Responding to Afghanistan's Opium Economy Challenge: Lessons and Policy Implications from a Development Perspective

Opium, Afghanistan's leading economic activity, lies at the heart of the challenges the country faces in state building, governance, security, and development. With their narrow law enforcement focus and limited recognition of development, security, and political implications, current global counter-narcotics polices impose a heavy burden on Afghanistan. This paper first provides a summary overview of Afghanistan's opium economy and the factors determining rural households' decisions on cultivating opium poppy. It then discusses the dynamic evolution of the Afghan drug industry in recent years, in particular its consolidation around fewer, powerful, politically-connected actors and the associated compromising of parts of some government agencies by drug industry interests. The paper reviews the experience with different counter-narcotics interventions, analyzes some proposals not yet tried in Afghanistan, and draws lessons and policy implications. Unfortunately there are no "silver bullets" -easy, quick, or one-dimensional solutions, and a longer-term horizon along with sustained commitment and resources will be required in order to phase out the opium economy over time. The paper concludes by putting forward some broad principles and approaches of a "smart strategy" against drugs in Afghanistan.


I. INTRODUCTION
The opium economy lies at the heart of the challenges Afghanistan faces in state building, governance, security, and development. Accounting for around a quarter of total economic activity in Afghanistan, its magnitude and importance are virtually unprecedented in global experience. Since 2002 efforts to reduce the size or even limit the expansion of the opium economy have failed. In the meantime Afghanistan's drug industry has evolved in directions which further exacerbate the threat it poses to the country's entire state-building and development agenda. Counter-narcotics measuresdesigned largely in isolation from the other interventions, implemented in a fragmented and often piecemeal or inconsistent manner, and suffering by all accounts from widespread corruption during implementation-contained the seeds of their own failure. This paper reviews the experience with counter-narcotics efforts in post-Taliban Afghanistan, derives some lessons from this experience, and draws out implications for policy. The main finding is that there are no "silver bullets" (easy or single-dimensional solutions) and that in order to have better prospects for success, the different counternarcotics instruments must be deployed in a much more intelligent way, with modest expectations and a long time horizon but strong and sustained commitment accompanied by adequate resources. The broad principles and approaches of a "smart strategy" in response to the drug industry in Afghanistan are put forward.
Since it deals with the opium economy in Afghanistan, this paper focuses very much on the supply side of the narcotics equation. Although there are concerns about growing use of illicit narcotics in Afghanistan (see MacDonald, 2007), which are touched on in this paper, the main threat to the country's development emanates from the cultivation, trade, and processing of opium and associated criminality and corruption. However, the difficulties in curbing the opium economy in Afghanistan are orders of magnitude greater because of the high world and regional demand for illicit opiates. Moreover, with their narrow law enforcement focus and limited recognition of development, security, and political implications, current global counter-narcotics policies impose a heavy burden on Afghanistan. And finally, even if the country were able to make progress in reducing opium production, in the absence of broader changes on the demand side production would most likely shift elsewhere, as has been demonstrated by international experience.
The rest of this introductory section provides some historical background, summarizes Afghanistan's opium economy from a development perspective, and highlights its strategic importance. Section II outlines the structure of the opium economy and recent trends. Section III analyzes determinants of opium poppy cultivation and the dynamic evolution of the drug industry. Section IV reviews the experience with counter-narcotics interventions in Afghanistan since 2001. Section V draws out some key lessons and puts forward implications for policy.

Historical Background
The genesis and subsequent history of large-scale opium production in Afghanistan have been intimately linked with the wars and upheavals in the country and in the surrounding region during the last two decades of the 20 th Century. Opium has been produced for a very long time in Afghanistan, but until the end of the 1970s this was traditional production on a small scale, largely limited to a few areas and primarily for local or regional consumption.
The Soviet Union's occupation of Afghanistan at the end of 1979, the emergence of a theocratic regime in Iran in the same year, and the development of the opium processing industry in Pakistan (which also cultivated opium poppy at the time), as well as developments farther away (for example in Turkey), together created the enabling conditions for massive expansion of opium poppy cultivation in Afghanistan. Opium became a lucrative source of financing for the mujahideen resistance forces fighting against the Soviet occupation, and the linkages to processing facilities in Pakistan paralleled those between Afghan resistance forces and Afghan political parties in Pakistan that were sponsoring and supporting the resistance. Iran's abrupt elimination of opium poppy cultivation at the beginning of the Khomeini regime, Turkey's shift to licit production, and Pakistan's more gradual phase-out of opium poppy cultivation (while remaining a very important location for opium processing and the narcotics trade) provided "space" in the world market for Afghanistan to emerge as a major exporter of opium (including to meet Iran's domestic consumption requirements). Although reliable data are not available, it is clear that Afghanistan became a very significant opium producer by the mid-1980s.
After the departure of Soviet forces in 1989 and especially after the collapse of the Najibullah regime in 1992, international financing for armed groups in Afghanistan was sharply reduced, further enhancing the relative importance of opium in providing funding for factions in the civil conflict which ensued. The Taliban regime, which took over Kandahar and much of the south in 1994, conquered Kabul in 1996, and controlled some 90% of Afghanistan's territory by the end of the decade, provided an environment in which opium production and trade could flourish. Essentially treating it as a legal crop, the Taliban collected religious tax (ushr) on opium at a low rate, as in the case of other agricultural products. Estimates of opium poppy cultivation, which were made on a more systematic basis by UNDCP (subsequently UNODC) starting from 1994, showed continuing increases to a peak of more than 90,000 hectares in the 1998/99 season, when Afghanistan accounted for close to 80% of total global illicit opium production.
Before the 2000/01 growing season, in what turned out to be their final year in power, the Taliban regime effectively banned opium poppy cultivation (but not trade) in the territories it controlled. While the motivation for this ban is subject to speculation, and major drug industry actors may have gone along with it in part because of oversupply and large stocks from previous bumper harvests, this was unquestionably the most successful and cost-effective short-run reduction in production of illicit narcotics achieved in history. However, the sustainability of this blanket ban was very doubtful, although the Taliban were overthrown before this question could be answered definitively one way or the other. There is evidence that the ban hurt the Taliban politically, and planting of opium poppy resumed in the second half of 2001 in many places even before the end of the Taliban regime. Moreover, during the ban opium poppy cultivation in the one province completely outside the Taliban's control (Badakhshan) increased by an estimated 160%, with some heroin processing facilities reportedly also moving there.
Thus the situation inherited after the downfall of the Taliban was one in which opium poppy cultivation had been almost completely eliminated in the previous year, but extensive planting of opium poppy was occurring, and within two years poppy cultivation and opium output were back to "normal" levels similar to those seen in the 1990s. The high farm-gate price of opium induced by the Taliban ban, which persisted for several years, as well as efforts by drug industry actors to diversify beyond the main production areas in the south, led to the emergence of extensive opium poppy cultivation in nontraditional growing areas in other parts of the country.

Strategic Importance and Development Perspective
The opium economy is one of several critical issues facing Afghanistan. It relates closely and in complex ways not only to the economic growth agenda and poverty, but also to state-building, the political process, governance, security, and counter-insurgency. The strategic integration of all these issues is essential for Afghanistan to make substantial and sustained progress in the face of a complex and inter-linked set of development challenges.
The opium economy and the insurgency both thrive in an environment where there is insecurity, lack of rule of law, and a weak and corruptible state. Thus even though their interests are by no means always intertwined, there are synergies between the Taliban and drug interests (including notably in Helmand province) that damage Afghanistan's state-building agenda. The close relationships between drug traders, warlords-turned-politicians, and corrupt officials in government agencies that have been partly compromised by the drug industry (for example the Police and Ministry of Interior) is another important example of the strategic linkages associated with the drug industry, discussed further in Section III.
The opium economy provides substantial incomes to segments of the rural population, stimulates aggregate demand, and supports the balance of payments, although it has only secondary and indirect benefits for government revenue. However, as argued by Martin and Symansky (2006), the opium economy's macroeconomic impact is less than might be expected from its sheer size, because much income beyond the farm level never enters Afghanistan in the first place, and some goes right out again in the form of capital flight or import financing.
The opium economy is also contributing to the "Dutch disease" in Afghanistan by providing an influx of financial resources and driving up rural wages. Labor in opium harvesting as well as (at relatively high levels of risk) opium trading earns such high returns that shifting to other, licit activities is discouraged. Moreover, as it has become entrenched in some areas and has been a major economic activity for some two decades, the opium economy affects asset prices (most notably the price and rental/sharecropping rates for agricultural land in and around opium producing areas) and non-opium business activities. However, even at present record levels of cultivation, opium poppy still takes up only a small proportion of Afghanistan's agricultural land overall.
In sum, the opium economy poses a complex development challenge. On the one hand, it contributes heavily to local incomes; on the other hand, its illegality and associated corrupt and criminal activities undermine the basic institutions of the state. In this context, poorly designed and implemented counter-narcotics measures can have an adverse development impact of a similar magnitude to the damage caused by the opium economy itself, possibly even greater. The poverty impact of such measures-resulting from reductions in the incomes of farmers cultivating opium poppy (most of them sharecroppers or tenants on others' land) and of wage laborers employed in opium poppy cultivation and harvesting-can be very significant. Both the nearly nationwide Taliban ban of 2000 and the 96% reduction in the cultivated area for opium poppy in Nangarhar Province in 2005 exacerbated poverty, both directly and through opium-related debt and through coping strategies like asset sales, as well as through multiplier effects on the local economy. Thus the development and poverty implications of both the opium economy and actions against it need to be fully taken into account in the development strategy and counter-narcotics strategy.

II. THE OPIUM ECONOMY: OVERALL PATTERNS AND TRENDS
After looking briefly at data and research issues, this section summarizes our knowledge of the opium economy and recent trends. Cultivation and production, trade and processing, opium prices, drug-related financial flows, and what little is known about the "commanding heights" of the drug industry are touched on.

Data and Research Issues
Quantitative information on Afghanistan's opium economy is limited and of varying quality and reliability. This is not surprising given its illicit and informal nature, as well as the weaknesses of Afghanistan's statistical system in general. Moreover, logistical and security constraints seriously hinder the collection of primary data on the opium economy. In addition there are technical issues, for instance related to the coverage and interpretation of satellite imagery. And identification and assessment of trends is complicated by varying reliability and sometimes changing collection and estimation methodology for data over time.
Nevertheless, data on the opium economy are generally no worse, and in many respects better, than the data available on the rest of Afghanistan's economy. Estimates of the opium poppy cultivated area are produced by UNODC 1 on an annual basis using remote sensing supplemented by a survey, although estimates of yields (and therefore of opium production) are less reliable. Opium price data also are collected on a regular (monthly) basis in an increasing number of provinces. Moreover rural households, smaller drug traders, and hawala (informal money transfer) dealers have been accessible for careful interviewing and information collection. Thus overall, data issues have not prevented meaningful research on Afghanistan's opium economy (see Byrd and Buddenberg, 2006, p. 4).

Cultivation and Production
Keeping in mind data limitations, summary information on opium in Afghanistan is presented in Table 1. Among the various estimates, those of the total area under opium poppy cultivation are the most reliable but still have significant margins of error. 2 Yield estimates have a greater margin of error, particularly when disaggregated to the provincial level. Compilation of the estimated farm-gate opium price introduces a further, though likely smaller, margin of error, so the end result is that the estimated farm-gate income has a considerably larger margin of error than the cultivated area estimate. Assumptions about border prices, from which the total potential export value and (as a residual) the gross income beyond the farm level are calculated, introduce substantial further unreliability into these numbers. As % of total agricultural land n/a N/a n/a n/a 1.6 2.9 2.3 3.65 4.27 Area under poppy / Area under cereals (%) 2.0 3.2 n/a 3.2 2.8 5.9 n/a n/a n/a  Government of Afghanistan (2004. 1 The US also produces annual estimates of the total area devoted to opium poppy cultivation, which in recent years have been fairly similar to UNODC's estimates. However, there are wide discrepancies between US and UNODC estimates of the opium poppy cultivated area in individual provinces. For convenience and consistency, UNODC estimates are used throughout this paper. 2 For example, in the case of the 2004 estimate of 131,000 ha of opium poppy cultivation, UNODC (2004, p. 21) indicated that the "range" of possible estimates was from 109,000 ha to 152,000 ha, implying a margin of error (90% confidence interval) of around plus or minus 16-17%. In 2006 the range of estimates was somewhat smaller, between 150,000 ha and 180,000 ha, for a margin of error of plus or minus 9% (UNODC, 2006, p. 115).
Amid annual fluctuations, total national opium poppy cultivated area has shown a generally rising trend since the early to mid-1990s, interrupted by the Taliban ban which almost wiped out the 2001 harvest, and reaching new peaks in 2006 and 2007. Estimated opium production shows broadly similar trends, although percentage changes differ, reflecting fluctuations in estimated opium yields. Estimated gross income per hectare and gross farm income rose very sharply after the Taliban ban (reflecting a supply shockinduced spike in prices, shown in Figure 2 later in this paper), and even after half a decade remain considerably higher than in the 1990s. This may reflect in part a higher risk premium in farm-gate prices as a result of criminalization of the opium economy and intensified (albeit fragmented and uneven) counter-narcotics efforts.
National cultivation trends mask major diversity across provinces, selected examples of which are shown in Figure 1. Cultivation estimates for some provinces tend to move together (at least fluctuating in the same direction), often with somewhat offsetting changes from year to year. In other cases, fluctuations across provinces are partially offsetting within a year. In 2005 for example, the year in which cultivation in Nangarhar Province declined by 96% due to a largely effective ban on cultivation imposed by the provincial authorities, cultivation in Kandahar and Balkh rose sharply and in Farah it more than quadrupled, largely offsetting the impressive decline in Nangarhar. There is also great diversity at local (district) level and, as demonstrated by extensive fieldwork, across households, although regularities in the parameters influencing decisions on opium poppy cultivation are evident Source: UNODC (2003UNODC ( , 2004UNODC ( , 2006UNODC ( , 2007.

Trade and Processing
Less is known about the trade in opium in and around Afghanistan, its conversion into refined products (morphine and heroin), and trade in these products. Nevertheless a rough picture can be gleaned from field research and interviews with (mostly smaller) opium traders (notably Pain, 2006b), reinforced by information on drug seizures in neighboring countries. It is worth emphasizing that unlike many other agricultural products, opium is a durable good, with a shelf life of several years-longer than heroin powder. 3 This means that sizable inventories of opium can be and are maintained, that opium can be and is used as a form of saving and even as "currency", and that speculation in and sizable capital gains and losses on opium inventories can occur with fluctuations in prices. In fact, observed changes in prices and smoothing of supplies in major consuming countries can be explained only by large adjustments of opium inventories in the face of fluctuations in production.
There are many thousands of smaller opium traders, typically operating on a parttime and seasonal basis (e.g. shopkeepers). At this level opium markets have been characterized by frequent entry and exit, and higher opium prices following the Taliban ban attracted more small traders into the opium business. Trade margins for smaller traders are relatively low, except where proximity to or crossing of borders results in significant risks of interdiction and associated risk premia. Research suggests that drug traders often have a background trading in licit goods, and that they do respond to financial incentives and risks in their decision-making about whether and how much to trade in opium and opiates. Based on fieldwork the most important source of risk for traders has been price fluctuations, although more recently the risk of seizure or theft by authorities appears to have increased (Pain, 2006b).
Moving up the "pyramid" of the drug trade in Afghanistan, fewer and fewer, and individually increasingly important, actors are involved (see Shaw, 2006, p. 204), culminating with no more than several dozen key traffickers at the top. There are important linkages between higher-level elements in the drug industry and some warlords and their militias, as well as with government officials and some of the figures active in the conflict-affected politics of Afghanistan.
There are no signs that the drug industry is a monolithic cartel or is functioning like a cartel in pricing or other behavior, but entry at the middle and upper levels, and in some areas even at the lower levels, is becoming more difficult (Pain, 2006b, andShaw, 2006). In addition, there are signs of cooperation and "regulation" which indicate that when it is in their interests, different elements of the drug industry can work together effectively, including across ethnic lines. By the same token, although some of the fighting in the south as well as elsewhere may be explained in part as drug-related conflict, all-out "drug wars" between criminal gangs of the kind seen in some other countries appear not to have been the norm in Afghanistan.
Finally, the span of control and influence of even the major Afghan drug traders does not appear to extend very strongly or far beyond the borders of Afghanistan. Prior to the early 1990s, the bulk of opium produced in Afghanistan was processed into morphine or heroin in neighboring countries, mainly Pakistan. In recent years, however, most Afghan opium has been processed in-country. This major transformation reflects in large part Pakistan's efforts to drive out heroin processing labs from its territory, which culminated in the mid-1990s (see MacDonald, 2007, pp. 86-87). As in the case of opium poppy cultivation itself, heroin processing activities have gravitated toward Afghanistan where the "enabling environment" in the form of insecurity, lack of rule of law, protection provided by armed militias, etc. remains conducive for such activities. However, after drug shipments cross the border, other trafficking groups, associated with the neighboring countries or more transnational in nature, appear to take over.

Price Patterns and Trends
Considerable data on opium prices are available (see UNODC, 2003 and and can be analyzed, albeit with caution. It should be noted that the farm-gate opium price comprises only a small part of the price of opiates at Afghanistan's borders, and a truly minuscule percentage of the wholesale or retail price in OECD consuming countries (Byrd and Jonglez, 2006, pp. 130-131).
As shown in Figure 2, there have been major fluctuations in prices of raw opium, most notably the sharp spike in prices associated with the Taliban ban. 4 This was followed by persistence of high prices for several years. More recently, farm-gate opium prices have declined but have remained at levels still well above those prevailing in the 1990s, despite large increases in production in the face of limited increases in global demand. This suggests that the "risk premium" associated with opium poppy cultivation may have risen considerably, probably reflecting criminalization along with significant albeit patchy and haphazard enforcement efforts including eradication, and likely greater extortion of "protection money" from farmers by various authorities. However, prices are currently being pushed down by the very large increases in output in 2006 and 2007.
Quantitative analysis of farm-gate opium prices, which makes use of several instruments ranging from simple correlation coefficients to linear regression and more sophisticated co-integration techniques (see Byrd and Jonglez, 2006), indicates that: • Opium markets are flexible and mobile; while actions against the opium economy can be effective locally and in the short run, they encourage shifts of production and trade to other areas. • Regional and, in particular, cross-border price differentials suggest that interdiction of opium trade, particularly at borders, can have a significant impact. • Available price data indicates that internal opium markets appear to have been more "integrated" (based on a technical definition of the term) during the 1990s than in recent years, perhaps reflecting the disruptive effects of counter-narcotics actions on opium markets. • Price data for recent years suggest that Helmand/Kandahar in the south is functioning as a "central market" for opium in Afghanistan.

Drug-related Financial Flows
The bulk of drug-related financial flows within Afghanistan, and also to and from neighboring countries (primarily Pakistan), occur through the ubiquitous hawala (informal financial transfer) system. Hawala is based on informal yet very solid networks of trust and business relationships, under which money transfers in opposite directions are offset against each other, and any remaining imbalances are settled through transfers between dealers (see Maimbo, 2003). Very little physical transfer of money needs to occur, hawala dealers can operate effectively with small cash reserves, and the system is remarkably efficient (as evidenced by small spreads in exchange rates quoted).
Recent analysis based on extensive interviews with hawala dealers (Thompson, 2006) provides insights into the nexus between the drug industry and hawala, and the considerable variation across different parts of the country. In the economically less developed province of Badakhshan, for example, field research indicates that at certain times of the year close to 100% of the liquidity in the hawala system is derived from drugs. On the other hand, in a much more developed province like Herat, only 30% of the hawala market's overall transaction volume appears to be linked to drugs, although the analysis of such linkages is complicated by use of drug money in the legitimate import business. In addition to being a center of opium production and trade, the southern region (Helmand and Kandahar provinces) is a focal point for money laundering: apparently about 60% of hawala flows are drug-related, and 80-90% of hawala dealers are involved in drug-related money transfers.
Beyond Afghanistan's borders, Dubai appears to be a central clearing point for international hawala activities, and various cities in Pakistan also are major transaction centers. Even payments for drug shipments to Iran enter Afghanistan from Pakistan. Transfers of funds from major drug consuming countries to regional countries like Dubai and Pakistan appear to occur largely through the formal banking system; hawala becomes dominant in the onward transfers of funds into and within Afghanistan.
The hawala system plays other important roles in addition to drug money laundering. Its positive contributions include serving as an efficient vehicle for remittances both during the long period of conflict and more recently; providing money transfer services in the many parts of Afghanistan where no banks exist; participating in foreign exchange and nascent treasury bill markets; and playing an instrumental role in the successful introduction of a new, stable currency for Afghanistan in 2002-2003.

Determinants of Household Decisions
Since opium poppy is an annual crop, rural households in Afghanistan make decisions every year on whether to plant opium poppy, how much to plant, and how to organize the required labor and other inputs, as well as on when and how to sell (or store) the output. Relative to its high value, opium poppy economizes on land and water use (although it requires decent, non-waterlogged soils and adequate water at the right times), but it is highly labor-intensive, and skilled labor is at a premium during harvest time. Market linkages for sale of the raw opium harvested are very strong (especially as compared with those for licit agricultural products), and drug traders also can make available key inputs-in particular credit and seeds-as necessary.
Extensive fieldwork conducted during the past decade has provided valuable insights into the various factors influencing rural households' decisions on opium poppy cultivation. The best of this research (notably by Mansfield, Pain), undertaken at great personal risk, has built up a significant degree of longitudinal knowledge-of provinces, localities, and even some households, as well as a wealth of cross-section information.
This research reveals that, although farm-gate prices of opium provide signals for producers and are a major determinant of incomes (see Byrd and Jonglez, 2006), a onedimensional price-based model of farm-level decision-making with respect to opium poppy cultivation does not fit the facts found in fieldwork, or even the broad trends seen in aggregate data. Changes in cultivation patterns at household and locality levels respond to many factors, of which the farm-gate price of opium is only one (albeit a very important one). These factors are intimately related to the development challenges confronting Afghanistan, and they highlight that a counter-narcotics strategy can only succeed if it is nested in and consistent with a broader development strategy.
However, eradication efforts and enforced production reductions sometimes do have major price effects that significantly affect cultivation decisions. Particularly if the reduction in cultivation is very large (exemplified most strikingly by the Taliban ban in 2000/01), the associated increase in farm-gate opium prices can be quite sharp (more than 1,000 percent in the short run at the time of the Taliban ban). This sends a very strong market-based signal for expansion of opium poppy cultivation in areas where the ban does not apply (non Taliban-controlled areas in the case of the Taliban ban) or is not enforced. High prices also encourage areas with more marginal potential to engage in opium poppy cultivation, as occurred in the case of Ghor after the Taliban ban (Pain, 2006b and. Household assets play a key role in guiding cultivation decisions, as argued by Mansfield ( , 2007a. These assets, broadly understood, include the number of ablebodied males and their labor skills (e.g. in opium poppy harvesting), agricultural land, irrigation water, proximity to labor markets, and jobs that pay regular salaries (e.g. in government), as well as more conventionally defined physical assets (e.g. livestock, vehicles). Households with relatively few such broadly defined assets have fewer (if any) viable alternatives to opium poppy cultivation or engaging in wage labor in the opium economy. More asset-rich households, on the other hand, have more choices and opportunities for viable licit livelihoods and hence will tend to be much less dependent on opium, even though they may cultivate poppy opportunistically to increase their incomes. The implication is that law enforcement efforts as well as political and moral pressure can encourage better-off households to eschew involvement in the opium economy.
Access to commodity markets also can be viewed as an "asset" which reduces households' dependence on opium. The growth and extension of local vegetable markets in areas of Nangarhar close to Jalalabad city provide a good example of how improved access to markets can lead to sustainable reductions in opium cultivation. There is also evidence of such factors at work near other provincial capitals, cities, and transport routes.
Another broadly defined "asset", which is important but affects a locality or area rather than households individually, is a modicum of security for persons and property, at least sufficient to conduct small-scale economic activities and transport agricultural produce. The massive expansion of opium poppy cultivation in southern Helmand province occurred when the Taliban insurgency there was intensifying, and other examples demonstrate the linkage between insecurity and opium at a more micro level (see Mansfield, 2007c for a study of two districts in Badakhshan province in this regard). 2007a) finds evidence of the importance of such assets in the initially successful effort to sharply reduce opium poppy cultivation in Nangarhar province in 2004/05. The Nangarhar opium ban has turned out to be largely sustainable in more central localities where most households are higher up along the asset spectrum and in particular have relatively good access to commodity and labor markets, and have shifted successfully and on a sustainable basis to licit economic activities. In fact, after an adjustment period usually of not more than 2-3 years at most, such households can actually become better-off than when they had been cultivating opium poppy, in particular by taking advantage of a combination of different production and labor market opportunities available when household labor is freed up from labor-intensive opium production.
However, more remote areas where households have fewer assets suffered severely from the ban and by the third year have been reverting to opium poppy cultivation. In the worst-off areas the ban was not fully implemented from the beginning.
Forcing households and localities that lie toward the lower end of the asset spectrum to forego cultivating opium poppy has led to drastic coping responses like asset sales, migration, and the like, which increase rather than reduce their underlying dependence on opium. Given their very meager assets and limited alternatives, the opportunity cost for such households of engaging in opium poppy cultivation is very low, and their decisions in this regard may not be affected by law enforcement actions or pressures Erosion or loss of some of the assets discussed above often constitutes an important "push" factor for households to become engaged in the opium economy. For example, in studying the main opium-producing areas in the northern province of Balkh, Pain (2006a and points to local population growth (including through return of displaced persons), and running down of irrigation systems which reduces water availability, as having made opium poppy cultivation a relatively more attractive alternative as compared with other crops. In the case of Ghor Province, loss of livestock due to the severe drought of the late 1990s was an important factor for both traders and farmers to become involved in the opium economy. Declining security in southern Helmand Province since 2005 appears to have been an important factor contributing to massive expansion of opium poppy cultivation in that province.
Historical and social factors also play a significant role in cultivation decisions. Pain (2007) argues that basic structures (agro-ecology, settlement history, and ethnicity of a locality), the social positions of individuals within a locality (including ethnicity within the local context and socio-economic position), and intermediary factors (community, markets, institutions, and behavior) together influence decisions on opium poppy cultivation (see Pain, 2007, Figure 1, p. 7). While recognizing that market-based price signals can encourage wider diffusion of opium poppy cultivation, as appears to have occurred in Balkh Province post-2001, he argues that informal "regulation" of markets and ethnic or other linkages with the drug trade play an important role. In particular, pre-existing ethnic or other ties that facilitate the drug trade and transfer of labor and techniques comprise another enabling factor. For example, ethnic Pashtuns who had been transplanted to Balkh decades earlier, but who retained ties with their tribes/ethnic group in the southern opium-cultivating provinces, facilitated the spread of opium poppy cultivation to Balkh (see Pain, 2006a). Moreover, existing trading networks for other goods, irrespective of ethnic connections, can help facilitate the opium trade when conditions are right. Pain (2006b) documents how many opium traders in Ghor had their origins in the livestock trade, which dried up as herds were decimated in the late 1990s due to the severe and protracted drought.

Broader Drug Industry Dynamics: Changing "Vicious Circles"
We now turn to dynamic patterns and trends in the drug industry as a whole. Figure 3 depicts a vicious circle involving the opium economy, warlords, and insecurity-broadly reflecting the situation as opium production rebounded in the first two years following the downfall of the Taliban. In this situation, payments from the opium economy strengthened warlords, who in turn undermined the state, while drugrelated corruption also undermined the state directly. In return for payments, warlord militias helped provide the enabling environment (often including armed protection) for the opium economy to operate. The weak government was unable to provide genuine security or rule of law, and this created a good environment in which the opium economy could continue to thrive. Thus the dynamic tendencies at work would perpetuate a large opium economy and a weak, ineffective state (particularly in terms of providing security).

Figure 3: The Vicious Circle of the Drug Industry
Source: Adapted from World Bank (2005) This vicious circle suggested that a multi-faceted strategic framework would be needed to effectively address the opium economy and the problems it causes for Afghanistan's development agenda. In addition to reducing the size of the drug economy through effective counter-narcotics measures more narrowly construed, this framework would have needed to include: (i) curbing warlords' powers by stopping payments and other support to them, Disarmament, Demobilization, and Reintegration (DDR) to take away their militias, and co-opting them into the Government as appropriate; (ii) building government capacity and effectiveness as well as resources; and (iii) reform and capacity building in the security sector (see World Bank, 2005, Figure 7.4, p. 127). A strategic framework along these lines appeared attractive, and several of the key elements were in place to some extent or at least initiated. However, improvements at the broader strategic level fell far short of what was needed. As a result, there has been massive further growth of the opium economy, and both the opium problem and its adverse impacts on the state building and development agenda have become worse.
As depicted in Figure 4, the transformation of warlords into politicians has been accompanied by compromising of parts of some government agencies like the Ministry of Interior and Police by drug industry interests. The strengthening triangle between drug interests, their political and other sponsors, and parts of the government reflects a trend that-primarily through widespread corruption in their implementation-counter-narcotics efforts have inadvertently contributed to drug industry consolidation (see Shaw, 2006). Security forces, most notably the police, are in part facilitating the activities of the drug industry rather than countering it. Overall, this dynamic evolution of the drug industry constitutes a profound threat to Afghanistan's state-building and development agenda. And the fundamental equation between a weak state (partly compromised by drug interests) and a thriving opium economy remains. Moreover, the expanding Taliban insurgency in the South (not depicted in Figure 4) adds complexity to the picture and helps provides an enabling environment of insecurity for the drug industry, further exacerbating the associated risks.

IV. COUNTER-NARCOTICS EXPERIENCE IN AFGHANISTAN
Afghanistan has a National Drug Control Strategy (NDCS), first prepared in 2003 and subsequently updated and refined several times, most recently in January 2006. The goal, priorities, and pillars of the NDCS are outlined in Box 1. There is much to be commended in the NDCS, including its explicit emphasis on the need for a multi-year time horizon and for sustainable progress against drugs, the importance of alternative livelihoods being available in order for eradication to play its role in the counter-narcotics strategy, the rejection of aerial and chemical spraying, etc. However, the NDCS is not specific on short-run prioritization and sequencing, regional targeting, or on the lessons learned from earlier experience with counter-narcotics efforts. Nor have adequate resources been allocated (or even estimated) for proper implementation. Annual detailed Counter-Narcotics Implementation Plans, which provide more operational guidance in the short run and are organized around the pillars of the NDCS, lack the strategic linkages, synergies, response to diversity, prioritization, and sequencing that is necessary to make the counter-narcotics strategy work effectively.

Exper
rior to their ban, the Taliban treated opium de facto as a legal ommodity. Moreover, the Taliban ban applied to opium cultivation but not to trade in pium experience provides seful lessons for the future. Below are brief summaries of experience with each of the ain co rops in the field. In fact, where arp reductions in cultivation were achieved, physical eradication accounted for only a consequence, most of the limited physical eradication of poppy crops that has occurred Box 1: National Drug Control Strategy-Objective, Priorities, Pillars iew inating the basis for the trade. The NDCS, which has gone through several versions and is expected to be revised again in the fall of 2007, puts forward a credible multi-dimensional approach, which is briefly summarized below.

Overall Policy Goal:
To secure a sustainable decrease in cultivation, production, trafficking, and consumption of illicit drugs with a v to complete and sustainable elimination. National Priorities: elim I. Disrupting the drugs trade by targeting traffickers and their backers and nd diversifying legal rural livelihoods. II. Strengthening a III. Reducing the demand for illicit drugs and treatment of problem drug users. IV. Strengthening state institutions both at the centre and in the provinces.

ience with Different Counter-Narcotics Instruments
Until the Taliban's comprehensive and (in the short run) highly successful opium poppy cultivation ban imposed in 2000, counter-narcotics efforts in Afghanistan were very marginal, consisting primarily of some small-scale alternative development projects of a pilot nature. P c o and opiates.
Efforts to shrink the size of the opium economy in the post-Taliban era have been significant but fragmented and detached from the development agenda, and unevenly applied over time and across the country. Given also the entrenched nature of the opium economy, it is no surprise that counter-narcotics efforts have failed to prevent the large increases in opium production seen in recent years. However, this u m unter-narcotics instruments deployed so far in Afghanistan.

Eradication and enforced cultivation reductions.
This category includes reductions in opium poppy cultivation achieved through pressure, persuasion, and threat of eradication as well as from outright eradication of c sh very small proportion of the decrease in cultivated area.
The two main instruments for physical eradication of opium poppy fields have been the Central Poppy Eradication Force (CPEF) and so-called "Governor-led eradication" implemented by police and other forces at the provincial level. The CPEF has limited capacity, often faces local resistance, and, lacking local knowledge, has to rely on local guidance with respect to where to focus its eradication efforts. As a has been under the leadership of provincial Governors. There are serious concerns however that due to the close ties between many local officials and drug interests, overnor-led eradication is especially vulnerable to corruption in implementation.
on has been a ecessary condition to achieve reductions through pressure and persuasion. e Taliban ban, and in other rovinces at the time of the dramatic reduction in Nangarhar. are uite mobile in response to the changing geographical focus of counter-narcotics measur oreover, such corruption tends to result in eradication disproportionately ffecting the poor, who lack political connections or resources to pay bribes to avoid eradication.

G
The vast bulk of reductions in opium poppy cultivation have been achieved through pressure and persuasion, including passing down orders through the provincial and district administrations of the government as well as traditional village and higherlevel committees of elders (shuras). Religious arguments, building on the widespread popular perception that opium as a narcotic drug is "against Islam", have often played an important part in such campaigns. However, the credible threat of eradicati n Overall, eradication generally has not had a sustainable impact. Within 2-3 years after the nearly complete cessation of cultivation almost nationwide under the Taliban in 2000/01 and in Nangarhar Province in 2004, poppy cultivation in both cases rebounded. Moreover, even at the time of the respective bans, cultivation increased sharply in other areas-in Northern Alliance-controlled areas in the case of th p There are three main reasons for the limited success of eradication. First, eradication is technically difficult. The opium economy has amply demonstrated that it is "footloose" both across space and over time, with impressive reductions in opium poppy cultivation being offset by increases in other areas and/or in subsequent years. 5 As opium poppy is an annual crop, cultivated on well under 10% of Afghanistan's irrigated area, it can easily shift locations-opium traders, wage laborers, and even farmers q es.
The second reason is political resistance to and corruption in the implementation of eradication programs. The unpopular Taliban ban undermined political support for the Taliban in key Pashtun areas, possibly making it easier for the regime to be overthrown in late 2001. The Government's campaign against drugs in 2004/05, led by President Karzai, which achieved by far its greatest success in Nangarhar Province, also carried significant political costs for the Government, especially in that province. Adverse popular sentiment about eradication can lead to a political reaction, which in the Afghanistan context can be exploited by anti-government interests and by the drug industry itself. Corruption in implementation aggravates these political repercussions, undermining the credibility and perceived legitimacy of the government among the rural populace. 6 M a 5 It should also be noted that like other crops, opium poppy is best rotated from time to time to maintain soil quality and high yields. Thus a "stop-go" pattern whereby there are sharp reductions in a province or locality in one year followed by a rebound in subsequent years often makes agronomic sense. The widespread reports of an excellent opium harvest in Nangarhar Province, in the third year after the near-complete ban imposed in 2004, provide a striking example. 6 See Anderson (2007) for an interesting account of a particular case where political and/or corruption considerations obviously influenced and constrained eradication efforts with adverse consequences.
The third reason is that eradication does not address the deeper determinants of opium poppy cultivation. More sustained success with eradication and enforced cultivation reductions has occurred in localities or provinces that were relatively new to opium poppy cultivation (e.g. Wardak Province in 2004) or that were better-off in terms of access to resources, assets, and opportunities (e.g. central areas of Nangarhar and other provinces). Based on experience in Afghanistan, eradication and enforced reductions in cultivation are economically unsustainable, except in better-off localities where people already have viable alternative livelihoods-including access to water, land, and commodity and labor markets. When imposed on poorer areas and households lacking opportunities for viable licit economic activities, eradication can exacerbate poverty and increase the underlying dependence on opium. At the level of individual households, this comes about through reduced incomes, forced asset sales, in some areas opium-related debt (see Zia et al, 2005), and other coping mechanisms, which weaken coping capacities and resilience, making it more likely that farmers will subsequently return to opium poppy cultivation. For example, there are cases reported of farmers whose opium poppy fields have been eradicated several times but who nevertheless continue to cultivate opium poppy because this is the only way they can manage (and have any hope of ducing) their opium-related debts.
and undermining the credibility of the government nd others involved in eradication. s a hostile ct against it, further driving a wedge between the people and the government. re Overall, the experience with eradication indicates that although it attacks the most visible part of the opium economy and sometimes achieves visible and quantifiable results in the short run, these results are not sustainable. Moreover, eradication often has adverse consequences, perversely exacerbating the underlying dependence of many rural households on the opium economy a In this context, it should be emphasized that the shortcomings of eradication are multiplied when chemical spraying is involved-especially aerial spraying-as opposed to manual or mechanical eradication as employed hitherto in Afghanistan. Patterns of human settlement and inter-cropping mean that it would be very difficult to avoid some impact on people, livestock, and other crops from chemical spraying. Even if the actual health and other effects are not significantly harmful, chemical spraying would provide a propaganda victory for anti-government interests. In a context where infant and child mortality rates are extraordinarily high, where there are frequent crop failures, and where livestock suffer from numerous diseases, all such problems encountered for many years to come could be blamed on chemical spraying. The insurgency undoubtedly would take advantage of what would be widely perceived by the affected rural population a a Interdiction. 7 Interdiction efforts in Afghanistan were limited at first. From 2003 onward, however, there has been considerable emphasis on interdiction, including destruction of heroin processing facilities. Strong efforts have been made to build up police forces, train judges for special counter-narcotics courts, and set up prison facilities for drug traffickers. These activities are having some impact, although it has not been possible to go after the larger actors in the drug industry. However, corruption in the implementation of interdiction activities is a serious issue. Cases have been reported of drug traders being arrested but then released in return for a payment, and of their drug shipments being confiscated, not for destruction but for onward sale by corrupt local authorities, including the possibility of returning part of the shipment to the trader concerned for an additional payment (Pain, 2006b). Implemented in a corrupt manner, interdiction actions as well as eradication have been inadvertently contributing to consolidation of the drug industry around fewer, powerful, and politically connected actors (see Shaw, 2006). Clearly, enforcement activities have been used by some local and regional power-holders to favor their own (if they are directly involved in the drug industry) or allied drug industry interests.
Alternative livelihoods. These frequently debated programs (see Mansfield, 2007b for a review) aim to assist farmers to shift from opium poppy cultivation to alternative sources of income. The earliest efforts involved simple "crop substitution" projects, which subsequently gave way to "alternative development" approaches. Although somewhat broader, these still focused on substituting other crops for opium, concretely involving relatively small-scale, localized rural projects. Given the economic and social forces that lead to opium poppy cultivation as discussed earlier, such projects were grossly inadequate. Even in cases where they were successful in narrow terms, they tended merely to displace illicit drug cultivation elsewhere. The shift to an "alternative livelihoods" concept was meant to encompass these broader factors, including access to assets like land, water, and credit, as well as markets. But this conceptual improvement has not been translated into practice, as alternative livelihoods programs have continued to focus on discrete projects mainly involving other crops. 8 Many recent efforts to attack the development roots of opium poppy cultivation have not only been narrow, but also far too short-term in their orientation. They have been used to try to (partly) mitigate the immediate income declines suffered by rural households that exit from the opium economy. Key examples include cash-for-work programs and provision of agricultural inputs (typically seeds and fertilizer). Focused on short-run incomes and not encompassing markets, assets, and financing, these programs do not change the long-run and deeper conditions that contribute to households' decision to cultivate opium poppy. Indeed, abundant research since the 1990s has demonstrated that a short-run "quid pro quo" approach does not work (Mansfield, 2002), but nevertheless it continues to be widely used. Moreover, the approaches taken have been counter-productive, since promises have been made and popular perceptions consequently have arisen that these programs would deliver immediate results. The inevitable failure to meet such unrealistic expectations has discredited the government, donors, and the counter-narcotics strategy more generally.
Initiatives to articulate and implement a broader, more long-run development approach as part of a counter-narcotics strategy have continued through "mainstreaming" of the counter-narcotics dimension in development activities, national development programs in particular. It is hoped that in this way development programs can have an enhanced and scaled up counter-narcotics impact, while avoiding "doing harm" (i.e. inadvertently supporting expansion of the opium economy). For example, as defined by the World Bank (2006) in its mainstreaming guideline note for Afghanistan, mainstreaming involves factoring opium considerations into all aspects of the World Bank's engagement with Afghanistan, including analytical work and policy dialogue.

Demand side interventions in Afghanistan.
While high demand along with criminalization comprises a critical enabling factor for the illicit narcotics industry at the global level and there is large regional demand for opiates in countries near Afghanistan, domestic demand within Afghanistan is very small relative to the size of the opium economy which is predominantly an export-oriented activity. Nevertheless, there are clear signs that problem drug use is a significant and increasing problem in Afghanistan, spurred by chronic insecurity and conflict, as well as by return of refugees who became drug users in neighboring countries.
Demand-side issues have been neglected in Afghanistan, although recently more attention has been devoted to them, including due to concerns about HIV/AIDS transmitted by intravenous drug users. Resources devoted to demand reduction in Afghanistan have been minuscule as compared with those devoted to eradication and interdiction. While demand measures will not have a significant impact on the size and importance of the opium economy since it is overwhelmingly an export activity, they will be important in reducing the adverse impact on problem drug use in Afghanistan.

Communications and education.
This is another neglected area, except that where successes have been achieved in reducing opium poppy cultivation through pressure and persuasion, communications down the line within government, and between local administration and village shuras (groups of elders), have played an important role. There are widespread indications from fieldwork that communications by radio and via mosques have been effective in making it clear to the rural population that opium poppy cultivation is illegal and may be subject to eradication. However, communication efforts have failed to manage expectations about delivery of development assistance and on the contrary have tended to fuel such expectations.
Moreover, based on anecdotal evidence it appears that other communication efforts have at least sometimes gone far off the mark, reflecting lack of sensitivity to the local (including cultural) context, language issues, aspects related to the target population (e.g. most of the rural population is illiterate), etc. A striking example is discussed by Mansfield (2007c, p. 28), where being unable to read, people looking at a counternarcotics poster gave widely varying interpretations unrelated to the intended message. For example, no one saw an armed, turbaned young man in the poster as a terrorist or insurgent (which was the intended depiction), as people with such clothing and carrying a weapon would be a normal part of the local scene.

V. LESSONS AND POLICY IMPLICATIONS
As discussed earlier, eradication and enforced cultivation reductions, as well as hasty and fragmented alternative livelihoods projects, have been major elements of the counter-narcotics effort in Afghanistan in recent years. The country's rich experience provides ample evidence that these are problematic instruments that carry significant adverse side effects. The key lesson is that there is no substitute for effective rural development over the longer term in weaning rural populations away from dependence on opium poppy cultivation. This has major implications for the design, time horizon, time profile of funding, and sequencing in relation to eradication of development activities intended to contribute to counter-narcotics objectives.

"Better" Counter-Narcotics Instruments Are Not in Themselves Solutions
In addition to these hard-learned lessons about eradication and alternative livelihoods, experience in Afghanistan also has demonstrated that other counter-narcotics instruments, which appear far more attractive for valid reasons, are not in and of themselves solutions to the opium problem.

Emphasizing interdiction rather than eradication:
This approach is attractive on a number of grounds. First, the number of "targets" is several orders of magnitude smaller than the number of opium poppy cultivating farmers. If an equivalent impact can be achieved by interdiction actions against a relatively small number of drug traffickers as in a large eradication campaign, interdiction would be much more cost-effective from a technical perspective. Second, rather than criminalizing farmers this option can target the elements of the drug industry that constitute the major threat to Afghanistan's statebuilding, governance, and development agenda. Third, interdiction is likely to increase the "wedge" between farm-gate and downstream prices, potentially even reducing farmgate prices in the short run, thereby discouraging cultivation. 9 In contrast, eradication tends to increase farm-gate prices. Finally interdiction, if effectively implemented in an even-handed manner, can enhance the government's credibility by going after criminal elements rather than farmers and wage laborers.
Despite these attractive features, interdiction is far from a panacea. Significant figures involved in politics and government may be involved in or beneficiaries of the drug industry. A serious interdiction effort is therefore likely to give rise to political resistance by powerful actors. Institutional development and capacity building in the concerned law enforcement agencies is essential, which will take time. And resources, albeit of a much lower magnitude than required for the other instruments, are needed on a sustained basis. Moreover, as seen from the experience of other countries such as Iran, interdiction can elicit strong and effective responses by the drug industry, ranging from armed resistance against police to assassinations, bribery and corruption, and political manipulations. Thus although it can reap important benefits in the short run-not least by sending strong signals of the government's commitment against the drug industry, a counter-narcotics strategy driven by interdiction can trigger an increasing spiral of drugrelated violence. And finally, as in the case of eradication, implementation of interdiction measures in an uneven and corrupt manner not only harms the credibility and perceived legitimacy of the government but also can be a vehicle for consolidation and strengthening of the drug industry.

Interception of Precursor Chemicals:
While transforming opium into morphine is a straightforward technical process with fairly simple requirements, processing opium into heroin is more sophisticated and requires precursor chemicals, most notably substantial amounts of acetic anhydride. With most opium produced in Afghanistan currently being processed in-country (a sharp contrast with the situation in the 1990s and earlier), interdicting and disrupting the flow of precursor chemicals into Afghanistan is often advocated, and may be seen by some as a relatively straightforward solution. However, Afghanistan's porous borders and the inability of the international community to stem cross-border flows of arms, insurgents, and illicit drugs themselves (including in the past large quantities of raw opium) suggest grounds for caution. Moreover, acetic anhydride is widely used for other purposes in many countries. And finally, since precursor chemicals account for a very small proportion of the price of heroin at Afghanistan's borders, even successful efforts to disrupt their supply and sharp increases in their prices will not necessarily curb heroin processing. Recent fieldwork suggests that with the unprecedented large opium harvests in 2006 and 2007, prices of these chemicals have risen sharply, but their cost still comprises a small proportion of the price of heroin in Afghanistan and neighboring countries, and higher prices do not appear to be a significant obstacle to availability of the chemicals and to heroin processing.

Proposed Solutions That Have Not Yet Been Tried in Afghanistan
There are a number of other proposals for dealing with Afghanistan's opium problem that have not yet been tried in the country. Unfortunately none of these approaches are "silver bullets," either.
Licensing Production: Not yet tried in Afghanistan, although strongly advocated in some quarters, is the idea of licensing the country's opium production for sale and processing in the legal market for pharmaceuticals. This already occurs in some other countries, most notably Australia, France, India, and Turkey. However, only India produces licensed opium by labor-intensive techniques similar to those currently used in Afghanistan; the other countries grow poppy straw rich in pharmaceutical ingredients using capital-intensive modern agricultural techniques. There are clear international rules under which licensed production of opiates can occur, including monopsony purchase by the government and stringent controls to prevent leakages into the illicit market. Opportunities for engaging in licensed production are supposed to be open to "traditional" producers of opium, a status for which Afghanistan would appear to qualify but this would need to be confirmed.
Unfortunately, although superficially attractive, the proposal for licensed production of opium in Afghanistan founders on several basic practicalities.
• The security, rule of law, and governance situation in Afghanistan is nowhere near adequate for licensed and effectively controlled production of opium. India, with a much better security, rule of law, and governance situation and internationally accepted control mechanisms in place, suffers from substantial leakages of opium from the licensed into the illicit market, estimated at around 30% (see Mansfield, 2001). In Afghanistan leakages could only be expected to be much larger. • More specifically, and given that only a small proportion (under 10%) of Afghanistan's good agricultural land is devoted to opium poppy cultivation, if the current level of production is licensed, the same amount of illicit production could spring up within a couple of years (i.e. total opium output very easily could double in a fairly short period of time). • The licensed price inevitably would be far lower than the price of illicit opium, reinforcing incentives for leakages and parallel production for the illicit market.
Even if these problems could be resolved, which does not appear feasible in the foreseeable future, there are equally daunting obstacles from the international side: • Stocks of licensed opium produced by India with labor-intensive techniques are building up. Moreover, Afghanistan and indeed India do not appear to have a comparative advantage in licit cultivation (see figures cited in Mansfield, 2007b) Thus the scope for large licensed production by Afghanistan using current techniques would appear to be limited. • More generally, although some have argued that there is a worldwide shortage of opiates for licit purposes like pain management in developing countries, this is subject to debate. Any shortage would not appear to be at all near the magnitude that could accommodate Afghanistan's recent or current production of illicit opium. Thus for licensed production of opium in Afghanistan to be accomodated from a demand perspective, the existing licensed producing countries would need to sharply reduce their output. But there is no sign of any willingness on their part to do so.

Buying Up the Opium Crop:
A somewhat similar proposal is that the international community, rather than putting large amounts of money into counternarcotics measures of doubtful effectiveness, should simply buy up the opium crop for one or two years as an interim solution. In addition to avoiding problems associated with eradication and other enforcement measures against the opium economy, this measure is seen as temporarily disrupting the drug trade and sharply reducing funding available for criminal and anti-state interests, as well as buying some time for development of sustainable alternative livelihoods to wean farmers away from poppy. Thus it would need to be accompanied by a crash program to develop viable legal livelihoods in opium poppy cultivating areas.
Although in some respects attractive, this proposal is problematic because, as discussed earlier, strengthening and diversifying legal livelihoods takes a long period of time, so buying the crop for only one or a few years will not do much good. Moreover, it would set in motion incentives that, as in the case of licensed production, would stimulate further growth of opium poppy cultivation. If it involves a commitment to buy all of the opium that is produced, costs would likely mount from year to year (or otherwise increasing amounts of opium would go back into illicit channels). And there are a host of technical issues like pricing which would not be straightforward to resolve, and would affect production incentives and potential for leakages. Finally, the public sector apparatus, presence, capacity, and level of governance required for this approach to be feasible administratively are not in place. Thus this proposal would not have lasting benefits. Moreover, by setting a precedent of the government and international community getting engaged in the purchase of opium, it would send very mixed signals to farmers as well as generating problematic incentives.

Blanket Agricultural Subsidies / Support Prices:
Another "silver bullet" which has not been tried in Afghanistan but is sometimes proposed is a subsidy / price support scheme for crops planted instead of opium. This option is seen as stimulating the development of licit crops to substitute for opium, and also as providing income support for former poppy farmers during a transitional period until their new cropping patterns get firmly established on a financially viable basis.
This proposal also raises a number of very serious issues. As has been amply demonstrated by Afghanistan's experience opium is footloose, so such a subsidy or support price system would need to cover the entire country in order to be effective. This would add greatly to its cost, and moreover would involve putting in place a national subsidy or price support system that Afghanistan would be unable to come anywhere near to affording on its own in the foreseeable future. And most of the poor in Afghanistan, as in many other developing countries, are net purchasers of grain, so depending on how it is designed this might well not be pro-poor in its impact. Moreover, based on international experience, agricultural subsidies and support price systems tend to be selfperpetuating, and they become politically difficult to terminate once they are put in place. Given the shortage of domestic revenue in Afghanistan and other development priorities, how long could the international community credibly commit to providing blanket subsidies or financing price supports?
Another important question is what crops would be subsidized and how they fit into Afghanistan's development agenda. Wheat is by far the most important agricultural product and tends to be the crop of choice for farmers who stop cultivating opium poppy, but wheat is not the crop of Afghanistan's future. As a low-value, relatively land-and water-intensive, and not labor-intensive crop, wheat does not mesh well with Afghanistan's resource endowment. Moreover, in good years Afghanistan already comes close to meeting domestic demand for foodgrains from its own production, and there are no export prospects for wheat in the region given that several neighboring countries are themselves sizable producers and exporters (including to Afghanistan).
What Afghanistan needs to develop instead in the agricultural sector for non-opium rural development is high-value, labor-intensive licit crops with good export potential (including through agro-processing).
Finally, Afghanistan's porous borders, and the same kinds of governance and security issues that affect some of the other proposals discussed earlier, would make a blanket subsidy or price support scheme virtually impossible to implement. The large production of wheat in neighboring countries, most notably Pakistan (where there is a subsidy on wheat flour) and Kazakhstan, would add further complications and risks. Costs would likely increase, and trade patterns would be further distorted.

Toward a "Smart Strategy" against Drugs in Afghanistan
The lessons from experience with different counter-narcotics instruments in Afghanistan and elsewhere, and the discussion above of other proposed solutions to the opium problem, clearly demonstrate that there are no easy answers. On the one hand, high global demand for illicit opiates shows no sigh of diminishing. On the other hand, the conditions that lead farmers to cultivate opium are deep-seated and not possible to change rapidly, or inexpensively. Moreover, many of the commonly proposed "solutions" could actually worsen the situation. However, while calling for modest expectations in the short run and for exercise of caution, this is not a recipe for inaction. As emphasized earlier, the opium economy is simply too important and too harmful to Afghanistan to neglect or downplay. Even though the illegality of the opium economy and high global demand for illicit opiates are likely to persist for the foreseeable future, experience does provide some grounds for hope that strategies exist that can reduce opium cultivation in Afghanistan.
Experience in Afghanistan and elsewhere suggests the following principles to guide the development of a "smart strategy" against drugs in Afghanistan (see Byrd and Buddenberg, 2006). One is to focus efforts against those parts of the drug industry that pose the greatest danger to the nation and its development agenda-i.e. the larger drug traffickers and their sponsors, who threaten to undermine state building through political corruption and compromising of state agencies like the Police and Ministry of Interior. These actors, not the farm households engaged in opium poppy cultivation and/or wage labor in the opium economy, constitute the real threat.
A second principle is to take fully into account the adverse side effects and distortions induced by counter-narcotics instruments, which could undermine or even negate any beneficial effects. A prime example is corruption in the implementation of counter-narcotics policies. There is no point in designing a plan that may work well but only if there is no corruption, when it is obvious beforehand (and based on experience) that there will be corruption in implementation.
A third principle is to minimize perverse incentives. If, for example, counternarcotics assistance (especially support for alternative livelihoods activities) is concentrated in the major opium-producing areas, other (non-cultivating) areas are likely to resent this and may even weaken their efforts to keep out the opium economy. 10 More generally, experience suggests that counter-narcotics efforts should not focus excessively on the major current opium producing areas but rather consider strong measures to discourage the spread of the opium economy to non-dependent provinces and localities.
A fourth principle is to respond to the diversity in the rural opium economy, based on the expanding body of knowledge that is being built up, and to exploit opportunities based on local resources, accessibility of markets, and improvements in security and governance. In areas where people have some assets (including not least in the form of able-bodied labor), there are land and water resources, commodity and labor markets are accessible, and there is a modicum of security and half-decent governance, it is possible to get away from dependence on the opium economy. Moreover, when these ingredients are in place, the shift from opium to sustainable licit livelihoods can occur within a few years. Where such essential resources, assets, and opportunities are not present in a locality, there is no alternative to longer-term rural development which inevitably will take much time to achieve.
Finally, it is essential to continue to monitor and build information and knowledge about the evolving opium economy, to engage in sound and careful policy analysis, and to utilize the findings of research and the lessons from experience in the design and implementation of counter-narcotics measures.
Based on these principles, a smart strategy against drugs in Afghanistan could include the following key elements.
With respect to eradication and enforced reductions in poppy cultivation, these should be focused on the better-off and new opium producing areas. Experience in such areas (for example with eradication in Wardak Province, a new opium producing area) suggests that these measures can be successful and have a sustained impact. When a locality is targeted for eradication on such a basis (the province is usually too big a unit for this purpose, but the district or sub-district level would be more appropriate), the target should be complete or near-complete rather than partial elimination of opium poppy cultivation in that locality-this will minimize the risk of corruption in implementation and associated distortions. Chemical spraying, especially aerial spraying, should be avoided for reasons explained earlier in this paper.
With respect to interdiction, and recognizing the political difficulties involved, these efforts should be focused against medium and larger drug traffickers and their sponsors. In addition to causing disruption of the drug trade in the short run, this will set a very positive example for counter-narcotics efforts at lower levels. Given the weakness of the judiciary and the difficulties in successfully prosecuting major drug figures (although some progress has been made in operationalizing special counter-narcotics courts and in training personnel), actors associated with the drug industry should at least be removed from their positions in government, which can have a significant impact in the Afghan context. Aside from the direct benefits, this will also send a strong signal of credibility for the counter-narcotics strategy.
With respect to alternative livelihoods, do not throw money at short-term alternative livelihoods programs but rather support sensible rural development, fully understanding that this will take time. Resources for effective rural development will need to be scaled up as sound programs are developed and refined through field experience. Lessons from international experience with rural development should be brought to bear in this regard. Part of the rural development effort could involve support for promising high-value horticultural products, such as almonds, raisins, pistachios, and others, as well as livestock products. Many of these goods would need to be exported, and innovative approaches to developing exports using the international and Afghan private sector could be explored for this purpose.
Mainstreaming of the counter-narcotics dimension in development programs is very important and will help in scaling up meaningful efforts on the development side. Mainstreaming should not be approached mechanically but rather in a flexible and results-oriented manner. And while moving forward expeditiously with necessary technical and support work to make mainstreaming a reality, expectations about progress in the short run should be kept modest.
As emphasized earlier, don't concentrate efforts and resources too much in the main opium producing areas, but rather consider a kind of "containment" strategy to progressively close off increasingly large parts of the country from vulnerability to dependence on opium. Over time, this will help narrow the geographical scope and range of the drug industry and restrict its response options to counter-narcotics measures.
Finally, and more generally, just as the counter-narcotics dimension needs to be mainstreamed in development programs, there is also a need to fully mainstream the development dimension in counter-narcotics strategy and actions.
Given the importance of the opium economy in Afghanistan, this kind of "reverse mainstreaming" of development and governance considerations in decisions, policies, and instruments for fighting against drugs is essential. In particular, it could help avoid problems like those encountered in the recent past when counter-narcotics measures have been designed and implemented in isolation from the broader development and state-building agenda.