This chapter examines what would happen to domestic and international markets over the medium term (2018-2030) if current public stockholding policies were to remain in place (baseline scenario) compared to a situation whereby the eight selected Asian countries in this report collectively increased (high-level scenario) or decreased (low-level scenario) their public stock levels. The analysis shows that the strongest impacts would occur during the three-year transition period when countries adjust their public stocks to the new levels. It also shows there would be structural impacts over the medium term, although at a lower intensity, on procurement, domestic and international prices, availability, private stock levels, and public expenditure. In the event of a global production shock, the model projects that the immediate impact on prices and availability would be less severe under the high-level scenario, but that recovery towards the no-shock situation would be faster and public expenditure lower when countries hold smaller public stocks.
The Economic Effects of Public Stockholding Policies for Rice in Asia
Chapter 3. Medium-term market effects of alternative public stockholding policies
Abstract
3.1. Description of scenarios
This chapter examines what would happen in domestic and international markets if all eight countries set their public stock levels to either a high or low level at the same time. The benefit of this scenario is that it establishes bounds to the global market impacts that derive from changes in public stock policies across the region. The fact that all these countries react in the same way could be the result of a period of relatively high (low) prices or price volatility. This was, for example, the case during the food price crisis in 2007-08 when several countries began to expand their public stocks in response to higher and more volatile prices.
To model these two scenarios, the norm for public stocks is adjusted to a high level (high-level scenario) or a low level (low-level scenario). Both scenarios are stylised experiments and are not designed to represent specific policies under consideration by specific countries. The level at which countries set their norms is expressed in days of national consumption. The advantage of this measure is that it allows for straightforward comparisons across countries. Furthermore, the norms of public stocks are often expressed in days of national consumption when they have a food security objective.
The level of public stock norms under the high-level scenario is set at three months of national consumption, while the level of public stock norms under the low-level scenario is set at two weeks. These values were chosen by examining public stocks over the past decades in the selected countries. The transition towards the high or low public stock norm levels is assumed to occur linearly during the three-year period 2018-2020.
Figure 3.1 shows the average public1 stock levels and norms under the baseline as well as under the low- and high-level scenarios after the transition period. The norms under the baseline differ by country and reflect the most recent information available, whereas the norms under the high- and low-level scenarios are set uniformly across countries, namely 91 days under the high-level scenario and 14 days under the low-level scenario. In the case of the People’s Republic of China (hereafter “China”), there was no historic information on public stock levels and norms. For the baseline, the norm in China was set at 52 days, which was selected because it is the midpoint between the low- and high-level scenarios.2 In Thailand, the baseline assumes that no public stocks are held and therefore the norm is set at zero.
For all countries, the high-level scenario translates into an increase in the norm (and public stock levels) compared to the baseline. In the case of the low-level scenario, the norm and public stock levels are lower than under the baseline in all countries except two, Bangladesh and Thailand. In Bangladesh, the low-level scenario and baseline are the same. In Thailand, both the low- and high-level scenarios lead to an increase in the norm and public stock levels compared to the baseline.
The scenarios examine what happens if countries decide to change their stockholding policies by changing the norm. They assume that the functioning of the stockholding policies remains unchanged, i.e. the acquisition and distribution rules (as described in Chapter 1 and used in the Aglink-Cosimo model) stay in place. For example, a country that traditionally has not distributed rice from public stocks at a subsidised price will not start doing this under the low- and high-level scenarios. The only way a country changes its policy is by adjusting the norm. This is a common stockholding policy practice as norms are set to ensure that stock levels are sufficient to meet certain acquisition or distribution targets.
Figure 3.1. Average public stock levels and norms under the baseline, low-, and high-level scenarios, 2021-2030
Note: Expressed in days of national consumption. Public stock data in Japan and Korea refer to public stocks composed of domestic rice.
Source: OECD simulations.
The exception to this assumption is Thailand. Under the baseline projections, there are practically no public stocks. In order to implement the low- and high-level scenarios, it was assumed that Thailand would resume the stockholding policies it had in place in the past. This means that rice will be procured exclusively on the domestic market at a procurement price and that all rice from public stocks will be exported. The procurement price was set below the market price.
Figure 3.1 illustrates how norms affect public stock levels. As described in Annex 2.A, public stock levels are not only a function of the norm but also depend on other factors. This explains why the norm and public stocks levels are not always the same. Furthermore, in certain countries, this disparity between the norm and public stock levels has historically been higher than in other countries, which is also reflected in the scenarios. For example, in Bangladesh the public stock levels and the norm have historically been at the same level, which is reflected in the baseline and the scenarios. In India, however, public stock levels have in recent years been higher than the norm, and this difference is also present in the baseline and the scenarios.
The market impacts of the two scenarios are analysed at the global and domestic levels. When examining the impacts on global markets, it is clear that changing stockholding policies in certain countries will have a greater impact than those in other countries. The effects will also depend on whether the country is a major rice producer, exporter, importer, or neither, and whether it uses imports to build public stocks. The impacts measured in each domestic market will be a combination of the country's own domestic policies plus the impacts of all other countries doing something similar. This implies that in certain smaller countries, the impact of domestic stockholding policy changes can be small compared to the impact of other countries behaving in a similar fashion.
With the exception of China, the scenarios are not symmetric. In some countries, notably Bangladesh and Indonesia, the low-level scenario is relatively close to the current baseline. For those countries, the move towards the high-level scenario has a much bigger impact than for countries where the baseline norm is situated between the low- and high-level scenario norms. In addition, the difference between the baseline and the high-level scenario is higher than the difference between the baseline and low-level scenario in most countries, which indicates that a larger adjustment is needed to reach the high-level scenario than the low-level scenario.
3.2. Impact on procurement
Under the low- and high-level scenarios, procurement levels are adjusted to meet the new norms. The change in procurement levels is projected to be most pronounced during the transition period, 2018-2020 (Figure 3.2). Aggregate procurement increases sharply in the high-level scenario during these three years, while it decreases under the low-level scenario. Once the transition period is over and the new norms have been met, procurement levels stabilise. However, there is still a difference in the level of procurement between the baseline, and the low- and high-level scenarios. This is because higher (lower) norms imply higher (lower) procurement levels, which in turn leads to higher (lower) amounts of distribution.
Figure 3.2. Aggregate procurement by all countries except China
Note: China is not included in this figure because there is no information on procurement. Shaded area indicates the projection period 2018-2030.
Source: OECD simulations.
The difference in aggregate procurement between the baseline and high-level scenario is higher than the difference between the baseline and the low-level scenario. This is because the baseline norm in most countries is relatively closer to the norm in the low-level scenario than to the norm in the high-level scenario. This is also the case in India, which highly influences the aggregate procurement levels in this figure. In all three scenarios, India accounts for over 80% of aggregate procurement. However, this aggregate volume does not consider procurement in China as no information on procurement or distribution is available.
Indonesia and the Philippines regularly import rice to build public stock. For these two countries, historic information starting from 2000 is available on the volume of imports they added to their public stocks. Figure 3.3 illustrates that, historically, the volume of imports by Indonesia and the Philippines has fluctuated considerably. Over the projection period, Indonesia’s procurement from imports is projected to decline slightly, following the country's overall import trend. In the Philippines, the baseline projections for public stock procurement from imports are twice as high as domestic procurement, which reflects the historical situation.
Under the high-level scenario, procurement from imports is projected to increase in both countries as the high norm cannot be reached via domestic procurement only. Under the low-level scenario, Indonesia will continue to increase its imports relative to the baseline during the transition period. As the norms of all countries are reduced under the low-level scenario, the global market is supplied with rice, which leads to a decrease in the world price. Since imports will be cheaper under the low-level scenario than under the baseline during these three years, Indonesia will increase its imports under the low-level scenario compared to the baseline. However, since the Philippines is more integrated with international markets (Furuhashi and Gay, 2017) and the price differential between the domestic and international markets is not sufficiently large, it will not increase its procurement from imports.
Figure 3.3. Procurement from imports in Indonesia and the Philippines, 2000-2030
Note: Shaded area indicates the projection period 2018-2030.
Source: Historic data for Indonesia from OECD (2012) and for the Philippines from NFA (2016a, 2016b, 2017).
3.3. Impact on private stocks
An advantage of the new model is that it separates private from public stocks and incorporates the link between the two types of stock. In particular, as countries build larger public stocks, the private sector is crowded out from stockholding activities. Accordingly, the aggregate level of private stocks is higher under the low-level scenario compared to the baseline since public stocks held by all countries drops (Figure 3.4).
The impact on private stock levels is expected to be most pronounced during the transition period, when public stocks are released on the market under the low-level scenario in order to achieve their new public stock norms. During these three years, the increase in supply under the low-level scenario causes rice prices to drop in domestic markets, triggering the private sector to buy rice at low prices. In the years following the transition period, rice prices rebound and hence the private sector sells its stocks. The opposite occurs under the high-level scenario.
The level of increase or decrease in private stocks varies by country. Figure 3.5 shows the average relative change (in percentage) for each country in the level of private stocks between the baseline and the scenarios during the projection period. In all countries, except for Bangladesh, Korea and Thailand, private stocks are projected to increase in the low-level scenario relative to the baseline. In Bangladesh, the public stock levels under the low-level scenario and baseline are the same. Korea has no private stocks under the baseline and is assumed not to have private stocks under the low- and high-level scenarios. In Thailand, private stocks decrease under the low-level scenario because under the baseline Thailand does not have any public stocks (only private stocks). Hence, the move for Thailand from the baseline towards the low-level scenario implies an increase in public stocks, and thus a decrease in private stocks. Private stocks will be relatively lower for all countries under the high-level scenario compared to the baseline.
In absolute volume terms, the change in private stock levels is projected to be most pronounced in China, which is assumed to place a large share of its stocks in private storage, whereas it will be relatively small in India due to the still very large public stock levels and the limit on private stockholding.
Figure 3.4. Aggregate level of private stocks of rice
Note: Aggregate includes all eight countries in the study. Shaded area indicates the projection period 2018-2030.
Source: OECD simulations.
Figure 3.5. Average relative change in the level of private stocks between the baseline and the scenarios, 2018-2030
Source: OECD simulations.
3.4. Impact on prices
Figures 3.6 and 3.7 illustrate how producer prices in the eight countries in this study are projected to change in the low- and high-level scenarios compared to the baseline scenario. The effects, expressed in relative percentage changes, vary across the countries but are expected to be strongest during the transition period and to dissipate in the years afterwards. In the low-level scenario, rice will be released from public stocks on the domestic or international market during these three years. The increased overall supply leads to lower producer prices. The inverse occurs under the high-level scenario, where producer prices are projected to be relatively higher than under the baseline during the transition period.
The change between the baseline and the high-level scenario during the transition period is much more pronounced than between the baseline and the low-level scenario. In the high-level scenario, producer prices are projected to increase between 5% and 18% compared to the baseline, while in the low-level scenario, producer prices will be between 2% and 6% lower compared to the baseline. The stronger reaction in the high-level scenario is partly a result of the asymmetric relation of the two scenarios vis-à-vis the baseline: the baseline norms in most countries are closer to the low-level scenario than to the high-level scenario. This explains why the increase in producer prices is highest in Bangladesh and Indonesia, the two countries where the baseline norm differs most from the high-level scenario norm. Another factor that might lead to a higher and more volatile reaction of prices under the high-level scenario is the fact that the private sector is less involved in stockholding activities. In China, the two scenarios are modelled symmetrically, with the baseline norm at the mid-point between the low- and high-level scenarios. Still, the producer price reaction is stronger under the high-level scenario than in the low-level scenario. This suggests the role of the private sector in stabilising prices in the low-level scenario.
The impact on consumer prices mirrors the effect on producer prices as both prices are projected to follow a similar trend over the projection period (see also Figure 2.3). During the transition period consumer prices are expected to decrease under the low-level scenario compared to the baseline because of the offloading of public stocks, whereas they will increase under the high-level scenario.
The eight countries in this study together accounted for approximately 45% of world rice exports during the projection period. Accordingly, if these countries collectively increase or decrease their public stock levels, this will strongly influence global markets. Figure 3.8 shows that the effects are projected to be strongest during the transition period and will be more pronounced under the high-level scenario than under the low-level scenario. On average, world prices are projected to increase by 10% during these three years in the high-level scenario compared to the baseline, which is a result of tighter global supply. However, they drop by 4% on average in the low-level scenario because of the collective offloading of rice stocks on domestic and international markets.
Figures 3.6, 3.7 and 3.8 show that after the transition period, domestic and international prices are expected to move towards the baseline projections as farmers adjust their output in response to the price changes. That is, the relatively lower prices in the low-level scenario decrease the expected return per hectare which causes farmers to reduce their output. This in turn is projected to lead to an increase in prices in the years after the transition period. The opposite occurs under the high-level scenario, where the higher price for rice motivates farmers to produce more for the global markets during the transition period, which consecutively drives prices back down to the baseline equilibrium.
Figure 3.6. Producer prices: Percentage change in low-level scenario compared to the baseline scenario, 2017-2030
Source: OECD simulations.
Figure 3.7. Producer prices: Percentage change in high-level scenario compared to the baseline scenario, 2017-2030
Source: OECD simulations.
Prices will not return completely to the baseline levels. Once supply has adjusted in the years following the transition period, prices under the low-level scenario are projected to consistently remain below the baseline levels while they will stay above the baseline levels under the high-level scenario. Prices are not expected to converge entirely to the baseline because procurement levels have changed under the scenarios compared to the baseline. The extent to which prices will stay above or below the baseline levels also depends on the effectiveness of the public stockholding programme. Under the current model specifications, a loss rate of 2% has been assumed for all countries, which means that average distribution levels are 2% lower than average procurement levels. However, if losses are higher, then less of the procured rice will reach markets, which will in turn induce stronger price effects, making the relative differences between the baseline, low-, and high-level scenarios more pronounced.
Figure 3.8. World rice prices during baseline, low-, and high-level scenarios, 2017-2030
Source: OECD simulations.
3.5. Impact on availability
Public stockholding policies often have a food security objective: public stocks are built with the purpose to release rice on the domestic market at a subsidised price. This is the case in Bangladesh, the Philippines, India, and Indonesia. In the latter two countries, public stocks also play a buffer role since rice can be released from stocks at prevailing market prices. When countries set a higher norm for their public stocks, they not only acquire more rice but also distribute more rice.
Figure 3.9 illustrates the impact on distribution in the low- and high-level scenario vis-à-vis the baseline by comparing the total amount available for distribution, expressed in per capita terms. In the four above-mentioned countries, the difference between the high-level scenario and the baseline over the projection period on average varies between 0.8 kg and 3.1 kg per capita per year. This means that between 0.8 kg to 3.1 kg more rice is projected to be distributed per person each year compared to the baseline. This scenario is a best-case scenario as it assumes that losses are minimal (2%).
Under the low-level scenario, distribution per capita is projected to be on average less than under the baseline. The difference between the baseline and the low-level scenario is small for Bangladesh and Indonesia since for these two countries, the norm and public stock levels under the baseline and low-level scenario are the same (Bangladesh) or very close (Indonesia). In the case of India, the difference between the low-level scenario and baseline is more pronounced given the overall size of the country's public stockholding program.
Figure 3.9. Per capita distribution of rice: Average difference between the baseline, low-, and high-level scenarios, 2018-2030
Source: OECD simulations.
Figure 3.10. Per capita rice consumption (availability): Average percentage difference between the baseline, low-, and high-level scenarios, 2018-2020 and 2021-2030
Note: Food availability is calculated based on FAO's Food Balance Sheets.
Source: OECD simulations.
Distribution of rice from public stocks is only one part of a country's overall rice availability. As indicated in the previous section, during the transition period towards the low or high-level scenario, prices are projected to decrease or increase considerably. This also affects availability. Figure 3.10 shows the relative change in per capita availability in each of the countries between the baseline and the two scenarios. A distinction is made between the transition period (2018-2020) and the post-transition period (2021-2030). During the transition period, per capita availability of rice is projected to increase under the low-level scenario compared to the baseline while it will decrease under the high-level scenario. This is a result of the offloading of rice during the transition period under the low-level scenario while under the high-level scenario more rice is procured in order to reach the high norm levels. Once the transition period is over and the new norms and public stock levels have been reached under both scenarios, per capita availability is expected to move towards the values under the baseline. Still, given that prices in the long term are projected to be stabilised at lower (higher) values under the low-level (high-level) scenario than under the baseline, availability for several countries is expected to remain above the baseline under the low-level scenario, while it will be lower than the baseline under the high-level scenario.
3.6. Impact on public expenditures
Public stock programmes require substantial amounts of government funds. The actual costs of these programmes vary by country and increase as the size of the programme expands. In this study, the costs for public stocks are estimated in order to obtain comparable amounts between the countries. They are calculated as the sum of estimated procurement costs, carrying costs, and costs as a result of losses minus potential revenues from selling rice from public stocks.
Procurement costs for each country are decomposed into the different channels of procurement. That is, procurement costs for a country that acquires rice from the domestic market at a procurement price and from imports will be decomposed into the amount procured in the domestic market times the procurement price plus the amount of rice procured in the international market times the import price. Likewise, the potential distribution revenues from selling rice from public stocks are decomposed for each country reflecting their respective distribution channels. Carrying costs are estimated at USD 84 per tonne3 of public stock held and costs from losses are estimated at 2% of the public stocks in the previous period times the producer price.
The resulting cost estimates are an underestimate of the actual costs as they do not account for other additional expenditures, which are difficult to quantify (e.g. under the high-level scenario, larger procurement and distribution will require extra storage space and improved infrastructure), and use a lower bound of 2% for all countries to estimate the costs incurred from losses. Costs of public stocks cannot be calculated for China because data on procurement and distribution volumes are lacking.
The aggregate cost of the public stock programme for the seven countries is estimated at USD 234 billion under the baseline for the entire projection period. Under the low-level scenario, these costs are projected to lower to USD 203 billion, while they increase to USD 322 billion under the high-level scenario.
Figure 3.11 illustrates how much more (or less) countries are projected to spend on their public stockholding programme over the period 2018-2030 if they implement the high-level (or low-level) scenario. The costs are accumulated over the projection period 2018-2030 and are expressed in USD billion. In India, almost USD 39 billion more are projected to be needed to obtain and maintain the high-level scenario over this period compared to the baseline, whereas USD 26 billion can be saved by reducing the programme to the low-level scenario. Bangladesh and Indonesia are projected to spend USD 15 and USD 22 billion more, respectively, over the entire projection period if they were to implement the high-level scenario. The savings for these countries by switching to the low-level scenario are nil (Bangladesh) or minimal (Indonesia) because their baselines are the same as (Bangladesh) or very close to (Indonesia) the low-level scenario. If Thailand were to re-instate the public stockholding programme under the conditions assumed in this report (i.e. procurement price at 95% of producer price and all rice from public stocks gets exported), then the projected costs over 2018-2030 are estimated to be almost ten times more under the high-level scenario compared to the low-level scenario.
Figure 3.11. Absolute difference in cost of public stockholding programs, accumulative over projection period 2018-2030
Source: OECD simulations.
3.7. Impact of a production shock
One of the main reasons why countries keep large amounts of rice in public stocks is to be prepared for emergency situations whereby domestic or global rice availability suddenly drops. In this event, they can release rice from their public stocks at either prevailing market prices or at subsidised prices. Maintaining larger public stocks of rice would provide a larger buffer against these shocks, but how much better can countries mitigate the impacts of these shocks when they keep large public stocks compared to smaller ones, and what would be the fiscal implications?
This section examines the impacts of a sudden drop in availability under the low-level scenario compared to the high-level scenario. The shock is modelled to occur in 2024. This year was chosen because by then the transitional impacts of moving towards the low- or high-level scenario will have dissipated. The shock is designed to correspond to the largest deviation from the global trend during the last 20 years: it simulates a 5% decrease in domestic production in each of the eight countries, which is equivalent to a global drop in rice production of 3%. After the shock, countries are assumed to rebuild public stocks to reach their low- or high-level scenario norm levels.
In the previous sections, one of the underlying assumptions was that the norm is used as a lower bound for the level of public stocks. Namely, it was assumed that public stocks do not go below 90% of the stock norm. However, the Aglink-Cosimo model also allows for emergency stock releases in case of a sudden drop in availability.4 Under this emergency situation, up to 90% of available public stocks can be released. This extra specification in the model reflects the overall functioning of public stockholding programmes, whereby the level of stocks is maintained at a certain threshold under a normal price environment, but allows the release of rice under extreme circumstances.
The eight countries in this study are responsible for 75% of global rice production and 45% of global rice trade, and hence a shock in these countries will affect global rice availability and world rice prices. Figure 3.12 and Figure 3.13 illustrate the impact of a production shock on global rice availability and world rice prices, respectively. They show availability and world prices under the low- and high-level scenarios during normal times (no shock) compared to a situation with a 3% global production shock in 2024 (with shock). Figure 3.12 and Figure 3.13 are both composed of two panels. The panels on the left show the evolution of availability and prices over a relatively long period in order to put the impact of the global production shock into perspective. The panels on the right focus on the period 2024-2028 when the effects are the most prominent.
Panel (a) of Figure 3.12 illustrates that global availability of rice is projected to be less affected by the global production shock in 2024 than by the collective move towards the high-level scenario during 2018-2020. Panel (b) shows that during the year of the shock, 2024, global rice availability is projected to decrease more under the low-level scenario than under the high-level scenario and in 2025, availability under the low-level scenario decreases even more. In 2026, markets start recovering from the shock but availability is still lower than under a situation without a shock. However, the relative decrease in availability compared to the “no shock” situation is in 2026 and 2027 less under the low-level scenario than under the high-level scenario.
The impact of a global production shock on world prices is expected to be initially stronger when countries keep low public stocks, but the effects are projected to persevere for a longer period when countries keep high public stocks. Figure 3.13 illustrates that world prices in 2024 and 2025 are projected to be respectively 5% and 6 % higher under the low-level scenario if there is a shock compared to a no shock situation. Under the high-level scenario, the world price for rice in 2024 and 2025 will be almost 5% higher in the event of a shock. Panel (a) shows that these price effects are nevertheless lower than the price increase experienced during the transition period towards the high-level scenario (2018-2020). In the years following the shock, the price effects are projected to dissipate more quickly under the low-level scenario than under the high-level scenario. By 2027, the world price for rice under the low-level scenario has converged to prices under the no shock situation while it is still above the no shock situation under the high-level scenario.
Under the high-level scenario, prices take longer to come down in the event of a shock since countries are assumed to start rebuilding public stocks in the years following the shock. In 2024, public stocks are drawn down to deal with the lower availability, but once the shock is over, the stocks need to be rebuilt towards the low or high norm levels. In case of the high-level scenario, the norm is set higher and hence more rice needs to be procured in order to rebuild the stock, which in turn has a larger impact on prices and availability.
The individual countries in this study are expected to experience similar impacts on availability and prices as the global market. That is, a 3% global production shock is projected to lead to a relatively larger decrease in per capita availability and larger increase in consumer prices in 2024 and 2025 under the low-level scenario compared to the high-level scenario. In the years following the shock, prices and availability recover faster (i.e. converge faster to the no shock situation) under the low-level scenario than under the high-level scenario.
Figure 3.12. Impact of a 3% global production shock in 2024 on the global availability of rice under the low- and high-level scenarios
Source: OECD simulations.
Figure 3.13. Impact of a 3% global production shock in 2024 on the world price of rice under the low- and high-level scenarios
Source: OECD simulations.
During the year of the shock, certain countries are projected to experience a stronger impact than others. Figure 3.14 shows for each country how much lower the impact of a shock is projected to be in 2025 if they had kept large public stocks versus small public stocks. The year 2025 was selected instead of 2024 because in 2025 the impact of the shock experienced under both scenarios is the strongest. Figure 3.14 also provides an estimate of the additional public expenses to experience the relatively lower impacts in 2025 under the high-level scenario compared to the low-level scenario. The values in this figure hence allow for a cost-benefit analysis.
Figure 3.14. Changes in per capita food consumption and consumer prices in 2025 following a production shock and additional costs to implement the high-level instead of low-level scenario
Note: The production shock refers to a global 3% production shock in 2024. Costs are estimated as described in this report. No costs could be estimated for China as there is no information on procurement and distribution volumes.
Source: OECD simulations.
Among the eight countries, Bangladesh is projected to be one of the most negatively affected by a shock under the low-level scenario. The simulations show that if it were to keep a low level of public stock, then in the event of a shock, it is projected that per capita food availability would decrease by 1% and domestic prices are projected to increase by more than 4% in 2025 compared to a situation without the shock. If Bangladesh had decided to increase its public stocks to the high-level scenario, then a shock would lead to relatively smaller changes in 2025 in per capita food availability and domestic prices. However, for Bangladesh to experience these relatively smaller impacts by switching to the high-level scenario, it would have to spend USD 8.7 billion more, which is the estimated difference in the accumulated cost of public stockholding between the high- and low-level scenarios over the period 2018-2024.
3.8. Concluding observations
Main findings
Simulations using the Aglink-Cosimo model examine what would happen over the period 2018-2030 if eight Asian countries were to collectively change their public stock levels of rice from their baseline level (baseline) to either a low level (low-level scenario) or high level (high-level scenario). In addition, a global drop of 3% in rice production is simulated to occur in 2024 and the relative effects of this production shock are compared under the low- and high-level scenarios.
The relative domestic impacts are projected to vary by country, but the analysis shows that there are several common trends which indicate how changing public stockholding policies can influence markets in the short and medium term.
First, the effects of changing public stockholding policies are expected to be more pronounced over the short term than over the medium term. That is, procurement levels, prices and availability will be most affected during the three-year transition period when countries adjust their public stocks to the new levels, than during the remainder of the projection period.
Under the high-level scenario, additional rice must be procured from domestic or international markets during the transition period, which reduces availability. This in turn leads to price increases in both the domestic and international markets. The biggest price effects are projected in Bangladesh and Indonesia, where average increases of around 14% in domestic rice prices are expected during the transition period compared to the baseline. Prices in the other six countries average between 8% and 11% above the baseline levels during those three years. But other countries are also affected as world rice prices are projected to increase by 10% compared to the baseline.
Lowering the level of public stocks, as under the low-level scenario, has the opposite effect. Public stocks of rice will be offloaded in the domestic or international markets, which will increase supply. During the transition period, domestic and international rice prices are projected to be on average 4% lower than under the baseline.
Second, changing public stockholding policies is expected to have structural impacts over the medium term. Whereas the effects are more pronounced during the transition period, the impacts on overall procurement, prices, availability, private stock levels, and public expenditure are expected to persist, albeit at a lower intensity, over time.
In order to reach the higher public stock norm under the high-level scenario, procurement must expand significantly during the transition period. However, to maintain the higher public stock level over the medium term, it will be necessary to continue to procure more rice than under the baseline. This has repercussions on the procurement channels, as domestic procurement from farmers or imports will stay at higher levels. The reverse holds under the low-level scenario, where lower procurement will be sustained over the entire projection period.
As relatively more rice is consistently acquired from the market in order to maintain the higher stocks under the high-level scenario, rice prices and availability adjust accordingly. The analysis shows that under the high-level scenario prices are projected to remain above the baseline levels in the medium term while availability will be lower than under the baseline. In contrast, under the low-level scenario, rice prices are expected to be lower and availability higher than under the baseline. The extent to which the price and availability levels differ from the baseline levels also depends in part on the effectiveness of the public stockholding programme, whereby a higher loss rate will further widen the gap between the scenario and baseline levels.
The private sector becomes crowded out from stockholding activities under the high-level scenario due to the increased involvement of the public sector. The level of private stocks is projected to drop under the high-level scenario and will stay below the baseline over the medium term. Depending on the country, the level of private stocks is projected to be 3% to 30% lower than under the baseline. In the low-level scenario on the other hand, the private sector will become more involved in stockholding activities, with varying levels of expansion among the countries in the study.
Expanding public stocks requires additional funds to acquire, store, manage and distribute rice. This study estimates the cost of public stocks for all the countries in the study except China. Over the next ten years, it is estimated that USD 88 billion extra would be needed to implement the high-level scenario in those seven countries. India would account for the majority of that amount, as it would need almost USD 39 billion more to obtain and maintain the high-level scenario over the period 2018-2030. A collective move towards the low-level scenario is projected to lead to an accumulated saving of USD 31 billion in these seven countries over the projection period.
Third, the immediate market impacts of a global production shock are projected to be less severe if countries hold larger public stocks than when they hold smaller public stocks, but recovery towards the no-shock situation is expected to occur faster when countries hold smaller public stocks. Furthermore, the higher public expenditure bill associated with the high public stocks questions whether these additional funds are not better invested in other policies that safeguard consumers from sudden supply shortages.
Policy implications
The main policy implications of this study are that countries considering changing their public stockholding programmes should take into account that the resulting impacts will not be limited to their domestic markets nor to the short term. Instead, the analysis has shown that upscaling these programmes can be costly and that changes to these programmes can structurally affect domestic markets and can have international spill-overs over the medium term.
A principal motivation that governments claim for keeping large public stocks is that these can act as a safeguard against sudden supply shocks. However, the present analysis has demonstrated that although keeping higher levels of public stocks might initially lessen the impact on price and availability from a global shock, keeping low levels of public stocks enables faster recovery towards the no-shock situation. In addition, keeping low levels of public stocks considerably reduces the burden on public finances, which frees up funds that can be used for other mitigation strategies to deal with (emergency) food shortages.
References
Food Corporation of India (2017), “Accounting Year-Wise Opening Stock Adjusted Weighted Economic Cost and Acquisition Cost”, available at: http://fci.gov.in/finances.php?view=23.
Furuhashi, G. and H. Gay (2017), “Market implications of the integration scenario of Southeast Asian rice markets”, OECD Food, Agriculture and Fisheries Papers, No. 108, OECD Publishing, Paris. http://dx.doi.org/10.1787/c81e00c6-en.
NFA (2017) NFA Palay Procurement 2011-2016, available at: http://nfa.gov.ph/images/files/statistics/palayprocurement.pdf.
NFA (2016a) NFA Palay Procurement 2006-2014, available at: http://nfa.gov.ph/files/statistics/palayprocurement06-14.pdf.
NFA (2016b) Summary of NFA Rice Import Arrivals 2000-2012, available at: http://nfa.gov.ph/files/statistics/summary.pdf.
OECD (2012), OECD Review of Agricultural Policies: Indonesia 2012, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264179011-en.
Notes
← 1. Public stocks in Japan and Korea refer to public stocks composed of domestic rice.
← 2. An additional set of scenarios was run with higher public stock levels (set at 50% of total stocks) and higher norms (3, 4.5, and 6 months under baseline, low-, and high-level scenarios, respectively) for China. Under the new specifications, the impacts were slightly higher in China but less pronounced in the other countries and in global markets. Given these results, the original analysis is maintained for this study in order to preserve the harmonised set-up across all countries. A future study could focus on China and examine how varying norms and public stock levels in China could affect domestic and international markets. A crucial part of such a study would be to incorporate information on the amount of rice procured at support prices and through imports, as well as information on distribution levels.
← 3. The annual cost of storage per tonne of rice is based on the carrying costs of wheat and rice in India, which were USD 84 per tonne (INR 566 per quintal) in 2016-17 (Food Corporation of India, 2017). Since no information on carrying costs could be found for other countries, these costs were applied to all countries.
← 4. See Annex 2.A. for the stock equations which describe these mechanisms.