Version classiqueVersion mobile

Société civile, associations et pouvoir local au Yémen

Sarah Ben Nefissa
Maggy Grabundzija
Jean Lambert

Chapitre 3 : ONG et pouvoir local

Financial Mechanisms of Decentralization and the Taxation of Qat

Peer Gatter

Entrées d'index

Texte intégral

1With the enactment of Law No. 4 of 2000 “concerning the Local Authority”, Yemen has launched one of the most ambitious decentralization programs in the Arab World. The law that initiates a comprehensive reform of the Yemeni governance system through the establishment of three levels of government (state, governorate, and district) is seen as a major step forward in nation building and democratic development.

2Local governance in Yemen is based on three main principles: the expansion of popular participation through elected local councils, administrative and financial decentralization, and the decentralization of service delivery.

3The extreme lack of financial resources at the local level is seen as one of the principal challenges in this endeavour. An efficient taxation of the production and consumption of qat (Catha edulis), a psychoactive stimulant widely consumed in Yemen, may help overcome this deficiency and could strongly contribute to financially more autonomous local representations. Ultimately this may foster the democratisation of the country. Already today qat revenue represents a large share in the budgets of a number of governorates.

  • 1 The Republican Decree for Law no. 52 for the year 1991 has sought to regulate local administration (...)
  • 2 GATTER & AL., 2002a, p. 43.

4Despite decentralization being one of the promises of Yemen’s unification in 1990, and despite being described in Yemen’s 1994 constitution as one of the pillars of the state instrumental for local development (articles 143-146), it took the country’s leadership a decade to cash in this pledge1. The unification of the Yemen Arab Republic (North Yemen) with the People’s Democratic Republic of Yemen (South Yemen) called for the merging of two very different political and economic systems. Strong tribalism and a precarious security situation in the north hampered the ability of the unified country to organize its administrative presence in many governorates and districts. Socialism in the south delivered to the new state a large and superfluous bureaucracy. Ultimately, Yemen’s unification resulted in an unbalanced and redundant administration with a high concentration of officials in the capital city and the major provincial centers. The economic aftermath of the Gulf War in 1990 with over 800,000 returnees and Yemen’s civil war of 1994 further aggravated an already difficult situation and delayed the government’s reform agenda. In particular this resulted in the need for the public sector to employ workers returning from the Gulf countries, the loss of purchasing power by civil servants whose salaries were eroded by rampant inflation, and the ultimate transformation of the civil service into an inefficient organization supporting a large number of poorly paid, little motivated and often unqualified staff2. After having addressed the challenge of stabilizing the economy with assistance from the IMF and the World Bank, the Government committed itself in the second part of the 1990s, to address Yemen’s most critical structural problems, in particular the modernization of the civil service and the restructuring of the country’s governance system.

Local Governance before the Reform

5While in socialist south Yemen elected local councils existed since the 1970s, in the north citizens were represented through an elected parliament only at state level. At governorate and district levels the government met public service needs through an elaborate system of de-concentrated field offices of ministries and agencies. Consequently, central control over most regions of Yemen’s north remained indirect and nominal. Despite attempts by successive Yemeni governments to increase their authority, local administration in many tribal areas has remained near unchanged since the times of Ottoman hegemony. Shaykhs continue to exercise their military and legal powers by maintaining tribal armies and private prisons, and government control over such areas is exercised only by military campaigns and by purchasing the loyalty of Shaykhs and their tribes. The arrangement of de-concentrated state powers was until the enactment of the Law of Local Authority (LAL) rather geared towards keeping up law & order than to providing services to citizens. It is thus not surprising that most governors and district directors have a military or police background.

  • 3 WORLD BANK, 2006, p. 40.

6The LAL also provided mechanisms for formalizing traditional, democratic organs such as the popularly elected Local Development Associations (LDA) that had evolved during the 1970s throughout northern Yemen due to the need of self-reliance of rural communities. LDAs collected funds and in-kind support from residents, migrant labourers, and foreign donors to build schools, health centers, water structures and roads with often a much greater reach than the state. When the government however centralized the administration of the zakāt tax in the 1980s (see below), the LDAs lost their most important funding source and their role in rural development rapidly diminished3. Building on their experience and on the strong tradition of local autonomy that the LDA movement had fostered, may help today to mitigate the concentration of power among a small number of tribal leaders.

The Administrative Organization after the Reform

7After unification in 1990, Yemen was made up of 18 governorates and some 280 districts. With the Republican Decree No. 23 of 1998 on the reorganization of administrative units the Government initiated a restructuring exercise carving out the Governorates of ‘Amrān and al‑Dāle’ and subdividing larger districts inherited from the PDRY into two or more administrative units. During 2001 and 2002 a number of municipalities were split into smaller constituencies and in 2004 the Governorate of Rayma was formed. Today, Yemen is made up of 21 governorates (including Sanaa Capital Secretariat) and 333 districts.

8Each administrative division has its own local authority, which consists of the centrally appointed governor or district director, the elected local council (LC), and the executive organs (branch offices of ministries and other government agencies). The new local authority system was designed to operate on the basis of people’s participation in decision-making, in the administration of local affairs, and in social, economic, and cultural development. Popular participation is assured through the democratic establishment of political representations that are closer to citizens and their needs, the powers of local councils to propose budgets, development programs and investment plans, and the accountability of executive organs to the local councils who supervise their performance.

9In some functional areas the elected councils may even be considered as local legislative bodies in their constituencies. In this regard, there are some similarities with the national legislative body, but while the parliament is to a certain extent independent and detached from the executive, the local councils along with the executive organs are considered an integral part of the local authority and are chaired by the centrally directed unit heads. Governors are appointed by the President, who is not held accountable for his choice. District directors are assigned by the Ministry of Local Administration (MOLA) and the directors and staff of executive organs are deployed by their respective ministries or agencies. Despite this major democratic deficiency the new law has for the first time ever clearly defined the functions and duties of governors and district directors. Among these is besides the administrative and executive functions the fostering of local development.

10The executive organs at governorate and district levels play new roles as facilitators in the decentralization of service delivery. They are now regarded as local organs and act as the administrative, technical and executive organs of the local councils. But even though operating under the councils’ supervision and control, they are not employed or paid locally. Yemen’s system of decentralization is thus rather a mixture of devolution and deconcentration.

Impact of the Reform on Central Government

11The central executive organs of the state remain unchanged, but are confined by law to the drawing of general policies, the enactment of regulating by-laws, oversight, and training. Based on the request of the local councils, they are to implement projects that local authorities on their own are unable to implement as well as projects of a national magnitude.

12MOLA is the government entity entrusted with the implementation of the decentralization process. The Ministry of Finance (MOF) reviews district and governorate budgets, controls local revenue collection, and allocates resources to the local level. The Ministry of Planning and International Cooperation (MOPIC) is expected to reduce its role in dealing with small projects and shall strengthen its governorate-level branches to provide support to LCs in preparing local development plans and ensure that these plans are congruent with national objectives. The LAL assigned new and greater responsibilities to MOLA and called for the development of new and improved coordination mechanisms among MOLA, MOF, MOPD, and the Ministry of Civil Service (MOCS). At present the coordination between line ministries at central level is weak, and overlapping portfolios are a source of conflicts.

  • 4 Author’s interview, Nov. 6, 2000 at MOLA.
  • 5 GATTER & AL., 2002a, Annex I, p. 4.

13Sādeq Amīn Abū Rā’s, Minister of Local Administration, pointed out in November 2000 that MOF was “reluctant to participate in the reform process and not ready to accept the new situation with taxes and fees remaining at local level, since it fears a loss of influence4”. Tensions over LC budgets increased over the following months and led in June 2001 to the announcement of resignation of the Minister of Local Administration as he felt that the MOF was obstructing his work and the progress of decentralization. The crisis was solved by mediation of the Prime Minister and after receiving funds from MOF, MOLA was able to launch the first training plan for LC members and Ministry officials in July 2001. MOF also allocated funds for all governorates and LCs were consequently able to draft their first budgets for 20025.

14There is still a need for revision of a large number of laws and regulations in order to align them with the legal framework for decentralization since inconsistencies have led also to a duplication of functions between central and local organs. In 2002 the Prime Minister has formed a commission for the harmonization of laws which has drawn up a list of 70 laws and regulations that need amendments. To strengthen local government and devolve some of their functions, ministries are also expected to re-deploy more qualified support staff from Sanaa to their branches. Over the past years little has been done in this respect, since few civil servants are ready to settle in the countryside with much less amenities.

The First Local Council Elections


15Following an intense public debate and broad media coverage on the country’s political and technical readiness for decentralization, President ‘Alī ‘Abd Allah Sāleh and the government of Prime Minister ‘Abd al‑Karīm al‑Iryānī decided in 2001 to proceed to local council elections in combination with an already planned referendum on constitutional reforms. While local council elections were seen as an important step towards democracy, the constitutional amendments were interpreted by Yemeni intellectuals and foreign observers alike as undermining democratic development. They further strengthened the President’s grip on power, as the President’s mandate was extended from five to seven years, making the next election only due in 2006 and extending President Saleh’s possible legal term of office until 2013. Observers believe that this will give him time to prepare his son, Ahmad, for office. The President was also given extended powers to dissolve Parliament without national referendum. The term of the Parliament was extended from four to six years, preventing the Yemen Socialist Party from entering the House of Representatives for a further two years (it had boycotted the 1997 elections). This also ensured that the religious-tribal Islah Party, with its growing support base, could until the next elections in 2003 not extend its own presence in Parliament. The Consultative Council (majlis al‑shūra) was almost doubled in size (now 111 members, all of whom are appointed by the President) and was given the power to jointly vote with Parliament on legislative matters6

  • 7 GATTER & AL., 2002a, Annex I, p. 2.
  • 8 CENTRAL STATISTICAL ORGANIZATION, Statistical Yearbook 2004, p. 350.
  • 9 WORLD BANK, 2006, p. 40.

16Referendum and elections took place on February 20, 2001. Though marred by some localized unrest and violence, they could be carried out satisfactorily and resulted in the elections of council members in all governorates and an overwhelming approval of the proposed constitutional reforms (73 % of the votes). The ruling party secured an absolute majority with 69.1 % of the vote in governorate local councils and with 60.7 % at district level. Islah secured 19.5 and 23.1 % respectively, with a higher dominance in the north. The Socialists were almost exclusively restricted to the south and won 4 % of the vote at governorate level and 3.5 % in district councils7. Teachers accounted for nearly 40 % of those elected to district councils and Shaykhs and civil servants each accounted for about 7 % of office holders. Of a total of 7,315 elected council members only 36 were women8. The overwhelming number of teachers among the ranks of council members suggests that the electorate voted on the basis of perceived qualifications, rather than social status9. However among the centrally appointed district directors, Shaykhs or their kin were over represented.

17Initially, the Government intended to hold its second local elections jointly with the general elections in 2003 after an experimental mandate of only two years. This was thought to allow government to assess the functioning of local councils and to develop appropriate measures to ensure that they will operate effectively over a full-term mandate. However local elections were postponed twice and councilors stayed in office until 2006.

The new Local Administrations – Tasks and Challenges

  • 10 MINISTRY OF LOCAL ADMINISTRATION, 2000, article 7, 13 & 17.

18The law has transferred responsibility for the provision of most public services to the new local administrations. According to the Executive Bylaws (EB) of the LAL, local councils are to direct, supervise and control the work of the executive organs in their constituencies. They are, among others, to propose development plans and evaluate the level of their implementation, encourage investment projects, attract tourism, and develop and protect water resources. The responsibilities and powers the district is assigned include the founding, equipping, management and maintenance of schools, health and agricultural extension centers, electricity networks, local water and sewerage projects, bridges and secondary roads, as well as public parks, slaughter houses, markets, and cemeteries, cleaning projects, and historical heritage sites. Under the supervision, management and control of local councils, the executive organs are to implement the general policies, plans and programs, and provide the citizens with municipal, health, educational, and agricultural services. They shall also provide electricity, water and sanitation services and maintain and build rural roads10.

  • 11 GATTER & AL., 2002a, p. 28-29.
  • 12 REPUBLIC OF YEMEN, 2005, p. 69.

19To assume these responsibilities at district level, Yemen and its local administrations still have to come a long way. Foremost, the country has to overcome an uneven and inequitable distribution of staffing. While the civil service is generally oversized at central level, the government is practically absent from other parts of the country where only teachers, some health staff and security forces are present. Inadequate or missing office buildings, shortages in equipment and supplies as well as low wages, are the major culprits for poor staff performance. Staff absenteeism, low education levels, inadequate women representation and a work culture that does not encourage initiatives, accountability or compliance with existing regulations also play their part. Another prominent contributor to low performance is chewing of the stimulant qat, which is practiced by 85.8 % of civil servants who chew at least three times per week and spend half or more of their official income on the leaves11. Training efforts implemented by MOLA have generally focused on explaining the Local Authority Law, rather than forming officials to its implementation. Many civil servants thus have a limited understanding of their tasks and functions in a decentralized system. A national training plan drafted in 2003 targeting some 16,600 senior officials and staff of local authorities, could not be implemented so far due to a lack of funding12.

Funding Sources of Local Administrations

  • 13 See also LOCAL GOVERNMENT OF DENMARK (LGDK), 2005, p. 37.

20Governorate and district local administrations are financed through recurrent and capital budgets, accounting for 16.8 % of the Government expenditure in 2004. The recurrent budget which is the administrative and operational budget for executive offices and local councils (salaries, office supplies, etc.) made up 98 % of these expenditures, reducing local authority funding primarily to salary transfers. These are determined and financed directly through central allocations and are beyond the control of local bodies (LAL, article 125)13.

21The meagre capital or development budget is to fund local capital expenditures, and consists of locally generated revenue and central transfers (LAL, article 123):

  • 14 This percentage share and those given in the following for other budgetary resources of local autho (...)

22District Local Revenues accounted for 19.8 % of the budget of local authorities in 200414. These are collected at the district level in favour of the district and consist of 27 different sources of revenue, in particular of different penalties, fees and taxes, endowments from individuals, assistance from donors through governmental agencies, and most importantly the religious zakāt duty (50 % of the collected zakāt revenues remain in the district, the rest is transferred to governorate level).

23Shared Governorate Revenues accounted for 34.1 % of local budgets in 2004. They are also collected at district level, but transferred on a monthly basis to the governorate finance office where these funds are quarterly divided for the benefit of all local authorities of the governorate. 25 % of the revenues return to the district where they had been collected, 50 % are equally distributed to all other districts of the governorate, and 25 % remain at governorate level (LAL, article 124). Shared governorate revenues consist of 28 revenue sources, including fees, penalties, grants, wills and donations (e.g. fees on well digging and passport issuance, fines for traffic violations, and fees on fishing and weapons permits). The by far most important ones are the “qat consumption tax” and zakāt (50 % of district zakāt collection comes in this pool).

24Shared General (Central) Revenues made up 20.8 % in the 2004 local budgets and are collected centrally by MOF on behalf of MOLA and distributed to local councils, cooperatives and local development funds. This revenue source consists of fees imposed on air and sea travel tickets, on petrol, diesel and gas sales, as well as of 30 % of the annual revenues of the Road Maintenance Fund, the Agricultural and Fisheries Production Promotion Fund, and the Youth Welfare and Sports Fund (LAL, article 123/3). Resources from these funds must be spent on purposes related to their origin.

25Central Subsidies feeding into the budgets of local administrations contributed 25.3 % to governorate and district resources in 2004. They are general‑purpose transfers to finance economic and social development plans and projects at the local level.

Resource Availability

  • 15 TAX AUTHORITY, 2005; MINISTRY OF LOCAL ADMINISTRATION, 2005; and LGDK, 2005, p. 38.

26Despite 55 different types of fees and taxes being collected at the local level, the revenue situation in most governorate and district administrations apart from some wealthier urban ones is precarious. District Local Revenues and Shared Governorate Revenue together accounted for just above one percent of total state revenue (YR 8.2 billion or US $ 45 million) in 200415, a rather inadequate figure for financing a decentralized system and for successfully conducting local affairs. Apart from zakāt and qat consumption tax (see below), many of the fees collectable at the local level bring such small yields that little efforts are taken to actually collect them (e.g. fees on cinema tickets or on digging seep holes). Since the level of taxation was set by cabinet resolution, local administrations have no liberty in adjusting taxes and fees to inflation rates. In 1990 the tax for opening a commercial place such as a general store was YR 480 (US $ 40), but due to the 1995 Riyal devaluation and rampant inflation this was equivalent to only $ 2.45 in 2007. In a number of governorates such as al‑Jawf and Mareb the collection of district and government revenues is further hampered by tribal power and frequent feuds.

  • 16 DLDSP, 2005, p. 28.

27Despite having experienced a sharp increase over recent years (from YR 452 million in 2002 to over 3 billion in 2004 and 2005), central transfers of Shared General Revenue transfers have fallen well short of plans16. Because shared general revenue are so limited it has become a practice in some governorates that these funds are not distributed equally among all districts, but pooled for one district in one year and for another in the following year.

  • 17 DLDSP, 2005, p. 28.

28Central Subsidies accounted for YR 3.85 billion during 2002–2004 and have reached YR 4 billion in 200517. Despite this small rise, central contributions have been declining in real terms. The total amount of these transfers is set in accordance with budgetary priorities as part of the annual state budget decision. Since this is an ad hoc exercise at central level, in times of declining state revenue local authorities are likely to be the first to suffer considerable budgetary cuts.


29The share of central contributions to local budgets varies greatly across Yemen. While in Sanaa Capital Secretariat and in al‑Hodeida Governorate with their large urban agglomerations central transfers account for only 20-25 % of local budgets in 2004, in governorates with largely rural character and weak infrastructure such as Abyan, Mareb, al‑Mahra and al‑Jawf, they contribute 80-96 % to local finances (see fig. 4)18.

  • 19 Due to the lack of statistical information the measurement of these parameters had certain limitati (...)

30The criteria upon which Shared General Revenue and Central Subsidies are distributed are laid out in the LAL (article 124). In 2004 an attempt was made by MOLA, MOF and MOPIC to translate these criteria into numeric indicators and the 2005 Central Subsidy (YR 4 billion, equivalent to ca. US $ 27 million) was distributed accordingly. 45 % were allotted upon considerations of poverty rates and the level of economic and social growth, 25 % for the abundance or scarcity of local revenues, 20 % for population density, and 10 % for the effectiveness in collecting and spending revenue19.

  • 20 LGDK, 2005, p. 38 & 42.

31The overall share of the local authority sector (recurrent and capital financing) in public budgets has been shrinking from 17.3 % in 2002 to 16.8 % in 2004. Observers find this deeply worrying since for an effective decentralization of Yemen’s governance system a steady increase in resources at local level is seen as indispensable. This decrease could indicate that tasks and responsibilities of local authorities are either not implemented or at a lower level than previously. It could also point towards a trend of re-centralization with less funding going through the budgets of governorate and districts administrations and being instead administered and spent directly by the line ministries20.

Budgetary Powers of Local Authorities

  • 21 LGDK, 2005, p. 37.

32Yemen’s present decentralized system assumes independent and competent local authorities with full responsibility for budget preparation and execution. To assist them in maintaining effective fiscal controls, the Ministry of Finance has established accounting units in all Governorates and a still limited number of districts21. These finance offices intensively monitor district and governorate budgeting and spending and ultimately control local revenue collection as well as resource allocation to the elected LCs.

  • 22 REPUBLIC OF YEMEN, 2000, articles 129-142; and MINISTRY OF LOCAL ADMINISTRATION, 2000, articles 7-2 (...)

33According to the LAL and its Financial Bylaw, each administrative unit at governorate and district level independently prepares its annual plan and budget which is subject to approval by its LC and the Governorate Plan and Budget Committee. After presentation before the High Budget Committee at central level, it is included in the state budget and submitted to Cabinet for ratification22. For the preparation of the plan the executive offices at district and governorate level have to define their needs for projects and submit them to the elected local council for discussion and approval. Local councils can allocate 20 % of the total annual district local and shared governorate revenues towards locally defined needs to promote their own development initiatives. Accordingly, an administrative unit can in principle “undertake all financial acts without reference to the central authorities so long as these are not inconsistent with the fundamentals of budget implementation” (LAL, article 137). These liberties concern however only locally generated revenues, since by law, central resources are not subject to the supervision of local councils and remain under the exclusive control of local MOF offices and their subordinate yielding authorities (LAL, article 127).

  • 23 LGDK, 2005, p. 16.

34The reality is somewhat different. Observers have described the real level of self-determination and self-governance of local authorities as very limited23. While the decentralized administrations at governorate level are working well, in most districts no significant devolution of responsibilities and resources had followed the enactment of the law due to the extreme lack of financial resources. Most local councils up to now thus neither plan development in their jurisdictions, nor do they budget their resources. This is especially true for those local authorities that cannot raise sufficient own local funding and heavily rely on central transfers. In rural areas where there are few commercial establishments, little agriculture, and especially no qat cultivation, to generate taxes, the level of local self-determination is extremely weak. This is true for most of the eastern and southern desert areas, and the lower western escarpment where qat cannot be grown.

Qat and its Role in the Budgets of Local Administrations

  • 24 Since enactment of the 10th decree regulating narcotic drug use (Betäubungsmittelrechts-Änderungsve (...)

35Qat (Catha edulis Forsk.) is a psychoactive stimulant that is grown in many of the highland areas of Eastern Africa, ranging from the southern Sudan to Madagascar and the Transvaal. It is also grown across the Red Sea in Yemen’s western highlands and the ‘Asīr mountains of Saudi Arabia. While being considered a drug in most Arab states (e.g. Saudi Arabia, Egypt, Sudan, Palestine), as well as in many European countries including Germany24, there is no legislation in Yemen controlling its cultivation and sale.

  • 26 Personal communications, Dr. Ali No’man, Director of the Qat Research Unit, Ministry of Agriculture
  • 27 GATTER, 2006, p. 36.
  • 28 LENAERS & GATTER, 2000, p. 13.
  • 29 Unpublished 2005 Qat market survey by the author.
  • 30 GATTER, 2006, p. 18.

36The hardy tree that is famed by farmers for its drought resistance, is grown on 12 % of Yemen’s agricultural area and its cultivation covered 121,000 hectares in 200325, though some Yemeni researchers believe that it may well be double as much26. Qat accounts for 10 % of the country’s GDP, for a third of agricultural GDP, and for 9.4 % of household expenditures27. In poor families qat related spending even reaches 28 % of the family budget28. The qat sector provides employment for one in every seven working Yemeni. In the capital Sanaa alone some 13,000 persons are involved in the sale of the stimulant29. On average 72 % of men and 33 % of women above the age of 12 chew the bitter leaves of the qat plant. Some 42 % of male consumers chew five to seven days per week and display compulsive habits30. As the predominant cash crop, the income qat generates plays a vital role in rural economies and prevents people in many of Yemen’s highland areas from drifting into the cities in order to seek work. But qat also depletes scarce water resources, contributes to soil degradation, and has crowded out production of essential food crops and agricultural exports. For its producers and consumers alike, qat is seen as one of the main health hazards in Yemen, mainly due to the unregulated use of pesticides. Given the economic importance of qat, it is not surprising that taxes stemming from the production and sale of the plant are substantial. The qat sector contributes to government revenue in four ways, by a religious tithe (zakāt) levied on qat production, a Public Cleaning Tax for keeping qat markets tidy, a Qat Consumption Tax, and a Youth & Sports Tax, levied on qat sales. While zakāt is imposed as a direct tax and collected at farm level by zakāt assessors, the other taxes are levied as indirect taxes at checkpoints on the roads leading into the cities and in qat markets.

  • 31 e.g. FOOD AND AGRICULTURAL ORGANIZATION, 2002, p. 107-108; and KENNEDY, 1987, p. 157, who makes no (...)

37Despite the importance of qat as a revenue source for governorate and district administrations, little is known about the taxation of this commodity. This is partly because the economics of qat have not received the attention deserved, partly because the government is highly secretive about revenues from the plant. As a consequence the literature on the subject matter is inadequate and beset with errors31.

A. Zakāt Revenue

38Next to faith, prayer, fasting and the pilgrimage, zakāt is one of the five pillars of Islam. It is a religious alms tax for purifying wealth and is enforced through moral obligation and fiscal measures. The sharī ‘a and Yemen’s budget distinguish between various types of Zakāt, such as the zakāt al‑fitr (for feeding the poor at end of the fasting month of Ramadān), the zakāt al‑bāten (due on wealth exceeding a certain threshold), the zakāt al‑mawāshī (on bee hives, honey, fish, and livestock), and the zakāt al‑‘ashūr (imposed on grain, vegetables and other crops such as qat, if their yield exceeds five wasūq ca. 140 kg).

  • 32 GATTER & AL. 2002b, p. 122; and REPUBLIC OF YEMEN, 1999, article 13.

39The sharī‘a sets the zakāt al‑‘ashūr at 10 % of net profits of rainfed farming. If additional means of irrigation are employed the zakāt due is reduced to 5 %, since water structures such as reservoirs, irrigation channels, dug wells, and in more recent times tube wells or diesel powered pumps raise production cost32.

  • 34 CSO, diverse Statistical Yearbooks: 1976/77, p. 180-184, 1979/80, p. 221-225, 1985, p. 198-200/ 198 (...)

40In the absence of a modern taxation system, zakāt was until the 1960s responsible for up to 84 % of Yemen’s public revenues33. When the Imamate made way for a republic, the revolutionaries were thus eager to enact the right of the state to oversee the administration of zakāt in the constitution of the Yemen Arab Republic in 1970 (article 135). But with the creation of a modern system of taxation and significant revenues from customs duties during the boom years of the 1970s and 80s, the relative importance of zakāt in government revenue dramatically decreased. During the 1970s zakāt still accounted on average for 5.1 % of government revenue, during the 1980s this had dropped to 2.6 %, and reached an absolute low during the 1990s with just 1.0 % not last due to increasing revenue from the emerging petroleum sector34. Zakāt was therefore long neglected and the government has only recently “rediscovered” this tax and started to recognize zakāt’s potential for rural development and poverty alleviation.

41In 2005, zakāt revenue amounted to YR 6 billion ($ 33 million). 38.2 % of total zakāt revenue was levied on wealth generated by private companies and individuals, 26.4 % came from the agricultural sector, 16.9 % was levied on profits generated by the public sector (e.g. water and electricity supply), 11.7 % was levied as zakāt al‑fitr after Ramadān, and 6.8 % came from other zakāt sources. In 2005, qat was responsible for over half of the government’s total zakāt proceeds from agriculture. Over recent years, qat’s share in zakāt revenue has been steadily increasing with YR 617 million (US $ 3.5 million) qat zakāt accounted for 13.6 % of the total zakāt budget in 2003 and reached YR 938 million (US $ 4.8 million) or 15.8 % in 2005.

42There are marked geographical differences in zakāt revenue from qat, which are easily explained since zakāt is levied on profits from its cultivation. Since qat is a highland crop adapted to moderate temperatures, which cannot be grown below 800 meters of altitude, it is neither found in the governorates of the coastal plain, nor in those of the hot and arid eastern desert. The governorates profiting most from zakāt generated by qat are Hajja, where qat accounts for 66.9 % of total zakāt revenue, al‑Dāle‘ (52,7 %), Sanaa Governorate (50,1 %), ‘Amrān (46,1 %), al‑Baydā’ (43,2 %), Sa‘da (38,7 %), and al‑Mahwīt (30,6 %) (see fig. 1A and 2).

  • 35 YEMEN TIMES, 1998.
  • 36 The Qur’ān states that 1/9 of the zakāt collection is to be allotted in favor of people who adminis (...)

43The administration of zakāt in Yemen remains archaic and a quarter of the collected revenue goes to tribal leaders and zakāt assessors to honour their role in overseeing its collection35. This is in clear violation of sharī ‘a provisions36. A new draft zakāt law currently being discussed in Parliament is attempting to address this deficiency and is to replace the current law (No. 2 of 1999), inconsistent with the LAL.

  • 37 DLDSP 2005, p. 11-12.

44In 2005, MOF and MOLA drafted a refined LAL which intends to spread qat production revenue more equally across the country. Zakāt was thus excluded from the funding sources of local authorities and in return for compensatory central subsidies is to become once again a centrally administered revenue37. So far this draft law remains to be discussed and approved by Parliament. If approved, however, the law would pave the way for a return to an inefficient system of central zakāt collection and would once again increase central government control over local affairs.

B. Consumption Tax

  • 38 Muhammad Mahmūd al‑Zubayrī, Revolutionary and Poet, born 1919, assassinated in 1965, called qat in (...)

45The taxation of qat sales has a long history in Yemen. The Ottoman treasury during the Turkish occupation of northern Yemen in the 19th and early 20th century as well as the British Colonial Administration in Aden, extracted handsome revenues from its trade. Under the Yemeni Imams who fought the Ottomans in the north, and the socialist innovators, heirs to the British in the south, this did not change much and qat remained one of the most important commodities for taxation. In order to discourage chewing after the republican revolution in the north, the tax rate for qat was raised from 10 to 30 % of its sales value in 1969. The revolutionaries initially associated qat consumption with the decadent Imamic regime which was reflected also in many poems of the time38. Once in power, many of the former revolutionaries however soon succumbed to the chewing habit and today even most ministries in Sanaa have a special room (mafraj) for the unique purpose of chewing. Many ministerial afternoon work sessions are thus held within a qat setting. As a consequence Law No. 14 of 1980 once again lowered the qat tax to 10 %.

46Upon the recommendations of an 1986 IMF mission that reviewed the Yemeni tax system, Law No. 70 of 1991, on the “Taxation of Production, Consumption and Services, and its Amendments” reformed the mechanisms of taxation, and its 1992 bylaws on tax collection and tax estimation were meant to make qat taxation more efficient. The Tax Authority however failed to significantly enlarge the tax base, and the government took the decision to raise the level of taxation to 20 % in 1996 (Law No. 14 of 1996). This led on the one hand to a substantial increase of qat tax revenue from YR 659 million in 1994 to YR 1.6 billion in 1997, but on the other hand resulted in widespread tax evasion despite provisions of the law for high fines. While qat tax revenues steadily increased during the 1990s, in real terms the government’s tax income dropped due to the devaluation of the Riyal in 1995. Qat taxes were equivalent to US $ 21 million in 1994, but to only $ 13 million in 1997.

47With the General Sales Tax Law No. 42 of 2005, the “consumption tax” of qat was converted into a “sales tax” and confirmed at 20 % of the sales value, without however changing the procedures governing its collection (article 64 and schedule 3).

48The revenues generated by the qat “consumption” or “sales” tax are even more important for the decentralization process than those generated by qat zakāt. They constitute the biggest single funding source for local authorities. At a national average qat was responsible for 62 % of locally collected tax (dharā‘eb) revenue in 2003 and for 58.5 % in 2004. The only other local taxes that generate yields worth mentioning are the real estate tax and the free profession tax accounting for 25.9 and 12.9 % of total tax revenue respectively in 2004.


49In only two of Yemen’s 21 governorates the share of qat in local tax revenue was below 40 % during 2003 and 2004. In many governorates qat even accounted for more than three quarters of the tax income, such as in the northern governorates of Sa‘da (76.5 %), al‑Hodeida (79.3 %), and Hajja (81.3 %), and even more so in parts of the former socialist south, with 81.6 % in Aden, 86.9 % in Lahej, 88.6 % in Shabwa, and as much as 89.8 % in Abyan (see fig. 1B)39.

50The fact that the share of qat taxes in the south is even higher than in the north, is yet more surprising, since the chewing habit is quite new to many areas of southern and eastern Yemen. Apart from Aden (Lahej and al‑Dāle‘), where consumption had been restricted to weekends by law, it had been completely banned in the other governorates of the PDRY between 1969 and 1990. This well illustrates the triumphal march of qat though unified Yemen during the past decade and a half.

51In 2005, revenue from qat consumption tax amounted to YR 2 billion ($ 11 million). Revenue from this sector is especially high in governorates with large urban centers such as al‑Hodeida (YR 423 million), Aden (341 million), Taez (305 million), and Sanaa City (303 million) (see fig. 3). The roads leading to these cities, particularly in the coastal areas, are easily controlled by checkpoints, and qat traders are taxed when passing though. Aden for example has only one access road linking it to the qat production areas in the north, and only one coastal and one desert road connect Hadhramawt Governorate with the qat producing areas of the western highlands.

52While qat consumption in urban, coastal, and remote desert areas is easily controlled and to some point efficiently taxed, this is not the case for the rural highland zone. Chewers in ‘Amrān, al‑Mahwīt, Rayma and Sanaa Governorates do not consume less qat than consumers in neighbouring Hajja, al‑Hodeida, Ibb or Sanaa City, where tax incomes are much higher, but consumption in rural areas is simply much more difficult to tax, since much of the qat produced there is consumed locally and does not cross checkpoints. Qat markets there are far too small and too many in number to be efficiently taxed by the government. Governorate figures on Qat Consumption Tax do thus say little about the spread of the qat consumption habit. Much rather they are an indicator for the efficiency or inefficiency of the Yemeni qat tax system.


53Dissatisfaction with the high availability of qat consumption tax in some Governorates and its absence in others has been voiced by local councillors and governors alike. For example, qat produced in al‑Dāle‘ and al‑Baydā’ is transported through Lahej and consumed and taxed in Aden Governorate. While qat consumption tax revenues were very low in the production areas with only YR 13.7 million in al‑Dāle‘ and 27.2 million in al‑Baydā’ in 2005, in Aden as much as YR 341.1 million were levied. In Lahej Governorate through which the qat vehicles pass daily to reach Aden, only 53.5 million were collected (see fig. 3)40.

C. Youth and Sports Tax on Qat Consumption

  • 41 DLDSP 2005, p. 28.

54In 1999, the Republican Decree No. 10 of 1996 concerning the Establishment of the Fund for Youth Welfare and Sports was amended to include qat as a funding source. Besides a 5 % tax on qat sales, the law also foresees a 5 Riyal tax on each packet of cigarettes and on each sack of cement sold (Law No. 36 of 1999, article 11/4-6). Qat’s share of the Youth and Sports Tax is collected together with qat consumption tax at checkpoints and in qat markets. As laid out in the Local Authority Law, the fund’s revenue can be allotted only to the purpose for which it was established (LAL, 123/3a). In 2001, Youth and Sports tax contributed YR 370 million (US $ 2.2 million) to the state budget and increased to 511 million ($ 2.8 million) by 2005. While 70 % of these funds are administered at by central government, 30 % were transferred to the governorates and equally distributed among the local authorities of these constituencies. In 2004, the Youth and Sports tax contributed with 20.1 % to Shared Central Revenue in local authority budgets and accounted for 3.2 % of total local budgetary resources41.

D. Public Cleaning Tax on Qat Sales

  • 42 MINISTRY OF FINANCE, 2001, p. 32 and CRANE, 2000, p. 3.

55A tax payable by urban companies and shops for city cleaning came in to effect in 1999 (Law No. 20 on the Establishment of a Fund for City Cleaning and Improvement). The fund is jointly managed by MOLA and the Ministry of Construction, Housing & Urban Planning. The tax is considered a local revenue and is collected at district level in the favour of the district (LAL, article 123/10). Local branches of the MOF administer the fund and in accordance with the law draft budgets for the use of its resources. An amount of YR 287 million was projected for collection in 2000, the first operational year the fund, and fell with a revenue of YR 261 million (US $ 1.6 million) somewhat short of expectations42. The cleaning tax is especially applied in qat markets and is paid in Yemen’s north by market owners who pass this expense on to the tenants of shops or sales stalls in the form of rent. In the south where market spaces are for the most part state owned, each qat merchant has to pay a small fee to participate in street cleaning. In Aden for example the tax is levied in form of a general fine for “public health infractions” which amounted in 2005 to YR 100/day (ca. $ 0.54) per qat merchant.

Qat Revenue and its Role in Democratization

56Yemen’s decentralization has significantly enlarged the circle of decision-makers throughout the country and has given local communities the ability to participate in decisions affecting their lives, namely in the field of health care, education, water supply and sanitation. The initial enthusiasm accompanying the decentralization process has however died down. Elected local council members (especially at district level) soon came to realize that extremely narrow boundaries were set for their scope for action. This was mainly due to the limited yield of locally collected fiscal resources, the inadequacy and unreliability of central transfers, as well as the extreme lack of qualified local administrative staff capable to monitor fiscal revenues and public expenditures, and trained in budgeting and financial management.

  • 43 GATTER & AL., 2002a, p. 37.

57Indeed, only few districts are currently able to exercise local self-governance, undertake considerable capital financing and to independently operate and maintain a broad range of services. The resources of most district authorities so far consist of the civil servants at local level and their salaries. An example is the Hadrami desert district of Thamūd with a population of 25,000. Apart from a handful of shops in the small and lethargic district capital there are few taxable economic activities. The local economy consists mostly of nomadic goat and camel herding for subsistence purposes. In 2000, the percentage of locally generated revenue was at just 1.3 % (YR 170,000 or US $ 1,060)43. Areas such as Thamūd will always remain heavily dependant on central funding and under the present system their local decision makers will also in the future have very few own decisions to make.

58The poor revenue and staffing situation is likely to prevail for years ahead. Since it is unlikely that central funding for local authorities will significantly increase over the coming years, and that existing training plans will be implemented, local administrations are pretty much on their own. The future success of Yemen’s decentralization process and its potential to foster democratisation thus depend on the financial resources that can be mobilized locally.

59At this time, the only way to boost local resources is by increasing tax revenue. Since local councils cannot raise the level of taxation unilaterally (see above), their only option is to make tax collection more efficient. Local authorities have a comparative advantage in tax collection over the central government, since locals are more willing to pay taxes if they know that these funds will not disappear in dark government channels, but will be used towards local development projects.

60The most potential for increasing local fiscal revenue is believed to be the boosting of zakāt and qat consumption tax revenue at present the most important contributors to local budgets. The role of zakāt in poverty alleviation and rural development is tremendous, but its full potential remains yet to be realized. The decentralized collection of qat consumption tax enables local administrations to levy taxes also in rural areas that until now have escaped taxation (see above). Local authorities are now able to tax qat sales in villages and monitor qat transport on secondary roads at comparatively low administrative cost and effort. This will however not go without resistance from qat farmers, sellers and consumers.

  • 44 Here only a comparison between governorates of a more rural character is made, excluding governorat (...)

61Governorates with more taxable economic activities (e.g. qat cultivation or sales) are less dependant on uncertain and fluctuating central transfers. There is a tendency that Governorates with intensive qat growing such as al‑Baydā’, Dhamār, Ibb, and Hajja, receive and need noticeably less central funding than Governorates that grow only little or no qat such as Mareb, al‑Jawf, Abyan and al‑Mahra (see fig. 4)44.

62The more important the share of own source revenues in local budgets is, the less dependant local decision makers become on central “spoon-feeding” and the more independent they become in allotting resources to projects they believe important. A slow democratisation of Yemen’s governance system is an expected result. Since the qat sector is today and is likely to remain for the coming years the most important source of local funding, one can say without hesitation, that qat revenue is the financial motor of the decentralization process and thus plays a decisive part in the democratisation of Yemen. Since the use of the qat leaf is notorious in local council meetings in many areas of Yemen (qat leaves are believed to stimulate ideas, generate understanding, help in the mediation of conflicts and in the sealing of agreements) one could even say that qat is also a social and organizational motor of decentralization and democratization.


63The financial means of local administrations are still present too insignificant to permit for an effective at both governorate and district levels.

64While governorate administrations supported by de-concentrated ministerial branch offices have assumed their responsibilities under the regime of decentralization in the provincial capitals, in most districts the inadequacy of financial transfers, lack of office facilities, trained and motivated personnel have so far not allowed for any effective local self-governance. Most district authorities – apart from some urban districts – thus remain heavily dependant on the support and guidance from governorate and central authorities.

65In governorate administrations, the blessings of qat taxation are already perceptible today. Qat the single most important taxed commodity at the local level fosters local decision making, makes governorate administrations less dependant on central financing and enables local councils to fund small-scale, but own development projects. In a third of Yemen’s governorates, qat was in 2003 and 2004 responsible for over 75 % of tax income. Despite some recognizable regional differences in tax composition, only in two of Yemen’s 21 governorates, qat’s share in tax revenue was below 40 %.

66As an agricultural product and taxable commodity, qat cultivation and trade are thus of outmost importance to Yemen’s rural economy and for the decentralization process.

  • 45 GATTER, 2006, p. 7-41.

67At the same time it should not be forgotten that qat as a consumer item is also at the root of a number of negative developments in Yemen. As a 2006 sample survey documented, qat consumption fosters corruption, divides families, contributes to poverty and malnutrition and thus impairs the possibility for development of children. Given this, it is not surprising that 61.2 % of Yemeni women and 53.2 % of men expressed the opinion that the government should actively limit the chewing of qat45.

68The government is well aware of the detrimental effects of the plant, but nevertheless has little moral scruples in using qat revenue to its own ends – not unlike its western partners that profit from the high taxation on cigarette and alcohol consumption.

69One of the future challenges for Yemen’s government will be finding a rational balance between the positive and negative effects of qat cultivation and its consumption. Making qat tax collection more efficient or even raising the tax level, under the present governance system, will have a very positive impact on local autonomy. It may in the long run also contribute to a certain decrease in the chewing of the bitter leaves.

Fig.1: Qat Revenue in the Budgets of Governorate and District Local Authorities within each Governorate

Fig.1: Qat Revenue in the Budgets of Governorate and District Local Authorities within each Governorate

Fig. 2: Zakāt Revenue for Qat Production by Governorate in 2005

Fig. 2: Zakāt Revenue for Qat Production by Governorate in 2005

Fig. 3: “Qat Consumption Tax” Revenue by Governorate in 2005

Fig. 3: “Qat Consumption Tax” Revenue by Governorate in 2005

Fig. 4: Number of Qat Trees per Governorate (in Million) and Shares of Central Transfers in the Budgets of Local Administrations (in Percent)

Fig. 4: Number of Qat Trees per Governorate (in Million) and Shares of Central Transfers in the Budgets of Local Administrations (in Percent)


2005: CSO/EU Agricultural Census of 2003.

1976/77, 1979/80, 1985, 1988, 1991, 1995, 1998, 1999, 2002, 2004.

A Study to Prepare Urban Development and Management Strategies for the City of Sana’a, Yemen. For the Ministry of Construction, Housing and Urban Planning and the World Bank.

Yemen Study of Policy Options for Finances and Financial Management Systems of Local Authorities, Draft Report August, 2005.

2001: Yemen Country Profile.

2002: National Conference of Qat. Towards the Formulation of a comprehensive Qat Policy in the Republic of Yemen. Technical Field Study.

Socio-Economic Assessment of Qat Chewing in Yemen. A Survey of Seven Governorates, Report for the World Bank.

GATTER, P. et al.
2002a: Assessment of Administrative Decentralization in Yemen. Prospects and Challenges. A Survey of four Governorates, Report for the World Bank, February 2002.
2002b: “
al‑Qat wa al‑Eqtesād al‑Kullī”, in: Gatter & al. (Eds.): First National Conference on Qat. Under the Auspices of H.E. the President of Yemen, Mr. ‘Alī ‘Abd Allah Sāleh. Proceedings and Conference Discussion Materials.

The Flower of Paradise. The Institutionalized Use of the Drug Qat in North Yemen. D. Reidel Publishing Company.

Qat and its Impacts on Nutritional Balance and Family Life in Yemen: With a Special Reference to Low Income Segments of Society. Report for the World Bank.

Yemen Study of Policy Options for the System of Sub-National Governance. International Consultancy Division Final Report, 15 July 2005.

2001: Financial Accounts for 2000.

2000: Executive Bylaws Republican Decree Nr. 269 of 2000.
Financial Accounts for 2004.

2003: Initiative 20/20. Allocating Resources to Basic Social Services in Yemen, Sanaa, April 2003.

1999: Law No. 2 of 1999 on Zakāt.
Law No. 4 of 2000 concerning Local Authority.
2005: Poverty Reduction Strategy. Progress Report for 2003 & 2004. (June 2005).

1993 « Les effets négatifs sur le développement économique et social du Yémen », Cahiers du GREMAMO, 11, 1001-102.

2005: General Tax Statistics for 2004.

2006: Governorate Qat Consumption Tax Revenue for 2005.

2005: Financial Governorate Accounts for 2004.

1998: « Zakāt – where does the money go? », Yemen Times, July 12, 1998.

2006: Republic of Yemen, Country Social Analysis. Report No. 34008-YE, Jan. 11, 2006.


1 The Republican Decree for Law no. 52 for the year 1991 has sought to regulate local administration in Yemen, but political struggles and separatist movements stood in the way of putting the provisions of the law into implementation (see also Yemen Human Development Report 2000/2001, p. 78).

2 GATTER & AL., 2002a, p. 43.

3 WORLD BANK, 2006, p. 40.

4 Author’s interview, Nov. 6, 2000 at MOLA.

5 GATTER & AL., 2002a, Annex I, p. 4.


7 GATTER & AL., 2002a, Annex I, p. 2.

8 CENTRAL STATISTICAL ORGANIZATION, Statistical Yearbook 2004, p. 350.

9 WORLD BANK, 2006, p. 40.

10 MINISTRY OF LOCAL ADMINISTRATION, 2000, article 7, 13 & 17.

11 GATTER & AL., 2002a, p. 28-29.

12 REPUBLIC OF YEMEN, 2005, p. 69.

13 See also LOCAL GOVERNMENT OF DENMARK (LGDK), 2005, p. 37.

14 This percentage share and those given in the following for other budgetary resources of local authorities have been calculated based on Tax Authority and MOLA financial accounts for 2004, and on statistics for central subsidies provided by the MOLA/UNDP supported Decentralization and Local Development Support Program (DLDSP). See also DLDSP, 2005, p. 29 (table 3).


16 DLDSP, 2005, p. 28.

17 DLDSP, 2005, p. 28.


19 Due to the lack of statistical information the measurement of these parameters had certain limitations (DLDSP 2005, p. 12).

20 LGDK, 2005, p. 38 & 42.

21 LGDK, 2005, p. 37.

22 REPUBLIC OF YEMEN, 2000, articles 129-142; and MINISTRY OF LOCAL ADMINISTRATION, 2000, articles 7-20.

23 LGDK, 2005, p. 16.

24 Since enactment of the 10th decree regulating narcotic drug use (Betäubungsmittelrechts-Änderungsverordnung (BtMÄndV, article 1e) on Jan. 20, 1998.


26 Personal communications, Dr. Ali No’man, Director of the Qat Research Unit, Ministry of Agriculture.

27 GATTER, 2006, p. 36.

28 LENAERS & GATTER, 2000, p. 13.

29 Unpublished 2005 Qat market survey by the author.

30 GATTER, 2006, p. 18.

31 e.g. FOOD AND AGRICULTURAL ORGANIZATION, 2002, p. 107-108; and KENNEDY, 1987, p. 157, who makes no distinction between zakāt and Qat consumption tax; SHEMSULDIN 1993; La dépêche internationale des drogues, 65.

32 GATTER & AL. 2002b, p. 122; and REPUBLIC OF YEMEN, 1999, article 13.


34 CSO, diverse Statistical Yearbooks: 1976/77, p. 180-184, 1979/80, p. 221-225, 1985, p. 198-200/ 1988, p. 297-299, 1991, p. 191-192, 1995, p. 308, 1998, p. 314-315, 1999, p. 342, and 2002, p. 379.

35 YEMEN TIMES, 1998.

36 The Qur’ān states that 1/9 of the zakāt collection is to be allotted in favor of people who administer this tax (Qur’ān 9: 60).

37 DLDSP 2005, p. 11-12.

38 Muhammad Mahmūd al‑Zubayrī, Revolutionary and Poet, born 1919, assassinated in 1965, called qat in his writings “the devil…. seducing the Yemeni man” (1958).



41 DLDSP 2005, p. 28.

42 MINISTRY OF FINANCE, 2001, p. 32 and CRANE, 2000, p. 3.

43 GATTER & AL., 2002a, p. 37.

44 Here only a comparison between governorates of a more rural character is made, excluding governorates with large urban centres. Areas such as the capital Sanaa, al‑Hodeïda, Aden, Taez and al‑Mukalla with considerable income from real estate tax were deliberately excluded. Lahej could not be included due to a statistical error in the CSO/EU Agricultural Census database.

45 GATTER, 2006, p. 7-41.

Table des illustrations

Titre Fig.1: Qat Revenue in the Budgets of Governorate and District Local Authorities within each Governorate
Fichier image/jpeg, 216k
Titre Fig. 2: Zakāt Revenue for Qat Production by Governorate in 2005
Fichier image/jpeg, 56k
Titre Fig. 3: “Qat Consumption Tax” Revenue by Governorate in 2005
Fichier image/jpeg, 56k
Titre Fig. 4: Number of Qat Trees per Governorate (in Million) and Shares of Central Transfers in the Budgets of Local Administrations (in Percent)
Fichier image/jpeg, 181k

© Centre français de recherche de la péninsule Arabique, 2008

Conditions d’utilisation :


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search