3. STRUCTURE OF THE CROSS-BORDER TRADE BETWEENETHIOPIA AND KENYA

Market structure can be explained in different ways depending on the purpose of the study. For instance, Cook et al, 1990 explained market structure for cross-border trade in some West African countries using trader's profile and goods traded. Similarly, Little (1997) described the structure of cross-border livestock trade between Kenya and Somalia using different actors involved and their roles in the marketing chain and markets. A Famine Early Warning System (FEWS) Somalia report described the structure using market channels and sheds (Steffen et al, 1998). All these studies used similar issues in explaining a given market structure albeit various degree of emphasis guided by variability in objectives and foci. In this study, elements of the cross-border trade are summarised to find out relationship with food security in the area and beyond. Thus, goods traded, traders' profile, markets, sheds, channels and currencies are identified and presented as follows.

3.1 Goods Traded

Observations were undertaken during November and December 1998 and January 1999 at various markets to identify the goods traded in the border areas. In addition, interviews were made to list important items traded under normal circumstances in these places. The evidence suggests that by encompassing only few items, the border trade be characterised by commodity concentration. Both primary and manufactured products in both directions are consumer goods and livestock. Livestock is also regarded as a consumer or a capital good. Consumer non-durables mainly food items and live animals are important commodities in the Ethiopia-Kenya border trade. The observation also shows that the commodity concentration is more in Ethiopian exports than its imports. Specialisation in the direction of the flow is presented in Table 2 below.

Table 2. Goods Traded in the Ethiopia-Kenya Borderlands

SOURCE: Survey Result, 1999

*stimulant green leaves

Table 2 shows that primary products predominate Ethiopian unofficial exports to Kenya crossing the border and livestock and livestock products are the most important items. This is similar to the structure of the country's official export trade, i.e. agricultural exports. Although livestock and livestock related exports are next to coffee in the country, livestock production in the southern and southeastern rangelands is a primary and important activity. On the other hand, goods unofficially imported from Kenya are manufactured or processed and most of them are manufactured in Kenya. These items are similar to those items officially imported by Ethiopia from Kenya. Cattle and camel milk from Ethiopia are also sold in border towns of Kenya such as Mandera, Ramu and Moyale. In Liban and Borana Zones, the northern limit for milk collection and sale extends as far as 50 km north of the border i.e. Kole and Boko Luboma (see Map 3). Milk fetched high price at the border markets as compared with prices further down to Marsabit and Isiollo in Kenya and in the southern and southeastern borderlands.

The structure of the goods traded shows that there are price incentives for unofficial importers from Kenya to Ethiopia through the border since most of the items officially imported are subject to taxes and duties. These impacts on government revenue and domestic industrial and trade policies and preoccupy governments that are on the disadvantage to impose control measures. These measures, however, involve both direct and indirect costs. Direct costs are government budget in financing the required institution. Indirect costs are more important and include foregone benefits by pastoralists from a larger market. Due to positive relationship between producers' incomes and wider and efficient markets, pastoralists in the area would benefit from cross-border trade. Cross-border markets offer better livestock prices. This would enhance food security for the pastoralists through increased income from livestock.

3.2 The Social and Economic Characteristics of Traders

Long-standing social relations based on ethnicity, kinship, and social variables help to shape the structure of cross-border trade. The profile of livestock traders in the borderlands shows that there is geographical, religious and ethnic concentration. This concentration is specifically high in markets near to the border. Along the Ethio-Kenyan border the Oromos (the Boran) and the Somalis (Gerri, Degodia and to some extent Marehan), dominate the cross-border trade. Most of these traders also have relatives in both countries and maintain multiple and a few carry triple country identity cards. These traders have access to larger markets in Kenya such as Garissa, Isiollo and Nairobi.

Table 3. Distribution of Livestock Traders by Wereda in the Ethiopia-Kenya Borderlands, 1998

SOURCE: Survey Result, 1999.

NOTE: Values in different rows are not comparable but values in different columns should sum up to 100 to reflect percent of the total interviews undertaken in the border markets. Percentages were computed from 100 trader interviews whose distribution by Wereda is indicated in Table 1.

In general, Table 3 and Table 4 below show that the cross-border trade is run by people who live in the area.

Table 4. Ethnic Distribution of Livestock Traders

SOURCE: Survey Result, 1999

People in this area speak primarily Oromiffa and Somali languages. These two languages are the most important medium of communication. Many traders speak both languages and some also speak Amharic in addition. While all traders of Somali clans are Muslims, the Boran are by large followers of Waqifata (religious practice by the Boran Oromo) and some are Muslims (Table 5).

Table 5. Religion of Livestock Traders by Ethnic Group (in percent)

SOURCE: Survey Result, 1999

Table 6. Level of Education of Livestock Traders

SOURCE: Survey Result, 1999

In terms of gender, livestock and grain-trade business are men's domain. Women's involvement in the border trade is limited in those small-scale ventures like sale of livestock products such as milk and butter, and other food items like vegetables that are primarily from own production.

Ethnic and religious concentration could entail substantial influence in the performance of production and trade in the southern and southeastern borderlands of Ethiopia. McIntire (1993) wrote that these social variables provide important coping strategies to missing markets for land, labour and capital in African pastoral production system. Cook et al (1990) and Shank (1997) also indicated the relevance of trust relationships among the same ethnic and or religious groups.

Table 7. Livestock Traders by Age

SOURCE: Survey Result, 1999

Table 8. Distribution of Livestock Traders by Number of Years Involved in

Livestock Trade

SOURCE: Survey Result, 1999 Range 1-25 years

Table 7 shows that traders are young which would imply the difficulty involved with contraband cross-border trade. In terms of experience, more than half of these livestock traders have started this business since the change of Government in 1991 Ethiopia (Table 8).

Another important socio-economic characteristic of the traders is their engagement in multiple activities. Trader interviews in the border markets indicated that almost all traders are engaged in other activities. However, the activities identified are very few and are limited in farming and small-scale business. These include, animal husbandry, crop farming, grain trade, retail shops of manufactured items and others. Out of these, farming accounts more than half of the activities (Table 9).

Table 9. Diversification/ Other Activities by Livestock Traders in Percent

SOURCE: Survey Result, 1999

Number of interviews=100

Trader's diversification would imply the risk and seasonality associated with cross border livestock trade. The risk emanated from drop in demand for livestock due to bans by importers in the Middle East, bans by the government to redirect the unofficial channel, theft and animal diseases. In addition, occasional droughts significantly affect livestock production in the area.

Seasonality in production and consumption is one of the manifestations of the livestock system in the border areas. Livestock supplies respond differently to dry and wet seasons of the year. Peak periods for consumption arise from relative importance of different time periods for meat consumption such as holidays with festivities like X-mass (Christian) and the Id (Muslim) in Kenya. Other demand shifters for livestock products in the country such as coffee and crop harvest are also seasonal (See Section-6 of this document for seasonality and livestock prices). Another reason for diversification could be that most traders from the Ethiopian side do not deal with large volume of sales that could be run as a full time venture and hence require additional activity to support their living.

There is also large scale of differentiation among the livestock traders as explained by high standard deviation and range.4 For instance, when we consider cattle, which is the most traded livestock group, we see that 55 percent of the livestock traders sell less than 200 cattle of various qualities and 35 percent sold less than 100 cattle (Table-10). Similar trend is also observed in small stock and camels (Table-11 and Table-12).

Table 10. Scale of Differentiation Among Livestock Traders Based on Annual Cattle Sales, 1998

SOURCE: Survey Result, 1999

Mean = 532; Standard deviation=1057; Range 10-7322; N=95 cases

Table 11. Scale of Differentiation Among Livestock Traders Based on Annual Sheep and Goat Sales in 1998

SOURCE: Survey Result, 1999

Mean=2388, Stdev=2686, Range 58-9883 (Min-Max)

Table 12. Scale of Differentiation Among Livestock TradersBased on Annual Camel Sales in 1998

SOURCE: Survey Result, 1999

Mean=116, Stdev=190, Range=4-980 (Min-Max)

In general, the livestock trade business that suffers from several constraints implying irregularities in the flow of income forced the traders to engage in other activities. This coping strategy allowed them to smooth consumption over time. Since other activities identified are also seasonal, the livestock trade business fills back the gap to sustain the flow of income. Thus the livestock trade business can also be regarded as a means of diversification of activities or opportunities for pastoralists and agro-pastoralists in the area when compared to other activities. This particularly holds true to small-scale livestock traders.

3.3 Markets and Market sheds

It is critical to examine both the markets and the areas (market sheds) from which these markets draw their supplies. As will be seen in this section, the market sheds for the cross-border trade with Kenya extends well into southern and south-eastern Ethiopia, with most of the livestock directed to the Nairobi market (See Map 2 & 3).

3.3.1 Markets

Livestock markets can be identified as bush, primary, secondary, terminal and export markets depending on who involves in the market (Ethiopian Livestock and Meat Board, 1971 and Gebremariam, 1976).

a. Bush or local markets

In these markets pastoralists and farmers congregate to exchange livestock for breeding, draught and slaughtering purposes either on barter or on cash basis. The exchange is between one farmer to another and between one stock-owner to another. The markets are dominantly of the breeding type and age. There is a very small flow to the trade stream. A few slaughter animals are also sold for local consumption.

b. Primary markets

These are the first points at which stocks enter trade. They are larger than bush markets. They are mostly found along livestock routes and are attended by all kinds of purchasers from farmers or pastoralists. Pastoralists buy breeding stock, butchers for local slaughter and traders for onward sale. In the pastoral areas distances to primary markets are considerable and if a producer does not sell he must trek his livestock to his holding, await the next market or trek to another market. Local farmers or pastoralists act as dealers and buy stock from farmers and pastoralists to resale in these markets. The dealers are in many cases also farmers or pastoralists whose trading activities are secondary. They have small capital resources and operate within a limited area.

c. Secondary or resale markets

These are the largest markets situated along stock routes leading to final destinations on all weather roads. They also help for the exchange of trade animals between the local traders and the larger traders who trek or truck onto terminal markets. They are attended to by farmers, pastoralists, traders, brokers, etc. and are usually located in farming and pastoral areas. Here, farmers buy breeding stock and draught replacements and sell draught culls, steers; and pastoralists buy breeding animals and draught replacements and sell bulls and steers. Traders buy culls from the farmers and slaughter stock from traders and move them to terminal or export markets.

d. Terminal, tertiary or consumers markets

Terminal markets are located outside or on the periphery of pastoral areas. They include markets in towns and cities and coffee growing areas; meat processing plants and industrial and mechanised farming areas. In these locations, the buyers are butchers, intermediary dealers and commission agents buying for processing plants, exporters and consumers.

e. Export markets

These markets are situated in the border countries and attended by traders from Ethiopia; and traders and trade livestock carriers from the neighbouring countries. When the export market is situated close to the borderlands, pastoralists attend markets.

Both livestock and other goods trade are effected in several places along the Ethio-Kenyan border. The most important livestock markets in Ethiopia are Negelle, Dollo-Ado, Arero, Dubluk, Mega, Teletelle, Finchawa, Wadera and Yabello area. These markets account the majority of livestock traded in southern and south east rangelands of Ethiopia.

The most important livestock markets in Kenya, for livestock traded in southern and southeastern rangelands of Ethiopia, include Mandera and Moyale (see Table-13). These two markets are the major recipients and it is estimated by GTZ/BLPDP and MoA that 70-80 percent of live animals sold in Mandera and Moyale (Kenya) originate from Ethiopia. Other border markets for livestock in Kenya that are found between Mandera and Moyale include Ramu, Banissa and Thakaba. These three smaller markets supply to Garissa, Isiollo and Nairobi markets.

Markets in Kenya are receiving markets for livestock from Ethiopia and supply markets for manufactured items produced in Kenya and unofficially imported to Ethiopia. Domestic terminal or consumer markets are found in Gedeo, Sidama, Konso, Dorze/Ochollo and Wolaita areas. Feedlot owners found in Nazareth and Dhera areas also purchase some cattle from Borana, Arsi and Bale area and these animals are fattened for three months for Addis Ababa and export markets.

Table 13. Classification of Livestock Markets for Livestock that Originate from Southern and Southeastern Rangelands of Ethiopia

Market visited			Classification of Markets
		Bush		Primary	Secondary	Terminal	Export	Livestock Traded
1. Border Market Catchment
1.1. Negelle		Alge, Kitta, Geda, Bulbul, Kalkalcha		Genale, Jidola, Harkallo Wadera	Negelle			Cattle, Camel, Sheep, Goats
1.2. Dollo-Ado/ Mandera		Bur Amino, Kole		Chiratte, Bare, Gode, Elkare, Filtu, Seddei	Dollo Ado		Mandera, Ramu, Thakaba, Banissa	Cattle, Camel, Sheep, Goats
1.3. Dirre		Selu, Boko Luboma		Mega	Dubluk			Cattle Sheep, Goats
1.4. Arero		Hirmaye/Obulo, Web		Matagafarsa (Arero)				Cattle, Sheep, Goats
1.5. Yabello		Didhara, Harobake Dambella Wachu		Yabello, Elwaya Suruppa				Cattle, Sheep, Goats
1.6. Teltelle		Bule Korma		Teltelle (Melami)				Cattle, Sheep, Goats
1.7. Moyale		Tuka, Kedaduma		Hidilola			Moyale-Kenya Isiollo, Nairobi	Cattle, Camel, Sheep and Goats
2. Domestic Market Catchment
2.1 Hageremariam				Finchawa	Hageremariam			Cattle, Sheep, Goats, Camels
2.2. Kebre Mengist				Wadera, Harkallo Jidola	Kebre Mengist			Cattle, sheep, goats
2.3. Wonago/Yirga Cheffe		Leku, Tulla				Dilla/Wonago/ Yirga-Cheffe		Cattle
2.4. Nazareth/Dhera					Diksis	Nazareth/Dhera		Cattle, Sheep and Goats
2.5. Addis Ababa (4 Markets)						Addis Ababa		Cattle

SOURCE: Survey Result, 1999

3.3.2 Market sheds

Market sheds are different regions in a given supply area whose consignments are destined effectively to different markets that are often located in opposite directions. The boundary between two market sheds is a place where a traded item fetches the same price if it is moved in both selling markets found in opposite directions (Steffen et al, 1998). A market-shed boundary can also be regarded as a supply region where the trader is indifferent to move its livestock to any one of the two directions. In the context of southern and southeastern rangelands of Ethiopia, three livestock market sheds can be identified based on three destinations, these are Somali markets, Kenyan markets and domestic markets. Kenyan markets are Moyale and Mandera and all others between these two. Domestic markets in Ethiopia include coffee producing areas of southern Ethiopia found to the north of southern and southeastern rangelands of Ethiopia, and other important markets like Addis Ababa, Nazareth and Dhera.

The market shed for the catchment in the Ethiopia-Kenya border market was identified using trader interviews. This part of the interview captured buying and selling markets by type of livestock traded. Accordingly, the market shed or the catchment area for the cross border trade include south of Hagermariam and Wadera line. This is the northern limit of the cross border livestock trade (see Map 3).

The boundary between the market sheds of domestic markets in Ethiopia and Kenya is different for different types of livestock. For instance, the boundary for small ruminants particularly for goats was further down to the borderlands in the 1980's for Ethiopia. Whereas, the market shed or the catchment area for the domestic markets is north of Yabello-Negelle line. This is the southern limit of domestic livestock markets (see Map 3). Expansion and shrinkage of market sheds or shifts in the boundary follow changes in relative costs and prices.

In general, the demand for livestock in the southern and southeastern rangelands comes from outside. Domestic markets are limited for livestock from southern and southeastern rangelands. One better price period is offered by the coffee and crop harvesting areas and the market shed expands down to the borderlands for the supply of livestock and shrinks in the non-harvesting season. The movement of the shed depends on the relative forces from both sides. For instance, livestock embargo by Saudi Arabia depressed livestock prices in 1998 and domestic markets were second best alternatives. Therefore, the supply was sustained irrespective of season.

3.4 Marketing Channel

The overall marketing channel of the cross-border trade is simple. In all types of commodities, it involves fewer transactions and ownership changes. For livestock trade, the border channel is crossed after, at least, two ownership changes. In the trader interviews, livestock traders ranked their two suppliers for different types of livestock bought in 1998. This part of the interview captured trader's suppliers. The suppliers were classified into two: farmers/pastoralists and other traders. The purpose of this classification was used to identify the complexity or simplicity of the marketing channel. From Table 14 we see that most of the cross-border traders buy from pastoralists/agro-pastoralists in the border area and hence after the second trader the border is crossed. The implication of this structure is that the integration between border markets and central and export markets of the neighboring countries are important to income and hence food security of pastoralists in the border areas of Ethiopia.

Table 14. First Rank of Traders for Source Supplier to Purchase Livestock (in percent)

SOURCE: Survey Result, 1999

Foodstuffs such as milk, vegetables and grain from Ethiopia are supplied to border markets in the neighbouring countries by pastoralists from their own production. The bulk of vegetables and grain originate outside the border markets, especially from the highlands. There are no other agents involved in the transaction of petty trading of these commodities. However, middlemen, transporters, feed and water suppliers and veterinary drugs and services are required for livestock trade. Other manufactured and processed food items and textile products are purchased from retailers in the border towns of the neighbouring countries.

The border markets in Kenya are supplied live animals by livestock traders and pastoralists in order of importance. Pastoralists and traders supply to both secondary markets in Ethiopia and border markets in Kenya. But the majority of the suppliers are traders and the role of pastoralists in the market diminishes as we move to the higher level of the chain. Supporting services are also provided by transporters or drovers and feed and water suppliers (see Figure 1 and Map 2).

Figure-1. Cross-border Livestock Marketing Channel from Southern and Southeastern Rangelands of Ethiopia to Kenya and Somalia

Export Destined

The marketing channel shows that the actors involved in the livestock trade increases at higher levels suggesting various labour and capital contracts. In the cross-border livestock trading from Ethiopia to Kenya there is a maximum of three ownership changes before the border is crossed. These are from pastoralists (herders) to the first trader usually in bush and primary markets and then to the second trader at secondary markets. Once the Somali and the Kenyan markets are reached, the involvement of Ethiopian traders drops substantially. The marketing channel on the other hand becomes more and more complex by involving different agents such as exporters, importers and consumers as well as support services required such as motorised transport, veterinary and quarantine services and customs. Thus, much of the value added from livestock sold at export markets is beyond the reach of the Ethiopian traders.

In examining the market channels and spatial characteristics of the cross-border trade, livestock are moved through several different routes. The study identified 12 export livestock trade routes to Kenya and Somalia and 7 domestic livestock routes that originate from Southern and southeastern rangelands of Ethiopia (See Fig. 2 and Map 2).

Figure 2. Cross-Border Livestock Export and Domestic Trade Routes from Southern and Southeastern Rangelands of Ethiopia

1. Cross-Border Livestock Export Routes from Ethiopia to Kenya and Somalia

ROUTE 1

NEGELLE MOYALE NAIROBI

ROUTE 2

FILTU SEDDEI NAIROBI

ROUTE 3

DOLLO ADO MANDERA WAJJIR NAIROBI

ROUTE 4

CHILANKO BANISSA WAJJIR NAIROBI

ROUTE 5

KEDADUMA THAKABA WAJJIR NAIROBI

ROUTE 6

AFDER LUUQ KISMAYU/MOGADISHU MIDDLE EAST

ROUTE 7

AFDER DOLLO ADO MANDERA WAJJIR NAIROBI

ROUTE 8

GODE/KELAFO WARDER BOSSASO MIDDLE EAST

ROUTE 9

DIRRE SOLOLO MARSABIT ISIOLLO NAIROBI

ROUTE 10

ARERO MOYALE MARSABIT ISIOLLO NAIROBI

ROUTE 11

FINCHAWA DUBLUK MOYALE NAIROBI

ROUTE 12

TELELLE HOBOK DUKENA NAIROBI

2. Domestic Routes

ROUTE 1

NEGELLE DILLA YIRGA CHEFFE

ROUTE 2

NEGELLE DELLO BEKOJI DHERA NAZARETH ADDIS ABABA

ROUTE 3

DIRRE (MEGA) DUBLUK FINCHAWA YIRGA CHEFFE WONAGO/DILLA

ROUTE 4

ARERO FINCHAWA YIRGA CHEFFE WONAGO

ROUTE 5

TELTELLE WATA WONDO SOUTH OMO

ROUTE 6

TELTELLE KONSO ARBA MINCH DORZE/OCHOLLO

ROUTE 7

TELTELLE YABELLO YIRGA CHEFFE WONAGO

NAZARETH ADDIS ABABA

The livestock border trade from Ethiopia to Kenya is one way. There is no livestock trade from Kenya to Ethiopia. Similarly, the flow of other products such as milk and butter, vegetables and grain is also unidirectional, i.e. from Ethiopia to Kenya. These products are consumed in the borderlands, whereas, a large proportion of the live animals is transported to the center towns including Nairobi and some are exported by Kenyan traders to the Middle East.

3.5 Currencies

Transactions in the Ethiopia-Kenya border are effected with most flexible form of payment, money. According to the results from the trader interviews, we conducted in December 1998 and January 1999, the instances of barter arrangements were very few. Out of 100 livestock traders interviewed in the border markets 86 percent of them sold one or more types of livestock in the neighbouring markets. Out of these exporters, 77 percent responded that, in return to their goods, they received currencies only. The rest, 23 percent, said that they mixed other goods, mainly foodstuff, in addition to money.

Three types of legal tenders namely the Ethiopian Birr, the Kenyan Shillings and the Somali Shillings are used in various degrees in different sites of the border markets. The Kenyan Shillings was paid to 86 percent of the exporters. In terms of distribution across markets, the Birr and the Kenyan Shillings are used in most corners of the border, while the Somali Shillings is restricted to the northeastern part of Kenya-Somalia border and the Somali inhabited areas of Ethiopia. Currency holdings are also ethnic specific. So far as the Somali Shillings is concerned, it is exclusively used by the Somali clans, who have easier access to Somali markets such as Baidoa and Mogadishu because of location, language and culture. This part of the Ethiopia-Kenya border links the other border trade between Ethiopia and Somalia and between Kenya and Somalia.

Table 15. Type of Currency Received from Cross-border Livestock Trade

SOURCE: Survey Result, 1999

From Table 15 we see that 86 % and 6 % of the traders received Kenyan Shillings only and Ethiopian Birr and Kenyan Shillings respectively. Ethiopian traders in most sites, in Kenyan border markets, received Kenyan Shillings in return for their livestock sales and they exchanged the currency for Ethiopian Birr. However, most of the Ethiopian-Somalis who sell in Mandera and surrounding markets of Kenya exchange the Kenyan Shillings to the Somali Shillings. There are two reasons for the preference of the Somali Shillings in this area. One is that the Somali Shillings is used as money in many areas of the Ethiopia Somali Region. The other is that the traders use the currency to import other manufactured goods and some food items from Somali markets.

The Ethiopian livestock trader sells livestock and receives money in Kenyan Shillings and exchange it to Ethiopian Birr at the point of sales and brings the money with him to Ethiopia unlike the view that the trader buys manufactured or other goods after selling his livestock. This is also true of other cross border traders who buy manufactured goods. This shows that there is some kind of specialisation. The survey reveals that there are no traders both on livestock and manufactured commodities considerations and vice versa.

There are three possible explanations for this. First is that livestock traders from the Ethiopian side deal with very small volume of sales, the return of which is required for immediate purchases like grain, breeding stock and other essentials, in Ethiopia. In this case, therefore, the border trade may be sought purely for price incentives from one way trade only.

The second reason could be convergence or shift in agent groups in the return trade. Thus, only few traders or other groups of people involve in the manufactured items trade from Kenya and Somalia to Ethiopia. This could be because the return trade is sophisticated and risky requiring different arrangements than the livestock trade. Moreover, since the return manufactured goods could have different channels in marketing and consumption as well as market sheds, this could have been inaccessible for livestock traders.

The third possible reason is that some items imported from other countries in Asia are cheaper in Somalia than in Kenya. This is why traders who sell in the northeastern part of Kenyan markets such as Mandera exchange the Kenyan Shillings, received from livestock sales, for the Somali Shillings. However, not all the Somali Shillings is spent for this purpose. Part of it is used to buy some items in Ethiopia since this currency is used in many areas of the Ethiopian Somali Region.

Despite this flexible mode of payment in any of the three currencies in the border areas, exchange rate fluctuations entailed speculations by different agents. For instance, a speculation for exchange rates between the Ethiopian Birr and the Kenyan Shillings is due to the fact that changes in the relative flows of all types of goods across the border that is not officially accounted and relative price changes in the two countries. Thus, when more goods flow, for instance from Ethiopia to Kenya, the price of the Ethiopian Birr in terms of the Kenyan Shillings increases or vice versa. Therefore, Ethiopian traders paid in Kenyan Shillings would loose the premium if their price expectations for their sales were more static. The situation implies that traders would do away with prolonged arrangements to insulate themselves against this risk.