Revenue Mobilisation Performance of Union Parishad in Bangladesh : Is It Convergence of Central-Local Relations ?

Despite several local revenue sources, Union Parishads (Councils) in Bangladesh are yet to perform optimal local revenue mobilisation. This paper evaluates Unions revenue trend and performance, budget and planning practices, effect of central grants on local revenue in the backdrop of central government recent initiative for Unions capacity building. Analysis includes three representative Unions, in a comparative perspective, and uses secondary and primary data (from Parishad functionaries, local citizen, government officials and national experts). Findings show that open budget discussion, discretionary and performance grants have positive impact on local revenue collection while this study raises question about sustainability of the revenue augmentation due to disparity of deconcentrated allocations system and feeble local democratic governance. Study recommends making adjustments in local revenue shares, increase discretionary grants, and validation of local participatory governance. This study has implications for local revenue mobilisation through convergence of central-local policy and strategy, specially for developing countries having similar local socio-economic and revenue source footings.


Introduction
Recent trend shows that policy makers and donors are emphasising on rural local government's fiscal capacity from local economic and rural development fronts to integrate development programmes through citizen participation, incorporate gender issue and disadvantaged groups, aid in poverty reduction, encourage local government's autonomy and innovation in exploring local potentials.In line with this trend, developing countries for the last two decades have been devolving finance functions and responsibilities to the local government and emphasising on the need for prudent mechanism in mobilising local revenues.Existing literature also supports that recent fiscal decentralisation move without some fiscal responsibility on the part of the local governments for local revenue mobilisation will lose much of their meaning (World Bank, 2009).In fact, fiscal capacity of rural local governments is related to fiscal decentralisation that commands prudent central-local relationships.Central fiscal and non-fiscal supports and local collective efforts, central polity and local strategy, central political choice and local innovation and freedom can accelerate local financial resources mobilisation.
According to Iben Nathan (in Blair, ed., 1989, p.50), "local resources mobilisation is the efforts of local administration to raise finances from a locality in order to fund activities within that area".The size and freedom of the local government are considered crucial aspects that determine the structure and generation of revenue and adequacy of local finance.In theory, it is contended that a sizeable local government in terms of human settlements and productive land area have better prospects of more revenue generation (Hye, 1985, p.180).But rural local governments' revenue generation in developing countries is always a great challenge and the mechanisms by which local revenue can be mobilised have raised some pertinent issues, like revenue raising ability, effect on economic efficiency, equity implications and administrative feasibility (Bahl, 1984).
Historically, local governments and local administrative units in South and South-East Asian countries were created for central administrative convenience.However, over the course of time the paradigm of rural local government in this region has got gradual change with regard to political and economic dimension.Democratic elected body is also taking place in many rural areas in offering realistic possibilities for citizen participation and mobilisation of local resources for local development.Bangladesh is not an exception.But resources mobilisaiton in rural areas in Bangladesh is still challenging task due to lack of real commitment from both central and local government.With the donors supported programme and in the changed economic context, government of Bangladesh has changed its traditional thinking with the philosophy that the issue of local government financing should take important place in central policy and local strategic move.In this respect, central government has taken various initiatives in raising Union Parishad (UP) revenue by means of project assistance and incentive grants mechanisms.
Recently, government of Bangladesh, to strengthen UPs fiscal role, has recognized UP as a primary economic and a part of administrative unit of rural local governments.Ministry of Local Government, Cooperative and Rural Development (LGRD) has introduced performance grants in 2003 and discretionary Basic Block Grants (BBG) in 2004 for each UP which are provided with flat rate basis.Besides BBG, normal Annual Development Programme (ADP) block allocations are also being provided to UPs.ADP allocation is appropriated as 90% for general and 10% for operational performance and general allocation is to appropriate based on population 50%, physical area 30% and backwardness 20%.(Note 1) ADP and own funded projects of Upazila and Zila parishad are also allocate to UPs, based on population, area and backwardness.(Note 2) Recently, government has replaced Local Government (Union Parishad) Ordinance, 1983 by the Local Government (Union Parishad) Act, 2009 with the aim of curtailment of external political interventions and providing more administrative and fiscal authority including citizens' participation and right to know information about UP affairs.To build UPs capacity, central government introduced Local Governance Support Project (LGSP) with a cost of US$ 218.12 million (2006US$ 218.12 million ( to 2011). ).LGSP is initiated from the experience and learning of Sirajgonj Local Government Development Project (SLGDP) (GOB, 2003).
Traditional revenue administration system along with the inability to explore the resources potential, UPs are continuing as chronically resource poor institutions and their reliance on central grants is still about 57% (GOB, 2007).Central government has declared policy to enhance local government expenditure but it is still below 3% of GDP while in the national level, the share of tax and non-tax revenue in the GDP has increased from 7. 1% in 19901% in /91 to 11.2% in 20071% in /08 (GOB, 2009)).The current share of local government and rural development in the total national public expenditure is only 7.2% (GOB, 2008).On the other hand, being a century old democratic body, UP is yet to be decentralised rural local government while this institution is provided with mandatory and discretionary functions including own revenue raising powers.Despite UP's long history, frequent changes in the composition of this elected body and its use for political mobilisation have also prevented it from pursuing long term development efforts in the rural areas.Even in the era of national democratic governance during last two decades, and thus, central-local relation remains weak in many fronts including ineffective policy and strategy on local resources mobilisation.
At present, there are Three rural local government tiers in Bangladesh, known as Zila Parishad (ZP) at district level, Upazila Parishad (UZP) at sub-district level and UP at the union of villages level.The field level administrative structure is comprised of six division, 64 districts and 483 Upazila.The total number of UPs are 4498 with an average population of 27,000 and average land area of 30 sq. km. in each UP (UNDP, 2002).The structure, functions and finance of rural local government institutions have undergone many changes from the ancient period to British colonial period and Pakistan regime to the present day.During the British rule , Bengal Local Self-Government Act, 1885 laid the formal legal foundation of Union for a group of villages (Shafi, 2001, p.3).During the regime of Pakistan (1959Pakistan ( -1970) ) under 'Basic Democracy' system Union councils were provided both revenue and development functions as the lowest tier rural local government (Rahman and Khan, 1997, p.8). Immediate after Independence in 1971, system of basic democracy was abolished and in the article 9, 59 and 60 of Bangladesh constitution articulated about the local government formation, functions and revenue powers where union council remain as the lowest tier rural local government.
Local government (Union Parishad) Ordinance, 1983Ordinance, (amended up to 1997) ) provided with powers to UP in levying tax and fees including women representation at UP.Each UP is now comprised with nine wards, one chairman, and nine members from nine wards including three elected women members representing each three wards.UP secretary is the only official staff of UP and chairman is the chief executive.Each UP has 13 different standing committees to perform its 38 different types of functions in which special participations are allowed from different occupational groups like, freedom fighters, cooperatives society, disadvantaged group, landless, destitute women, farmers, fishermen or weavers etc, as non-voting representatives.UPs were allowed to continue with 28 revenue sources until the creation of Upazial Parishad (UZP) in 1982.After the creation of UZP, fragmented revenue structure provided with UP only six major sources.For expenditure side, Upazila Development Coordination Committee (UDCC) is designated as the final authority for the finalisation of UPs project selection and resources allocation.UDCC is consisted of Upazila chairman and UPs chairmen of Upazila jurisdiction.
However, there has not been done any specific empirical study on UPs revenue mobilisation subsequent to the changes being experienced in UPs recent local revenue activities and central governments grants mechanism including renewed policies.This study is considered crucial, both for central and local perspectives.Therefore, the objectives of the study are to examine UPs own revenue mobilisation trend and performance, to evaluate UPs budgeting and planning practices, to assess impact of central grants on UPs revenue mobilisation and finally to provide policy implications based on findings.

Selection of UP cases and Data approach
Based on the record of Ministry of Local Government, Cooperative and Rural Development, three UPs are purposively selected from three different regions, districts and Upazila (sub-district) of the country.An average annual revenue performance criterion is used to classify three cases, based on five years (2002/03 to 2006/07) actual revenue collection while compared and validated it with published revenue performance data of other UPs.(Note 3) Hence, three UP cases are defined as High Revenue (More than Tk.1000 000) Performing (HRP), Moderate Revenue (not less than Tk.500 000) Performing (MRP) and Low Revenue (low Tk.250 000 to nearly Tk 500 000) Performing (LRP) UPs.This broad categorization is made to show the comparison on UPs own fiscal capacity and revenue efforts including revenue performance of three cases (Patton, 1990;Yin, 1989;Miles and Huberman, 1984).Others selection criteria used for sampled UPs include local socio-economic activities, good documentation of revenue records, human settlements, natural disaster, NGO's activities, distance from town or district and Upazila headquarters, and existing resources base with potentials of local resources generation.The study is based on field works (2007/08), and UP cases are analysed in a comparative perspective to find specific and common findings.Among selected cases, SLGDP assisted program was carried (2002)(2003)(2004)(2005) in MRP UP for capacity building while in HRP and LRP UP, there were no such projects being carried, though LGSP has been implementing in all three UP cases from 2006.However, in general, the sampled UP cases are, more or less, said to be a representative character of other UPs.Selection procedures, profile and key statistics of three UPs is shown in Table 1.
Table 1 can be here Both secondary and primary data are used in the study.Primary data obtained from total 227 respondents consisting UP functionaries (18%) as key informants, local people (62%) (comprising Households, Businessmen, Local leaders/elites, NGO workers and members of Marginal groups), relevant Upazila, District and Ministry officials (18%) and national experts (2%).To obtain data, researchers interacted with district and Upazila officials, UP members as well as with purposive sampling of local citizenry of various strata.Networks were developed with key informants through a constant process of iterative interviewing and cross checking to build an understanding of the dynamics of local resources mobilization.The reconnaissance survey, observation, questionnaire survey, interview, FGD including written information where it required, are used in data collection.Data analysis reviews available literature, reference of previous study, published report and relevant regulation/policy documents.Quantitative data on local revenue trend, performance and efforts are explained through illustrative tables and descriptive statistics.Pearson's correlation coefficient is applied to examine the UPs tax revenue collection relationships with conditional and unconditional grants, tax assessment, and development expenditure.Qualitative data are used to complement the quantitative analysis.

Resources mobilisation trend and performance
It is observed and understood from UP functionaries and local people that recent policy of central government have induced UP representatives and local people in bringing about positive change in UPs tax collection.UP functionaries informed that the increase in assessment and collection of local taxes, onward 2003/04, is due to introduction of performance and discretionary grants (BBG) and the new system of rewarding UP chairman for operational performance.In addition, UP representatives have motivated local people to pay fees and taxes with the commitment of implementation of more local development projects.This commitment is given by UP as a local counterpart of the national assisted projects.During study it is found from the calculation of UPs revenue estimation and collection records that in all three UPs, onward 2003/04, estimation and collection of fees and tax revenue have increased in size though there are some erratic fluctuations in the annual growth and collection efficiency.For LRP UP, collection efficiency is erratic and only in 2005/06 it shows highest (76%) efficiency.The average collection efficiency is 126%, 83% and 61% for HRP, MRP and LRP UP respectively.Functionaries of LRP UP noted that due to occurrence of flood in 2003 and 2006, taxpayers were exempted from 50% tax liability which reduced the amount of collection is in those years.
To examine the UPs revenue mobilisation performance, tax effort index is important which is calculated from actual level of taxes divided by the estimated level.Over the five years, with similar revenue structure, average tax effort index of HRP UP shows 21% above average while for MRP and LRP UP, it is 8% and 43% respectively below the expectation level (see Table 2).

Table 2 can be here
It is found that in HRP UP, per capita own revenue is significantly higher due to non-agriculture based economic activity and higher population (World Bank, 2004, p.12).But under similar revenue structure and effort, in compare to 2001/02, the average annual growth of revenue and collection performance in 2006/07 of MRP UP (85%) is higher than HRP (16%) and LRP UP (56%).Despite agriculture based economic activity, low population, and natural disaster prone remote rural area, better performance of MRP UP is due to some innovative efforts and strategies being taken in dealing with local leaders, politicians and common people through SLGDP.This UP have initiated some strategies like, employed young-new tax collectors, prepared house to house assessment rolls, motivated tax payers to participate open budget discussion, discouraged tax defaulters in providing any certificate (like, nationality, death, birth or inheritance, etc.) from UP, encouraged tax payers with tax rebate as incentive and made local publicity including demonstration of development projects implemented from UPs own fund, and a social motivation for all categories of taxpayers with minimum tax burden to establish tax compliance culture.
On the other hand, in HRP and LRP UP tax collectors have been working for about 10 to 25 years whose average age is more than 60 years and insignificant local strategy were adopted to motivate local people in participating UPs open budget discussion until government has imposed condition under BBG system.HRP and LRP UP functionaries mentioned that the collection of local tax and fees traditionally has got little importance by the previous office bearers and owing to longer term tax non-compliance culture constrained the way in enhancing collection, within any preceding particular regime.In LRP UP, representatives also recognised that own initiative of UP chairman has compelled UP members and secretary to increase the assessment and collection without adopting any new measures.On the other hand it is known from HRP UP that all residential households of UP jurisdiction are not under holding tax coverage though all industrial and commercial establishments are in holding tax coverage, and thus, in HRP UP, real picture of holding tax is yet to display in the accurate amount.
Existing revenue instruments for all UPs show that viable revenue sources available to UPs and utilised by UPs are quite limited with very lower contributions (see Table 3).Ratio of tax and non-tax revenue in HRP, MRP, and LRP UP are 59: 41, 25:75, and 33:67 respectively.Among total revenue sources, it is found that in HRP UP, though holding tax constitutes the highest proportion but it able to capture only about 46% of the existing potentials.In the case of MRP and LRP UP, shared revenue constitutes the highest proportion and these two UPs, in comparison to their existing potentials, also able to capture only 70% and 50% holding tax respectively.(Note 4) It is also observed that the non-tax revenue share from market auction and land registration fees are lopsided shares to all UPs.(Note 5) Table 3 can be here From Table 3 and calculation from UP and Upazila records, it is found that the general trend of share from land registration fees in three UPs are increasing, and the average annual growth of this source over the five years are 48%, 41% and 25% for HRP, MRP and LRP UP respectively.Majority UP functionaries' as well as local people opined in favour of increasing UPs share from land registration fee from 1% (i.e.0.025% of the sale value) to the break of 10% to 15%.If it is considered that the existing share to be increased to 5%, the contribution of this source in UPs total revenue will be increased to 55%, 73% and 84% as against current share of 20%, 35% and 54% for HRP, MRP and LRP UP respectively.Local people argued that UP should receive an appropriate share, since the land is located in their jurisdictions, and the activities of them have an effect on the value of property.Similarly, it is found that majority UP functionaries of HRP, MRP and LRP UP expressed their demand to increase the share of market auction fees from current 5% to the break of 21% to 60%, while majority local people of HRP, MRP and LRP UP have views of increasing the share to the break of 11% to 20%.If it is considered that the existing average annual revenue of this source will be increased to 20%, then the contribution of this source will be 42%, 64% and 34% respectively for HRP, MRP and LRP UPs total current revenue.
Local Government (Union Parishad) Act, 2009 includes provision for UPs share from Land Development Tax (LDT) which is yet to be decided.Based on annual average LDT collection, if UP is provided with percentage amount, there will be differential revenue impact on each category of UP.Calculation from UP land office record shows that if there is an option of providing 3% share, total revenue will increase by 5.5%, 1.9% and 1.7% for HRP, MRP and LRP respectively and if the option is for 10% share, then the impact will be 18.3%, 6.2% and 5.8% for HRP, MRP and LRP UP respectively with an average14.5% for all UPs total revenue.(Note 6) National expert and UP functionaries also contended that these are the sources which can be adjusted in term of fixed amount plus percentage to minimize horizontal imbalance.
The collection of business licences fees, non-motorized vehicles fees are deficient in compare to the existing base and rate potentials.UPs own statistics show (see Table 1) that in HRP, MRP and LRP UP, there are good number of commercial shops and small industries, which are not being fully taken into tax coverage and it shows UPs inability or lack of efforts in tapping those untapped revenues.Similarly despite potentials, revenue sources like, fees on profession and tax on signboard have never been collected by any one of the three UPs.In FGD, four main reasons are identified in this regard, which are: firstly, tradition of weak enforcement has induced non-compliance which has now posed a threat of potential conflict and challenge to establish a compliance culture; secondly, establishment of compliance is not feasible due to lengthy time, costly process, perception of local political pain and above all, UP has to rely on local administration and police authority for legal supports which require prudent local coordination; thirdly, at present, commencement of any business enterprises does not require UPs clearance, some enterprises are regulated by sectoral authority and in some cases fees are levied by ZP, UZP and District administration, which also revealed the overlapping and weak coordination of revenue administration in the local level; and fourthly, since no visible benefits are linked to tax payment, UP functionaries are not perceived any accountability to central government or to electorates and even they do not want to antagonise wealthiest part of the society due to their self interest.
It is observed that existing regulations empowered UPs to mobilise resources from local sources through assessment and levy of taxes, leasing of local hats and bazaars (marketplaces), water bodies, etc, but they do not receive the total resources generated from their entitled sources.In addition, controlling and administration of local sources still remain with deconcentrated agents, like, share of land registration fees, market auction, land development tax, marriage registration fees, etc.This clearly indicates that the UPs own revenue sources definition is somewhat misleading and virtually UPs have no direct control even over resources generated from respective UPs jurisdictions.Such a practice of regulating and controlling of the financial resources by the central government functionaries keeps the local UPs ever resource poor and resource dependent on the central government.

Table 4 can be here
Despite limitations, recent trend shows that in all three UP cases, revenue assessment, collection and development expenditure have increasing trend which is reinforced by the introduction of open budget discussion, discretionary and performance grants though the sustainability of this trend depends on both central financial and non-financial support including local extended efforts in exploiting resources from additional and new sources.The increasing trend of UPs revenue assessment and collection has impact on UPs development expenditure.It is found that in HRP UP the higher per capita assessment of revenue has strong positive linear association with per capita revenue collection with significance at 0.05.Similarly very strong positive associations are also being observed in the case of MRP and LRP UP with significance at 0.01.On the other hand, the linear association between per capita revenue collection and per capita development expenditure shows positive relationship in the case of HRP and LRP UP while it is negative in the case of MRP UP with no significance level (see Table 4).The negative relationship of MRP UP is due to its higher non-development expenditure in comparison to other two UPs and UP chairman said that these expenditure are related to motivation of local people, community involvement strategy and motivational programme like, court yard meeting with female households, and training on local resource planning which, they thought, will have longer term impact on UPs future revenue flow.About UPs expenditure pattern for 2004/05 to 2006/07, it shows that development expenditure of HRP, MRP and LRP UPs are 85.5%, 61.8% and 77.8% respectively, and for all three UP cases, it is 70.7% in an average.The rest is recurring expenditure and major portion of this expenditure is staffs salaries and office expenditure which varies with the activity level of each UP.Composition of development expenditure shows (see Table 5) that for all three UP cases, major portion accounts for rural earthen road construction and maintenance followed by education, socio-economic infrastructure, public health and social services.For individual case, rural road/culvert construction and maintenance of HRP and LRP UP is higher than that of MRP UP.In the case of other expenditure, HRP UP has emphasized on socio-economic infrastructure followed by human resources development and education and MRP UP has emphasized on education followed by public health, social services and socio-economic infrastructure while LRP UP has emphasized on education, social services and public health.6) that during 2008/09 to 2010/11, total basic expenditure need for HRP, MRP and LRP UP is 4.7, 3.0 and 2.8 times respectively higher than previous three years actual expenditure.All UPs have emphasised on rural road communication whereas, this policy is not consistent with the central spending guideline which states that UP can spent only 20% of their discretionary grant for earth work.(Note 7) It also shows that average one-third of UPs expenditure need is currently supported as compared to the projected expenditure need.This indicates that UP is running with two-third deficit of their required basic small infrastructure development expenditure in every year.UP functionaries opined that imposed condition for discretionary block grants seriously limits the scope of local level planning while they have to compromise in satisfying the immediate needs of the community with highly constrained resources.National experts have opined that interference in UPs functional autonomy might discourage their revenue effort and consequently weak local participation.UP functionaries also added that UP's capacity for the development works is restricted by the limited ADP allocation and the complexities involved in the timely flow of fund and this tradition, to some extent, has undermined UPs image to the local community and UP thus, has failed in capitalizing from the potential role of local self-governance.
However, the overall pattern of UPs development expenditure shows a common fashion of infrastructure biases.The other remarkable issue evident from the development expenditure is that there are considerable variation in terms of need and demand for the particular expenditure as well as annual per capita expenditure which is due to variation in population size, land area as well as resources allocation differentials.The attributes of expenditure pattern is also provides result that UP with lower population size but has higher land area can spend less in total (Tk.6229,000) term and more in per capita (Tk.66.74) terms which is evident in MRP UP (Bahl et. al., (2010).On the other hand, UP with lower land area but have higher population can spend more for infrastructure and can receive more allocation (Tk.7771,000 for HRP and Tk.6855,000 for LRP) form central grants though it seems less on per capita (Tk.21.03 for HRP and Tk.52.87 for LRP) basis which is evident in HRP and LRP UP cases.

Participatory budgeting and planning
Initially UP has started open budget meeting and participatory planning process under LGSP assistance programme, discretionary and performance grants system which is now taking formal shape through learning by doing and learning from mistake.(Note 8) It is evident that the new system of participatory budgeting and planning practices in UP has been facilitating democratization of local planning including a process of decision making power deemed has transfer effect from central to the local government.UP functionaries said that the new system of budgeting and planning has induced local people to make strategic plan in using UPs financial resources and realize taxes which were absent before.It is observed that the system has created local ownership and a stake in project selection and implementation with their local knowledge.In the development project implementation cycle though Union Development Committee (UDC) is the final scheme selection authority, they have to depend on local participatory selection process (Figure 1) which is ensured through Ward Development Committee (WDC) and in the implementation process there is separate Scheme Supervision Committee (SSC) who has to keep in touch with local people and local people can exercise their rights to know information about each and all projects that are implemented by the committee (Figure 2).For development scheme, UP can spend five per cent for selection, preparation and visit of scheme through participatory planning process.
Figure 1 can be here Figure 2 can be here In the local fiscal power, autonomy and efficiency of budgeting are crucial for local government.Though UP functionaries are held responsible for preparing annual budget as per guidelines embodied in administrative and procedural orders but it is still subject to scrutiny of Upazila Nirbahi Officer (Head of Upazila administration) at the sub-district level, Deputy Director (Local Government) and Deputy Commissioner at the district level.UP functionaries said that they have not been provided with training on operational side of budgeting like, defining budget cycle, make revenue estimates, classification of expenditure as recurring and capital and conduct open budget discussion.(Note 9) UP functionaries said that they have introduced participatory local planning in selecting and implementing project but in general, they had never been prepared five years plan.UZP supposed to compile all UPs five years plans as per Upazila planning book 1999, but this had never been prepared.(Note 10) This results in a general lack of supervision and control.Recently UP representatives have started ward based open budget meeting and planning where BBG and UPs own tax revenue are prioritised for development activities according to the choices of local people.In reality participatory system of budgeting and planning is still a difficult task which depends on local socio-political milieu.Participation can be effective when people have a greater say in how their local institutions and development priorities are addressed and managed.To understand the UPs current participatory budgeting and planning this study also focuses on weak areas that need to be resolved in the process.Thus, a pertinent question being raised in FGD with local people is 'Do rural people of different stratum participate and influence budgetary allocations and do their views find a place in open budget discussion?' Local people replied, 'We don't want to go with confrontation with political clout and muscle power of UP chairman and due to the same reason socially we don't want to go for friction with fellow residents'.This reply certainly reflects the less interest of commoners and disadvantaged groups and women to participate in UP budget meeting.In FGD, local people also commented that only some influential and party line political leaders articulate and express their views in the gatherings and common people are usually passive in such meetings.
However, overall responses show that open budget discussion has brought about diverse views among local people which also command citizens learning for meaningful discussion and to make UP budget as an actual planning tool.(Note 11) The national experts and ministry officials also noted that the discretionary basic block grants and open budget discussion in many cases have induced local people in sharing ownership of development projects as well as a learning forum for UPs resources and expenditure planning.

Disparity and effect of central grants on local tax revenue
Financial resources entirely used for the implementation of development projects in the UP jurisdiction are channelled through agents of line ministry and departments of sectoral ministries and these resources are coordinated at Upazila level, not at UP level, except discretionary block grants and UPs own revenue.UP functionaries and local people said that major part of the development resources that are utilised by Local Government Engineering Department (LGED-an agent of line ministry), and even by sectoral departments in the UP area, are on the arbitrary basis resulting to a highly politicised system, without proper consultation with UP and for such huge expenditure, democratic governance is ignored in practice.(Note 12) In addition, they said that the selections of LGED projects as well as sectoral allocations are biased to resource rich and politically influential UPs and thus, they capture more allocation than remote rural UPs that contradicts with the government block grant allocation policy which is based on land area, population and backwardness.(Note 13) Local people also said that influential local informal leaders or party line UP members use their connection with parliament member to gain membership on union-level resource committees and allocate resources for their interest and thus, outsiders intervention become apparent into the UPs revenue mobilisation domain (Lewis and Hossain, 2008, pp.33-51).
UPs overall financial resources mobilisation for development activities during five years period show (see Table 7) that there are considerable variations in the proportion of resources raised locally and allocations from higher tiers or central government.It shows that average annual per capita own tax revenue contribution for development works of HRP (Tk.12) is 1.5 times higher than MRP UP (Tk.8) and 6.0 times higher than LRP UP (Tk.2) which indicates that MRP and LRP cannot create their required own revenues for local development expenditure.Per capita basis central discretionary grant (BBG) also shows considerable variations in three cases (2:8:6 for HRP, MRP and LRP UP respectively) while it is equally allocated to all UPs irrespective of fiscal capacity/performance which ignores equity issue of UPs expenditure need, rather it accentuate only equality in nominal term.
Similarly, based on aggregate amount, resources from ZP and UZP are also discriminating.From UZP, per capita allocation is provided for HRP, MRP and LRP UP based on formulae set by the local government ministry but from ZP, the per capita allocation does not follow strict criteria which show that MRP UP has received nearly three times higher than HRP and two times higher than LRP UP.Accordingly, resources flow to UP jurisdiction through LGED shows that HRP and LRP UP is 2.48 times and 1.52 times respectively higher than MRP UP which indicates that for the last five years in MRP UP, LGED has implemented very few projects or projects having very low cost.(Note 14) The sectoral allocation shows that HRP UP is 2.0 times higher than MRP and LRP UP but in terms of per capita comparison, MRP is 2.0 times and LRP is 1.4 times higher than HRP UP.On per capita basis sectoral allocation, if provided with a view to confiscate the slack economic conditions of the rural areas then more allocation to MRP and LRP said to be rationalised but when own revenue capacity, land area or local development indicators are concerned, higher allocation to HRP UP might raise question about the rationale of allocation system.Thus, empirical evidence shows that there is no system works in redressing such vertical imbalance and allocation disparities.

Table 7 can be here
Resources flow from different channels to individual UP also shows that due to higher population, HRP UP, in all cases enjoying higher aggregate allocation and generating higher own revenue than the other two categories of UP but based on per capita comparison, HRP UP is getting comparatively low per capita except LGED allocation.HRP UP is a sub-urban area and its population is 4 and 3 times higher than MRP and LRP UP respectively, while land area of HRP is 2.8 and 3.8 times less than MRP and LRP UP respectively.MRP and LRP UP functionaries also raised their other arguments about the central allocation policy where they noted that despite HRP UPs prospect in raising higher own revenue, it is still getting higher total allocation as compared to other categories of UPs in the backdrop of similar development functions being assigned to them.National experts opined that as per present context, there might be rate support grant for some UPs in addition to block grant system, (Note 15) which is due to their varying needs of expenditure for similar nature of projects having harsh climatic impact and natural geographical conditions and difficult terrain (Bambang and Martinez-Vazques, 2002).

Table 8 can be here
Despite disparity in central grants allocation and low incentive for local resources mobilisation it is found that the declaration and commissioning of discretionary grant and performance grant have created a competitive environment in enhancing tax revenue collection.It shows that despite fluctuation of conditional grants, discretionary grant has stimulation effect on UPs revenue augmentation onward 2003/04 (see Table 8).For HRP UP, tax revenue generation in 2001/02 and 2002/03 is 5% and 66% whereas in compare to 2001/02, tax revenue generations during 2004/05, 2005/06 and 2006/07 has increased to 157%, 166% and 188% respectively.For MRP UP, the effect of discretionary grant is higher than HRP UP which shows that in 2004/05 to 2006/07 the tax revenue generations have increased to 372%, 475% and 730% respectively as compared to 2000/01.Similar effect also evident in the case of LRP UP, which shows that in 2004/07 revenue collection have increased to 277%, 645% and 534% respectively, as compared to 2000/01.(Note 16) Table 9 can be here It is also found that there are strong positive relationship between discretionary grant and UPs tax revenue collection for MRP UP (0.993) with significant at 0.01, for HRP and LRP UP this relationship is also showing positive (0.858 and 0.826 respectively) with significant at 0.05 but the relationship between discretionary grants and conditional block grant for HRP is negative and for MRP and LRP UP, it shows positive weak relationship with no significance level (see Table 9).

Conclusion, Observation and Policy Implications
Union Parishads being century old institutions and democratically elected bodies, working closest to the rural people, have little achievements in relation to serve rural people to whom they are responsible by the statute.Central government, donors, international agencies and higher tier rural local governments for the last one decade have been showing their serious concerns over how this body can operate, and to what extent UP can fulfill expectations of rural citizen.To this end, the introduction and implementation of fiscal decentralisation in UP is still challenging.This study, however, have tried to uncover some common features and findings through comparative study of three typical categories of UPs.It is believed that the commonalities of analysis and findings have implications for UPs in Bangladesh.Similarly, from development countries perspective, analysis and findings of this study has policy implications with respect to participatory budgeting and planning, local feedback of discretionary and performance grants, central government's orientation to local revenue sources adjustments and bringing innovation in local resources mobilisation as per individual country's social and economic context of the lowest tier rural local government.
Study has revealed that in all three UP cases introduction of discretionary and performance grants including participatory budgeting and local development planning system become convergence to the central-local relations that resulted in local revenue augmentation while there are divergences still being evident from disparities in central allocations, weak central support for local revenue administration, imperceptible coordination of higher tiers transfers, external political interference in UPs revenue domain, lack of understanding about authority and functional jurisdiction between deconcentrated agents and Union Parisahd, existence of bureaucratic spiral in the local revenue raising powers and budgeting, non-consulted LGED and sectoral allocations including lack of citizen learning in the participation of preparing UP budget and plan.The other divergences are being evident from central government's real political choice in the fiscal devolution rather a kind of pretend in showing respect to donors' preferences.Similarly, the intergovernmental fiscal system is yet to design for different tiers of rural local governments with a view to match functional responsibilities and fiscal resource transfers.Central government should have realistic information about extent of local tax bases to define fiscal capacity and fiscal need of each UP.From the finding, it is however, difficult to make conclusive comment regarding substantial revenue performance of particular UP because tax revenue potential and collection trend showed that there are still existence of huge untapped revenue that can be exploited with extended local efforts and central incentive mechanism.
It is also found that conditional grants in various forms have shown negative relationship in the case of high revenue generating UP and very weak positive relationship in the case of moderate and low performing UPs which suggests that increase of discretionary block grants to encourage UPs own tax revenue collection and on the other hand, strategy is required to provide incentive based conditional grants or substitute part of conditional grants by discretionary grants keeping an eye on UPs own revenue efforts.This implication is also similar with the study findings of Capuno (2002), Orial (2002) and Faguet (2003) that show how fiscal decentralisation has brought about enormous changes in local revenue rise.There are also lessons from fiscal decentralisation reform in Indonesia, South Africa and Tanzania regarding formulation of local government budgeting based on expenditure needs and minimum service standard including allocation of central resources, based on equitable share in the form of unconditional block grants that has to use on the basis of demographic and economic development profiles of local governments (Martinez-Vazques et.al., 2004).
It is observed that there are many avenues still available for the increase of UPs share from local sources such as land registration fees, market auction fees, which are considered as easy to do and have a big impact on the local resources available to UPs with only a marginal effect on national budgetary resources.Share of land development tax and marriage registration fees can be determined with rational basis which are other extended areas of UPs stable local revenue flow.In addition, UP functionaries also need to provide more efforts towards improving internal revenue sources within existing legal frame and sought for new sources for capitalization including instill transparency and accountability in their revenue management structure.This can be effectively carried out through formal local participatory budgeting/governance system and by strengthening UPs standing committees.Finally, it can be commented that like many other lowest tier rural local governments, UP cannot be a complete self-sustained body with its legally assigned sources but this body should try to raise own financial resources to the optimum level from the existing bases, so that, more development schemes can be implemented in rural areas than the embarrassing position now prevailing.World Bank. (2004) Menon (2008) and GOB (2007) in their study on 30 and 10 UPs respectively, showed that average annual own revenue of each UP is Tk 470 000-540 000 with minimum Tk 145 000 and Maximum Tk. 1719 000.Similar flow of UP revenue during 2003-07 was found in the study report of Rupantor (2008), a NGO working with a project known as 'Improving Local Level Governance by Strengthening UP and Creating Citizen's Awareness Including SIDR Reconstruction Activities' under USAID finance, covering 50 Union Parishad in southern coastal part of Bangladesh Note 4. Since ability to pay method is used by UPs instead of any particular valuation method, Holding tax revenue potential is calculated from UPs assessment rolls where eligible HHs are multiplied by the minimum average holding tax rate that based on per capita HH which is applied to each eligible HH and this minimum average rate is estimated (Tk.75.00) and agreed by UP functionaries and local people during FGD.
Note 5. Report of the 'Local government institutions strengthening committee', government of Bangladesh, recommended in April 1998 that UP should receive lease money from market @50%, land registration fees @10%, share of land development tax @10% and accordingly report of 'Local Government strengthening committee' in November 2007 again recommended that land registration fees @2%, lease money from stone or sand extraction @3% and share of land development tax @3% can be provided to UP.But none of these recommendations were implemented yet by the government.Note 6. Report of the 'Local government institution strengthening committee', of the government of Bangladesh in April 1998, had recommended to provide UP a share of land development tax @10%, and later on another report of 'Local Government strengthening committee' in November 2007 recommended to provide a share of land development tax @3% to UP.Note 9. Open budget meeting is organised at the UP office in the last week of May.Around 120 to 200 people normally attend in a meeting.UPs draft budget is made public by distributing draft printed copy from UP office.There is a fixed date to discuss the draft budget.In discussion, participants review the last year budget including implementation progress, incoming year budget with the annual projects/plan that come from participatory planning of ward level UP representatives.Participants can ask for the clarification of different revenue and expenditures priorities stated in the budget document, review relevant other documents and assert their comments and recommendations to incorporate in the budget.
Note 10.Information is based on FGD with Upazila officials'.Note 11.A nationwide survey in 2005 by Unnayan Shamannay (a Bangladeshi NGO) on 480 respondents selected from 52 UPs and 28 municipalities revealed that most of the female respondents (above 85%) do not possess basic knowledge regarding participatory budgeting, while a significant number of male respondents (roughly 75%) have some knowledge on such an important issue.
Note 12.This finding is also supported by previous study conducted by CARE Bangladesh that conducted a research on 'In Pursuit of Power: Local Elites and Union-Level Governance in Northwestern Bangladesh'.Under guidance of Mick Howes (Consultant), August 2000.Note 13.Eminent economist Professor Mahmud, W (2002) observed that resource allocations from higher tiers of government to localities (such as Special relief programmes and the ADP allocation to UPs) are often determined by personal or party political imperatives rather than by need.Note 14.The study of Mahmud et.al. (2008) have confirmed similar finding in other areas of the country.
Note 15.Since the present grant system is not sufficient to finance UPs development activities, local government strengthening committee in April, 1998 had recommended to provide an annual fixed percentage of 40% to UPs, 30% to Upazila Parishads and 30% to Zila Parishads from fixed 20% of the national revenue which is followed in many other developing countries like India, Thailand, Philippines, Indonesia.But unfortunately this recommendation has not yet been implemented by the government.

Note 7 .
Government of Bangladesh, Ministry of LGRD circular no.116 date: 5 February 2005.Note 8. Ministry of Local Government, Rural Development and Cooperative issued a guideline for the use of development grants which provides condition for participatory budgeting and planning (p.9) with effect from 2004-05 and government in 'Rural Development policy 2003' has emphasised on participatory planning at UP level (para.5.1).

Table 1 .
Note 16.Approximately similar trend of UP tax revenue generation during 2003-04 to 2006-07 was found by Rupantor (2008) -a NGO working with a project known as 'Improving Local Level Governance by Strengthening UP and Creating Citizen's Awareness Including SIDR Reconstruction Activities' under USAID finance, the study report has covered 50 Union Parishad in southern part of Bangladesh.Selected UPs profile Computed by author from UP office Data, 2008 a Number of observation is 5 (2002/03 to 2006/07) *Correlation is significant at 0.05.**Correlation is significant at 0.01.

Table 5 .
Composition of development expenditure a Source: UP office, 2008 a Amount shown in thousand Tk. excluding rural works programme (like Rural maintenance programme, Vulnerable group development, Vulnerable group feeding and Test relief), food grants and sectoral ADP; and figure in parentheses are percentages of total.

Table 6 .
Composition of projected development expenditure a a Amount shown in thousand BDT excluding rural works programme (like Rural maintenance programme, Vulnerable group development, Vulnerable group feeding and Test relief) and food grants and figure in parentheses are percentages of total.

Table 7 .
Disparity of average annual financial resources inflow for UPs development scheme Amount shown in thousand BDT (2002/03 to 2006/07) and figure in parentheses are percentages of total.b Per capita calculation is based on census 2001 with adjustment for subsequent population growth c Performance grant received by UPs 2004/05 -2006/07. a

Table 8 .
Effect of central discretionary and performance grants on UP's tax revenue collection a Source: Computed by Author from UP office data, 2008 a BBG is Basic block grant and PG is performance grant (in Tk.) and both are discretionary grants, figure in Parentheses are revenue collection growth in percentages in compare to 2000-01.