huke, 1997). Part of this increase in the
share of irrigated area in total area was due to the absolute decline in rice area
planted in marginal ecosystems such as deepwater and upland. The area planted to
deepwater and upland rice declined by more than 4 million ha between the early
1980s and mid-1990s, a fall of 25%. Reliable supplies of water have substantially
reduced production fluctuations relative to a situation where production relies solely
on the vagaries of rainfall.
Production stability has been further enhanced by the development of modern rice
varieties that have become progressively more resistant to pests and diseases. The
first modern high yielding semi-dwarf variety, IR8, suffered frequent attacks by dis-
eases and insect pests. IR36 was released in 1976, and this new variety incorporated
resistance to multiple pests and diseases. It was so successful that it is still grown
in many areas today. Other derivative varieties (such as IR64) also incorporated
much of this resistance, and this has proved to be a successful approach for stabiliz-
ing yields (Khush, 1995). Pesticide use has also increased over time, although it is not
clear that use of these chemicals has always been successful in reducing crop losses.
Table 4
Stability of Asian per capita rice production
Period Average absolute value of RMSE (from regression of per capita
annual changes (%) production vs. time trend)
normalized by average per capita
production
1951–1964 4.4 0.047
1965–1981 3.7 0.025
1985–1998 1.6 0.020
Sources of raw data: FAO (2002) for 1961–1998, Palacpac (1977) for 1951–1960. Time periods do not
correspond exactly to those in Table 1 because of incomplete data on production for 1950.
362 D. Dawe / Food Policy 27 (2002) 355–370
To summarize, both the level and the stability of per capita rice production in
Asia have increased substantially over time, and these developments would appear
to explain why world prices have been so low and stable for the past 15 years. Yet,
this cannot be the whole story. Why were world rice prices relatively stable from
1950 to 1964 in spite of very unstable production?
The fall and rise of commercially oriented rice exporters
An active world market: 1950–1964
For most of the 20th century, the major rice exporters in the world market have
been the nations of mainland Southeast Asia: Thailand, Burma, Cambodia, and South
(or southern) Vietnam. During the 1950s, Burma and Thailand dominated world rice
exports, with Cambodia also being an important player (Table 5). More important,
exports were a large share of domestic production for all of these countries (Fig. 3).
From 1950 to 1963, the average share of exports in domestic production was 40%
in Burma, 32% in Cambodia, and 24% in Thailand. South Vietnam was also a net
exporter in the 1950s, but its total exports were roughly only one-third of Cambod-
ia’s, and its share of exports in domestic production was also quite low.
During this period, agriculture dominated the economy of these poor countries,
and rice dominated the agricultural sector. Furthermore, because a large share of
production was exported to world markets, these countries needed to be commer-
cially oriented and participate actively in the world rice market as reliable suppliers.
2
Not only was rice an important share of the economy, but it was also a key source
of foreign exchange earnings and government revenue. For example, in Thailand
taxes on rice exports consistently accounted for more than 10% of all government
revenue in the years 1950–1965, and the share occasionally reached more than 25%
(Siamwalla, 1975). Thus, whenever there was a shortfall in Asian rice production,
one or more of these countries would typically step in to fill the breach and prevent
world prices from spiraling out of control. For example, there were large declines
in per capita production in 1954 and 1957 (due to a major La Nin
˜a event in
1954/1955 and a major El Nin
˜o event in 1957/1958; NOAA, 2000). To meet the
shortfall in 1954, Burma stepped into the market and exported a then record of 1.7
million tons, followed by another record of 2.0 million tons in 1955. In these 2
years, its share of exports in domestic production surged to 49%, relative to an
average of just 33% in the preceding 3 years.
In 1957, Asian per capita production fell by 4%. Aggravating matters, production
dropped sharply in the two major exporters, by 15% in Burma and 33% in Thailand.
Nevertheless, these two countries responded. Exports from Burma fell compared to
2
The phrase ‘commercial orientation’may have different meanings and may be difficult to measure.
For the purposes of this paper, a commercially oriented rice exporter is defined as an exporter with a
relatively high ratio of exports to domestic production, as this indicates that a relatively high level of
attention must be given to marketing supplies to international customers.
363D. Dawe / Food Policy 27 (2002) 355–370
Table 5
Leading rice exporting and importing countries, selected periods, with average annual level of
exports/imports during each period (in millions of tons of milled rice)
Country 1950–1964 Country 1965–1981 Country 1985–1998
Exports
Myanmar 1.52 USA 2.03 Thailand 5.12
Thailand 1.38 Thailand 1.75 USA 2.55
USA 0.80 China 1.62 Vietnam 1.60
China 0.69 Myanmar 0.59 India 1.42
Cambodia 0.47 Pakistan 0.58 Pakistan 1.29
Egypt 0.22 Italy 0.34 China 1.10
Italy 0.19 Egypt 0.32 Italy 0.61
Vietnam 0.16 Japan 0.30 Australia 0.47
Pakistan 0.12 Australia 0.18 Uruguay 0.39
Brazil 0.06 North Korea 0.17 Myanmar 0.29
World 6.05 World 9.23 World 16.72
As percent of world total (%):
Top 5 80 71 72
Top 10 93 85 89
Imports
Indonesia 0.70 Indonesia 1.02 Iran 0.77
Japan 0.66 Vietnam 0.62 Indonesia 0.75
India 0.58 South Korea 0.51 Brazil 0.68
Malaysia and 0.53 India 0.42 Saudi Arabia 0.47
Singapore
Sri Lanka 0.47 USSR 0.40 China 0.47
Hong Kong 0.31 Hong Kong 0.36 USSR/CIS 0.47
East Pakistan 0.25 Sri Lanka 0.35 Iraq 0.46
Cuba 0.20 Bangladesh 0.31 Philippines 0.42
West Germany 0.13 Malaysia 0.29 Malaysia 0.40
Philippines 0.11 Singapore 0.24 Senegal 0.40
World 6.10 World 9.30 World 16.05
As percent of world total (%):
Top 5 48 32 20
Top 10 65 49 33
Sources of raw data: Palacpac (1977) for 1950–1960, FAOStat on-line electronic database (2002) for
1961–1998. World exports and imports are not equal in any particular year in original data sources. USSR
is not included in 1950–1964 imports because of data availability constraints.
the record high of the previous year, but they still reached 43% of domestic pro-
duction in spite of the production shortfall. Thailand enacted quantitative restrictions
at this time (Siamwalla, 1975), but they were not very severe in their effect as exports
hit a near record of 1.5 million tons, with 40% of production being sent to the world
market. As a result of exports from these two countries, world prices barely budged
from the mid- to late 1950s.
Another major shortfall in per capita production occurred from 1959 to 1961, but
this was primarily due to the policies of the Great Leap Forward in China that led
to a dramatic collapse of production. Because of China’s isolation at that time, it
364 D. Dawe / Food Policy 27 (2002) 355–370
Fig. 3. Ratio of exports to domestic production, 1950–2000: Thailand, Myanmar, Cambodia, and Viet-
nam.
did not enter world markets to try and make up the deficit with increased imports.
Again, world prices did not increase substantially.
Exit of the exporters: 1965–1981
The situation had changed considerably by the mid-1960s, when a major El Nin
˜o
event led to a sharp fall of 6% in per capita Asian production in 1965. By this time,
Burma, the leading rice exporter in the 1950s, was well into a period of sharp decline
due to the restrictive policies of General Ne Win, who had seized power in a coup
in 1962. Exports were falling, and by 1967, they had declined to just 11% of domestic
production (Fig. 3) as world prices surged 30%. The proportion of Cambodia’s pro-
duction that found its way onto world markets was also in a period of decline (Fig.
3). South Vietnam banned exports in 1965, and it did not return to the world market
as an exporter until the late 1980s. Perhaps more surprisingly, even Thailand was
becoming less commercially oriented. By 1967, revenue from rice export taxes had
fallen to just 6% of total government revenues, and this declined to just 1% by 1971
(Siamwalla, 1975). Since the government was no longer so reliant on export tax
revenue, it had more flexibility to constrain exports in the interests of domestic price
stabilization. As a result, during the world price spike in 1967, Thailand raised its
rice premium (a form of export tax) to levels more than double than that of its
previous high.
By the early 1970s, the world market was even more unsettled. The proximate
cause of the world food crisis of 1973–1975 was a severe El Nin
˜o in 1972–1973,
followed by major La Nin
˜a events in 1973–1974 and 1975–1976. But this situation
was exacerbated considerably by the behavior of the traditional commercial rice
exporters. Thailand’s exports fell to just 10% of domestic production from 1973 to
1975, reaching their lowest point in the post-war period, and it banned exports com-
365D. Dawe / Food Policy 27 (2002) 355–370
pletely for a few months in 1973 (Slayton, 1999). By this time, Cambodia had joined
South Vietnam and completely exited the market, while Burma was also out for all
practical purposes (Fig. 3). The situation was so difficult for countries wishing to
import rice that the experience of this period continues to shape policies and attitudes
in the region to the present day.
Re-emergence of market stabilizers: 1985–2000
By the middle to late 1980s, this situation had changed considerably for the better.
Thailand’s commercial orientation has increased steadily since the world food crisis,
with exports now typically accounting for 40% of domestic production. Vietnam has
re-entered the world rice market, with exports accounting for approximately 20% of
domestic production during the past 5 years. The presence of Thailand and Vietnam
as commercially oriented rice exporters was a major factor in stabilizing the world
market in 1998 in the face of a major El Nin
˜o event that led to a fall in per capita
Asian rice production. Thai exports surged to a record of 6.4 million tons, while
exports from Vietnam jumped to 3.8 million tons. The devaluation of the Thai baht
played an important role in spurring exports, but Fig. 3 shows that Thailand is also
now much more commercially oriented compared to the mid-1970s. This is reflected
in current Thai policy, which now allows free trade in rice even if the consequence
is increased domestic prices (i.e. export taxes have been abolished). During the recent
financial crisis, domestic rice prices in Thailand rose by more than 50% in real terms
between November 1997 and January 1998. Despite this rapid rise, the government
allowed domestic prices to track world prices one for one. This is an important policy
change that adds considerably to the stability of the world market.
Myanmar and Cambodia are yet to return to the world market to play a major
role. Especially in Myanmar, more liberal domestic policies would increase domestic
production and exports substantially. Nevertheless, other exporters have emerged to
complement Thailand and Vietnam. The share of exports in production in Pakistan
has steadily increased during the past two decades, and is now typically 40%. China
and India have also emerged as important exporters in recent years. Although the
vast majority of rice production in the world’s two largest countries is consumed
domestically, both have proved able to export large quantities in recent years. India
was the world’s second largest exporter (behind only Thailand) in 1996 and 1998,
exporting more than 5 million tons in 1998. China exported more than 3.5 million
tons in 1998. The apparent willingness of these countries to supply world markets
lends added stability in times of crisis.
3
The renewed presence of several commercially oriented rice exporters is reflected
in the share of world rice production that is traded on world markets. Between 1961
and 1993, world trade fluctuated between 3.5 and 5% of world production (on aver-
age, it was 4.3%). Since 1994, however, the ratio has exceeded 5% every single
3
Although the US is also a major exporter, most US rice is too expensive to enter the Asian trade
on a purely commercial basis.
366 D. Dawe / Food Policy 27 (2002) 355–370
year, and the share traded has averaged 6.1%. This does not make the world rice
market as heavily traded as world wheat and maize markets, but it still represents
an important increase. Furthermore, although the world rice market is less heavily
traded than other grain markets, world rice prices are now more stable than world
wheat and maize prices, in contrast to earlier periods (Table 6). Thus, while world
market price stability has improved for all three major grains, it has improved most
for rice.
Whither the world rice market?
The recent plunge in world prices
What does the future hold for the world rice market, in terms of the level and
stability of prices? In terms of levels, prices have declined substantially in the past
3 years. From 1998 to 2001, the price of 100Bs (the highest quality rice of those
that are widely traded) fell 48% in real terms. The price of 25% brokens, a quality
roughly equivalent to what is eaten by average consumers in major importing coun-
tries such as Indonesia and Philippines, declined by a similar amount.
The magnitude of this decline is nearly equivalent to the precipitous decline that
occurred from the early to mid-1980s, and the reasons behind it are similar as well.
First, after importing 6 million ton in 1998, Indonesia greatly reduced its exposure
to the world market in succeeding years. The reasons were different this time (a
recovery in production from the El Nin
˜o drought, a depreciation of the exchange
rate, and a tariff on rice as opposed to a production surge caused by new technology
and infrastructure), but the consequences were the same: lower world prices. Second,
the Thai baht was devalued during the Asian financial crisis. Third, rice production
has surged recently in Bangladesh, Vietnam, Pakistan, and India. From 1995 to 1999
(just 4 years), production surged 30, 26, 30, and 15% in these four large countries,
respectively. Much of this recent surge was underpinned by public and private invest-
Table 6
Comparative stability of world grain markets
Commodity 1957–1964 1965–1981 1985–1998
Average absolute value of annual percentage price changes (%)
Maize 5 12 15
Wheat 4 16 15
Rice 7 24 11
RMSE (from regression of real price vs. time trend) normalized by average price
Maize 0.068 0.222 0.155
Wheat 0.066 0.284 0.147
Rice 0.065 0.341 0.111
Source of raw data: IMF (2002). Time periods do not correspond exactly to those in Table 1 because
shorter time-series of prices are available for wheat and maize than for rice.
367D. Dawe / Food Policy 27 (2002) 355–370
ments in water control. For example, the proportion of irrigated area in Bangladesh
was just 25% in the early 1990s, but it is now more than half due to the expansion
of private sector shallow tubewell irrigation. Vietnam has recently invested heavily
in constructing sluice gates to control salinization and new canal systems to increase
rice area in the Mekong Delta.
A simple structural model of world rice prices
The similarity of events surrounding the two large price declines of the past 20
years suggests the possibility of isolating the contributions of each factor through
the use of a multiple regression model (Table 7). The model was estimated in both
levels and first differences. The results were similar for both, but only the results
for the level specification are presented.
4
The price series used was for 25% brokens
because this is the most common quality imported by the two main Asian developing
country importers, Indonesia and Philippines. (Similar results were also obtained
with the price series for 100Bs, the highest grade of indica rice on the world market.)
The time period covered by the analysis was 1984–2000. Earlier data were not used
because the structure of the world market was different before 1984, as argued above.
Explanatory variables included 1 and 2 year lags of per capita production, a 1
year lag of the logarithm of the nominal Thai baht versus US$ exchange rate, and
Indonesian imports. (A measure of the real exchange rate of the baht versus the US$
was also used, with similar results. Time trends were included in both the level and
first difference specifications, but were not significant.) Since population growth is
the main source of demand growth in the Asian rice economy, production was nor-
malized by population, i.e. per capita rice production was used as the independent
Table 7
Coefficient estimates from structural model of real world rice prices
Variable Coefficient estimate Standard error t-Statistic p-Value
Constant 3331 411 8.11 0.00
PCP(-1) −6.73 2.31 −2.91 0.01
PCP(-2) −7.47 2.20 −3.40 0.01
lnBahtUS$(-1) −222 64 −3.45 0.01
IndonImports 8.56 5.33 1.61 0.13
Estimation uses annual data from 1984 to 2000 (FAO, 2002; IMF, 2002; USDA, 2000). Dependent vari-
able is real price (25% brokens). PCP is per capita rice production for an aggregate of 17 large Asian
countries (see note to Fig. 2). lnBahtUS$ is the natural logarithm of the Thai baht/US$ exchange rate.
IndonImports is net rice imports by Indonesia. Number of observations=17; R
2
⫽0.86; adjusted R
2
⫽
0.81; DW statistic=1.31.
4
Dickey–Fuller tests indicate that the null hypothesis of a unit root for real prices cannot be rejected,
but weighted symmetric (t) tests indicate that real prices are I(0). Since ttests dominate Dickey–Fuller
tests in terms of power (Pantula et al., 1994), the level specification is reported in the text.
368 D. Dawe / Food Policy 27 (2002) 355–370
variable. This normalization keeps the demand curve fixed and allows estimation
without resorting to a system of simultaneous equations. Net trade in rice for Asia
is a very small percentage of total production.
Thailand is the world’s largest exporter and Indonesia the world’s largest rice
importer, justifying the use of important economic variables from these two coun-
tries. Policy and structural differences between the two countries account for the
different manner in which the Indonesia and Thailand variables enter the model.
First, Indonesian imports were hypothesized to have a contemporaneous effect on
world prices, while the value of the Thai baht was hypothesized to affect the world
price with a 1-year lag. This is because import demand can increase world prices
in a very short term, while changes in production incentives for Thai farmers trans-
mitted through the exchange rate require time before farmers can adjust planting
decisions. Also, during the past 20 years, private sector rice exports from Thailand
have been largely unrestricted. Thus, export quantities are endogenous to world
prices and it is necessary to use the value of the Thai baht instead of exports in
order to have the independent variable be exogenous. On the other hand, net Indone-
sian imports are used because import and export decisions were under monopoly
control of the Indonesian government until very recently and were made in response
to domestic needs, not the level of the world price. Thus, Indonesian imports are
effectively exogenous to world prices.
All estimated coefficients are of the expected sign and have relatively low p-
values, although the coefficient on Indonesian imports is not statistically significant
at conventional levels. Furthermore, the model captures turning points in prices over
the interval quite well. The magnitude of the coefficient estimate for the variable
measuring net Indonesian imports implies that an increase in Indonesian imports of
1 million tons is associated with an increase in world prices of US$8.56 per ton,
other things being equal. The magnitude of the coefficient on the value of the Thai
baht implies that a depreciation of the baht by 10% in year tis associated with a
decline in world prices of US$22 per ton in year (t⫹1). The coefficients on the
two lags of per capita production imply that an increase in per capita production of
1 kg paddy per capita in year tis associated with a decline in prices of US$6.73
per ton in year (t⫹1) and a further US$7.47 per ton in year (t⫹2). (Current per
capita production is approximately 165 kg paddy per capita.)
The future of the world rice market and policy implications
If this model is at least approximately true, it would imply that prices will remain
in the neighborhood of their current low levels for the medium term, barring a large
fall in per capita rice production, a major appreciation of the Thai baht, or Indonesia
becoming an important exporter. The latter two alternatives seem unlikely, and there
are no concrete signs of per capita production falling from current near record levels.
The one possible countervailing factor is the long-term slowdown in yield growth
that has occurred throughout Asia. This phenomenon has been most pronounced in
countries like Indonesia and Philippines that were among the first to embrace the
modern varieties of the Green Revolution (Otsuka, 2000). In both of these countries,
369D. Dawe / Food Policy 27 (2002) 355–370
rice yields are no higher today than they were 10 years ago. If yield growth continues
to decelerate throughout Asia, and does so faster than population growth, then per
capita production will probably begin to decline and this may cause rice prices to rise.
Furthermore, it seems likely that world prices will generally remain stable in the
near future, just as they have during the past 15 years. The improvement in stability
has come about due to the greater importance of irrigation in rice production, the
improved pest and disease resistance of modern varieties, and the re-emergence and
strengthening of the commercial orientation of major rice exporting nations. None
of these trends are likely to be reversed. Although the growth of irrigation is slowing
in Asia, the share of irrigated land in total rice area is still increasing (albeit slowly).
Plant breeders continue to improve the insect and disease resistance of modern var-
ieties, and biotechnology holds out hope for even more improvements in this area.
Finally, as the world economy moves toward freer trade and increasing integration,
it is unlikely that Thailand and Vietnam will turn their back on the world rice market.
And with appropriate changes in policies, Myanmar and Cambodia may emerge to
become important stabilizing forces once again sometime in the next decade.
Reduced price variability does not necessarily mean that the effects of instability
are now entirely negligible. Price fluctuations still cause serious problems for many
poor Asians, and the resultant effects on producer and consumer welfare can have
political repercussions. And, even if the world rice market is relatively quiet in the
future, increased financial market liberalization may lead to greater exchange rate
fluctuations in the future (Stiglitz, 2000). Under free trade, changes in exchange rates
translate to changes in domestic rice prices just as much as changes in world rice
prices, and Asian governments will need to formulate cost-effective policies to deal
with these issues (Dawe, 2001). Nevertheless, it seems likely that the combination
of low and stable prices on the international market will continue for the medium
term. This suggests that there is less risk if rice importing countries decide to rely
on the world market more heavily than they have in the past.
Acknowledgements
I thank C. Peter Timmer and Scott Pearson for many helpful suggestions. The
standard disclaimer applies.
References
Barker, R., Herdt, R.W., Rose, B., 1985. The Rice Economy of Asia. Resources for the Future, Wash-
ington, DC.
Dawe, D., 2001. How far down the path to free trade? The importance of rice price stabilization in
developing Asia. Food Policy 26 (2), 163–175.
FAO (Food and Agriculture Organization of the United Nations), 2002. Electronic on-line database. Avail-
able from http://www.fao.org.
Falcon, W.C., Monke, E., 1979-1980. International trade in rice. Food Research Institute Studies 17 (3),
271–306.
370 D. Dawe / Food Policy 27 (2002) 355–370
Hayami, Y., Kikuchi, M., 1978. Investment inducements to public infrastructure: irrigation in the Philip-
pines. The Review of Economics and Statistics 6 (1), 70–77.
Huang, J., Rozelle, S., 1996. Technological change: rediscovering the engine of productivity growth in
China’s rural economy. Journal of Development Economics 49, 337–369.
Huke, R.E., Huke, E.H., 1997. Rice area by type of culture: South, Southeast and East Asia. International
Rice Research Institute, Manila, Philippines.
IMF (International Monetary Fund) 2002. International financial statistics CD-ROM Database. Wash-
ington, DC.
Jayne, T., 1993. Sources and effects of instability in the world rice market, MSU International Develop-
ment Paper No. 13. Department of Agricultural Economics and Department of Economics, Michigan
State University, p.104.
Khush, G.S., 1995. Modern varieties: their real contribution to food supply and equity. GeoJournal 35
(3), 275–284.
Monke, E., Pearson, S., 1991. The international rice market. In: Pearson, S., Falcon, W., Heytens, P.,
Monke, E., Naylor, R. (Eds.), Rice Policy in Indonesia. Cornell University Press, Ithaca, NY.
NOAA (National Oceanic and Atmospheric Administration) web site, 2000.
http://www.cdc.noaa.gov/ENSO/enso.mei—index.html.
Otsuka, K., 2000. Role of agricultural research in poverty reduction: lessons from the Asian experience.
Food Policy 25, 447–462.
Palacpac, A., 1977. World Rice Statistics. International Rice Research Institute Department of Agricultural
Economics, p.140.
Pantula, S.G., Gonzalez-Farias, G., Fuller, W.A., 1994. A comparison of unit-root test criteria. Journal
of Business and Economic Statistics 10, 449–459.
Pindyck, R.S., Rotemberg, J.J., 1990. The excess co-movement of commodity prices. Economic Journal
100 (December), 1173–1189.
Rosegrant, M.W., Sombilla, M.A., Perez, N., 1995. Global Food Projections to 2020: implications for
investment. Food, Agriculture and the Environment Discussion Paper No. 5. International Food Policy
Research Institute, Washington, DC.
Rosegrant, M.W., Svendsen, M., 1993. Asian food production in the 1990s: irrigation investment and
management policy. Food Policy 18 (2), 13–32.
Sarris, A.H., 2000. World cereal price instability and a market based instrument for LDC food import
risk management. Food Policy 25 (2), 189–209.
Siamwalla, A., 1975. A history of rice policies in Thailand. Food Research Institute Studies 14 (3),
233–249.
Siamwalla, A., Haykin, S., 1983. The world rice market: structure, conduct, and performance. Research
Report 39, International Food Policy Research Institute, Washington, DC, p. 79.
Slayton, T., 1999. The outlook for the rice trade in the new millennium. In: Paper presented at the Fourth
Asia International Rice Conference, Cebu, Philippines, August.
Stiglitz, J.E., 2000. Capital market liberalization, economic growth, and instability. World Development
28 (6), 1075–1086.
Timmer, C.P., Falcon, W.P., Pearson, S.R., 1983. Food Policy Analysis. Johns Hopkins University Press,
Baltimore, MD Published for the World Bank.
UNDP (United Nations Development Programme), 1997. Human Development Report. Oxford University
Press, New York.
USDA (United States Department of Agriculture), 2000. Daily market quotations from GAIN (Global
Agricultural Information Network) report.
Zhang, B., Carter, C.A., 1997. Reforms, the weather and productivity growth in China’s grain sector.
American Journal of Agricultural Economics 79 (4), 1266–1277.