Report No. 587-TA E PP F E -T-- - -- -: - . _: -- .I A r L Lw tu Id[ V-did: Frisldi AsPe§O of Decentraiization March 19, 1975 Eastern Africa Department Not for Public Use Document of the international Bank Tor Recolstruct.ion andu Developrnent International Development Association This report was prepared for official use oniy by the Bank Group. it may not be published, quoted or cited without Bank Group authorization. The Bank Group does not accept responsibility for the accuracyr Or romnlptoness of the report This reportn , prepared by Mr .7-e P Gandhi, is based on the findings of his Special Mission to Ta.nzar.ia during July-Augus t, 1974. TATfTL OF rCONTTS1Tr INTRODUCTION Page No. Ph-loso1ph1y Dehi-nd D-ece..tralization - - - - - - - - - - - - - - - - -1 Objectives ol' Decentralizatio.. - - - - - - - - - 2 Prossible Dangers o'L D-ecentralization - - - - - - - - - - - - - - - - - Organization of the 'Report - - - - - - - - - - - - - - - - - - - - - - I. FINANCIAL ARRANGEMENTS UNDER DECENTRALIZATION Local Government Finances Prior to Decentralization - inequities of Local Finance -- Local Finance Reforms, 1969/70 - - - - - - - - - - - - = u Decentralization: Administrative Structure - - - - - - - - -9 Decentralization: Planning System - - - - - - - - - - - - - - - - - 14 Decentralization: Financial Arrangements - - - - - - - - - - - - - 17 Taxation and User Charges - - - - - - - - - - - - - - - - - - - 17 Level and Pattern of Development Expenditures - - - - - - - - - 19 Regional Development Fund - - - - - - - - - - - - - - - - 20 Development Budget - - - - - - - - - - - - - - - - - - - -21 Level and Pattern of Recurrent Expenditures - - - - - - - - - -23 II. DEMOCRATIZATION OF EXPENDITURE DECISIONS: IMPACT ON SECTORAL ALLOCATIONS Development Expenditures … - - - - - - - - - - - - - - - - - - - - - 25 Recurrent Expenditures - - - - - - - - - - - - - - - - - - - - - - - 32 Conclusions … - - - - - - - - - - - - - - - - - - - - - - - - - - - 37 TABLE OF CONTENTS (Cont'd) III. FINANCIAL AUTONOMY UNDER DECENTRALIZATION AND SELF HELP IN DEVETOPN4F.NT Page No. Expenditure Discretion - - - - - - - - - - - - - - - - - - 38 Tender Procedures - - - - - - - - - - - - - - - - - - - - - - - - - 39 Procurement System - - - - - - - - - - - - - - - - - - - - - - - - - 1 Self-Help Contributions - - - - - - - - - - - - - - - - - - - - - 42 TV TMTR-R17ITfNAT EVITTYV TNT QrOT AT LID lT 'CES ~ L.tU. o JA2 LViUEI.J Inter-Regional Imbalances - - - - - - - - - - - - - - - - - - - - - 44 National Targets - - - - - - - - - - - - - - - - - - - - - - 46 Fiscal Implications of National Targets - - - - - - - - - - - - - 47 Equity-Oriented Allocations - - - - - - - - - - - - - - - - - - - - 52 Parastatal Expenditures - - - - - - - - - - - - - - - - - - - - - - 56 Conclusion - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 57 V. WORKING OF THE FINANCIAL CONTROL MECHANISMS Control Mechanisms at the Formulation Stage - - - - - - - - - - - - 57 Conditions for "Supplementary"… - - - - - - - - - - - - - - - - - - - 58 Control Mechanisms During Implementation Stage - - - - - - - - - - - 59 Working of Control Mechanisms - - - - - - - - - - - - - - - - - - - 60 Failure to Produce Final Accounts - - - - - - - - - - - - - - - 60 Problem of "Excess Votes" - - - - - - - - - - - - - - - - - - - 61 Conclusion - - - - - -62 VI. SUMMARY AND CONCLUSIONS Tanzania's Own Brand of Decentralization - - - - - - - - - - - - - -63 Goals of Decentralization - - - - - - - - - - - - - - - - - - - - -64 Generalizations: - - - - - - - - - - - - - - - - - - - - 54 Achievements - - - - - - - - 4 - - - - - - - - 4 Limitations - - - - - - - - - LIST OF TABLES Table ,4o. Page No. 1.1 Tanzania: Degree of Ministerial Decentralization Since 1971/72 .11 1.2 Tanzania: Revenues to be Collected by Regions (Budget Estimates). 18 1.3 Tanzania: Sectoral Guidelines for Regions, 1974/75 . .22 1.4 Tanzania: Regional Recurrent Expenditure Allocations, 1973/74 - 1974/75 .2.h...... . . \/ 2.1 Tanzania: Regional Development Expenditure by Planning Stages. 1974/75 ... . . . . . . . . . 26 2.2 Tanzania: Dodoma Region Develonment Expenditure Requests and Budget Allocations, 1973/74 and 1974/75 .. . ........ - . 28 2.3 Tanzania: Coast Region Development Expenditure Requests and Budget Allocations, 1973/74 and 1974/75 .- .-.. 29 2.4 Tanzania: Arusha Region Development Exnnritiirp Requests and Budget Allocations, 1973/74 and 197L/7= - - 30 2.s Tanza.nia: Setnrnl Allocations of (Budgeted) Regional Development Expenditure, 1973/74- 107A/I7r 197 175 . . . . . . . . . . . . . . . . . . . . . . - 2.6 Tanzania: Dodoma Region Recurrent Expenditure Requests and Budget Allocations . . . . . . . . . . 33 2.7 Tanzania: Arusha Region Recurrent Expenditure Requests and Budget Allocations . . . . . . . . . . 2.8 Tanzania: Coast Region Recurrent Expenditure Requests and Budget Allocations . . . . . . . . . . 35 4.1 Tanzania: Regional Socio-Economic Indicators,L (Former) Eighteen Regions Around i973i/74 . . . . . 4 Tanzania: Cost Estimates or Capital ProJects . . . . 49 4.3 Tanzania: Capital Cost Estimates of National Targets- 50 4.4 Tanzania: Share of Regional Budgets in the National Budget, 1972/73 - 1974/75 ............. . 53 LIST OF TABLES (Cont'd) Table No. Page No. 4.5 Tanzania: Regional Expenditures Per Capita, 1972/73 - 1974/75 . . . . . . . . . . . . . . 74 4.6 Tanzania: Regional Development Allocations for Social Services, 1974/75 . . . . . . . . 55 i RO wDtUUTUIUN 1. Since the Arusha Declaration of 1967 Tanzania has set before itself the goals oI socialism and self-reliance. In achieving these goals the Government has recently emphasized an urgent need for the development of the rural sector where the bulk of the countryis popu- lation lives. It has also recognized the limitations of centralized planning in bringing about widespread rural development consistent with people' s participation. Philosophy Behind Decentralization 2. Tanzanian leadership has expressed its belief that peqple must be given "power over their own lives and their own development"±LI and this naturally called for a reform in the existing system of planning where "to the mass of the people, power is still something wielded by others -- even if on their behalf". The decentralization of administration of Tanzanian Government, announced by President Nyerere in May 1972,_/ should theret'ore be considered a major institutional change in order to democratize the decision-making process for planning and foster development from below. 3. Before July 1972 the planning and control of development of the Tanzanian economy was concentrated in Dar es Salaam and so were the bulk of Government's technicians, engineers, bureaucrats and finances. The country did have a system of local governments but their role in economic development was really minimal, their finances very weak and their staff extremely poor. While many ministries from the Center did post s9ome of their own statf in the regions and districts, such officials were given no authority of their own nor did they enjoy any discretion. In fact each such functional officer felt more responsible to his own ministry in Dar es Salaam than to working in coordination with others in solving local rural problems. _/ All the quotes in this Chapter have been taken from President Nyerere's speech, entitled "Decentralization" given in May 1972. 2/ The original Act in respect of decentralization was passed in July 1972. It was called "The Decentralization of Government Administration (Interim Provisions) Act, 1972" and was to expire on July 1, 1973. The life of this Act has now been extended up to December 31, 1975. 4. The wishes of the local people did not seem to influence the r Jor ords - lans -L. a.j si.iLficant- m=arer. ConsenterCl.yr people felt no involvement in or conmitment to their development plans. It wa3 thrfr diffic-' L±lal peopLle tJo responUd U±U LUw.LLUith ULbo ent,h---usilasr. to a call for development work which may be to their benefit, but which hLLas bL-eer, decided uapon ari-d plara-tldu b. arn authl-orituy 1h-1-ndreds of ,M,l-es away". This lack of local participation in development seemed to Tanzanian leaders contrary to the country*s ideal of democratic socialism. Objectives of Decentralization 5>. It was to remedy these weaknesses of the existing administrative structure that decentralization of the planning process as well as implementation of the plans was considered absolutely essential. In announcing the new structure the Government hoped that the following objectives of development would be simultaneously furthered: (a) As under the new system "the Regions and Districts will be free to make their own decisions about priorities and methods of work", this would mean that the process of decision-making would be substantially democratiT.ze it would also mem that the wishes of the local people from now on would greatly influence the formulation of the local plans. In addition, to the extent lower level (regional and district) administrators would be provided with the qualified functional staff from the Center to achieve inter-functional coordination and "for the im- plementation of their approved part of the plan", more effective project preparation and implementation could be expected to result. On the whole, thus, the major purpose of decentralization was "to ensure that nf`ture economic planning stems from the people and serves the people directly." (b) If "local responsibility is to become a reality, greater financial control must be vested in the Regions and the Districts. They must also be allowed to exercise self-discipline in financial matters, and be allowed some discretion in the use of money allocated to them." This would mean that Regional and District Administrations would enjoy substantial spending authority and financial discretion under the new system so that "the anount of red tape and bureaucracy" would be cut down and "the thicket of red tape and the tyranny of 'the proper channels'" eradicated. As a result it was believed that marked improvement would take place both in plan implementation and speed of development spending. Besides, the regions and districts would also have some incentive to cost- effectiveness, for "If they are able to economize . . . while maintaining acceptable standards, they will be allowed to use their savings' on other projects." (c) Once the local people can see their needs being fulfilled, they would be eager to participate and contribute voluntary labor to the development projects. Consequently self-help element in the development effort would sub- stantially grow. (d) The abolition of the former system of local governments and its replacement by the decentralized administrations would permit the Central Government to implement its policy "of a gradual equalization of well-being between different regions" and the achievement of inter-regional equity in the availability of essential public services. Possible Dangers of Decentralization 6. All these great hopes from decentralization were, of course, not without dangers. The Government showed awareness of at least two possible dangers and took the necessary safeguards. 7. First, as the lower levels of administration would not have any revenue raising authority of their own, it is possible that they might not fully appreciate the "value" of public funds or the "costs" of raising tax revenues. 8. In order to make them fully conscious of the need for fiscal discipline, it was clearly laid down that "Districts or Regions or Divisions or Ujamaa villages, as the case may be, will not be allowed to spend more than they are given." Besides, it was emphasized that they would be held fully accountable to the public. "If a region gets the money for five dispensaries and health centers, they will have to show five dispensaries and health centers at the end of the year or produce the money in lieu . . . (i.e.) at the end of the year they will have to be able to show projects for all the money they have been allo- cated and have spent - together with the accounts!" 9. Secondly, in a situation where the desires of the masses are to be given due recognition there exists a real possibility that public resources might in some sense be "over-allocated" to the nrnvisinn of social services at the cost of productive investments. Or, in other words, decentralization could possiblv carrv with it t.he seds fnr -n undesirable "imbalance'" between social consumption and economic investments in the use of limited public resources. 10. Once again, this danger was foreseen and it was stre d +.ha1t. "Each District Plan will have to be prepared in such a way that, in general terms, the extra production of goods which will reuilt from it could be expected to do much more than pay for the future recurrent costs of the non-revenue earning investments it proposes." Organization of Report 11. When President Nyerere announced the decentralization of Tanzanian Government administration he observed that "The details of the financial arrangements are still being worked out." The present report therefore begins with a description of the financial, as well as admini- strative and planning arrangements which have evolved since May 1972 (Chapter I). 12. It then attempts to evaluate, though somewhat qualitatively at this stage, the achievemenLs of decentralization in terms of its above mentioned objectives viz: (a) Democratization of the decision-making process and its impact on sectoral allocations of government resources (Chapter II); (b) Financial autonomy and discretion given to the regions and districts and its impact on self-help contributions for and local participation in development (Chapter III); and (c) Decentralized Planning and its impact on the achievement of inter-regional equity in public services (Chapter TV). 13. The report also indicates certain problems with the working or the financial control mechanisms and highlights the distinct possibility of expenditure explosion following decentralization (Chapter V). 14. Finally, the report concludes with certain hypotheses regarding the imDact of decentralization which should be tested as more data become available (Chapter VI). ('rA-PTHR T. FTNANGTAT, ARRAN Mz.Fa S TThNfl.? TWNETRAT.T7ATTnNU 1.01 A description of the financial arrangements between Central Government a nd iPgionn1 and D strict Adm.inistrations stv begin with a statement on the state of local government finances before decentraliza- tinn anci the ecnnomir implic-atJons nf su'rhin a -1Kiti+AnV had it been allowed to continue for long. A brief description of the present administrative and nlannninp' qtmr,tnrPre is nlso relevant to an 11 wersta¶e, o, the present financial arrangements. Local Government Finances Prior to Decentralization 1.02 Before the decentralization of administration in 1972. local government system of Tanzania was organized into 58 rural district councils and 15 town councils. The rural districts diifered from each other widely in terms of area and population. Masai, the largest rural distErict; for e,ramnl- was more than 120 times Mafia TslAntL the .ma11lest in area. Similarly, Kilimanjaro, the largest in population, was over 28 times MnfianITlnd. gnm.e ys al so trie of to..m coArnd-c-ilc -- Dar es Sala-am. the largest municipality was 33 times the smallest town of Bukoba. 1.03 All local authorities undertook activities generally fit to be pursued at the local level; n add-tion, the ad-in4istration of priary schools and health services had also been made their responsibility since 'lOh To 1 f ce these e s r 1 Al -" -A'_ ^ A 4 - .___ bulk of their revenues from local rate -- a sort of graduated poll tax -- and p -vI-ce cess. 'lust of the .e4.e ofthe 4-- -1 on t 1' ~ ~ ~ U ~~1 U W~ 4 V .LAU~Q %J4 LILAU V%JW.ll U1IJJL UJLIt, other hand, came from three separate property taxes -- site rates, urban housing tax, anA Uu-icipal house taA -- and liUe, fees and other charges. In addition, both rural and urban councils received loans financed from thIe l-oca'l gover-nment s ILroans Board. Publ--ished dAt for 1967 suggest--- W" Ao^ > s oGU W Jiass:XDo U3 L- XuLLs- LI a ltU UdlU LUI -l7U[ Z /UuggeS;U the following relative importance of different taxes in local government ..ances: local rate 36%, produce tax 6%, urban property taxes 6%, urban licesnes, fees and charges 8%, market dues, primary school fees and other adm-inistrative ch-.arges 14%, and income from property and investments 4%/ Central Government grants-in-aid accounted for the remaiLJnlng 260. / The expenditure pattern oI local authorities, in turn, was as follows: education accounted for;,6%, admrinistration 29%, health 13%, roaVU 77 au natural resources 3%._ 1.05 The Central Government in Tanzania was not satisfied with the financial position of the local authorities and this for many reasons. 1/ About 82$ oi -rnese grants were given ior eaucation krinancing teacherst salaries in particular) and another 12% for health services and the rest for road maintenance and towards the salaries of senior administrative staff. See Tax Reform Planning in the United Republic of Tanzania (a study by Ibrahim M. Kaduma and Ji2tendra R. Modi presented at the Meeting of the Expert Group on Tax Reform Planning, 8 - 14 September, 1970, under the auspices of United Nations, New York), Table VIII. 2/ These percentages relate to the year 1966 and are taken from IMF's Report on Local Government Taxation in 'lanzania, prepared by S.B. S5cutton (January 2d, 1969). 1.06 First, as the 1966/67 Report of the Public Accounts Committee oi the Pa'-aimen' s4tressed !"Ife ---gavl concerned aL LI-- detrira --! --. financial situation of the Local Government Authorities as revealed in this an.d previous reports. hllere are evidently serious short-comangs in both the collection and spending of revenue, and we consider that to remedy these, the uejRnxive re-orgwaization of the system of local govern- ment is necessary."-/ The local authorities were not generating enough revenues on their own and as a resuit their dependence on Central Government grants was rapidly growing. Whereas in 1961, Central Government was called upon to contribute only ll,o of total local authority expendi- tures, in 1967, its contribution had grown to about 26%.T! The "own" revenues of the local governments did not show much elasticity -- they remained at about 4% of GDP all through the decade of sixties and, as a share of total (central and local) revenues declined from 21% in 1961 to 151 in 1968. The Central Government transfers to local authorities absorbed almost 40% of the former's total current expenditure in 1968/69. Despite all this help from the Center the liquidity criqis of local authorities was forcing many to ignore capital spending altogether.-' 1.07 Secondly, the manpower and administrative costs of collecting most local taxes, local rates and urban property taxes in particular, were found to be so high and tax evasion so wiaespread that it did not seem worthwhilp to continue collecting many of these taxes without major reforms.~Y Inequities oI Local Finance 1.08 hnird, and this perhaps was the most important problem, local authority finances perpetuated many inequities. There were significant differences in the levels of public services offered by different districts as indicated by the local expenditures per capita. Whereas, for example, the cost of local government was Sh 18 per capita in the rural districts in 1966, it was Sh 101 per capita for all towns. The substantial dif- ferences between rural and urban areas were thus apparent. Even among themselves, there were substantial differences. Among rural districts, for example, local expenditure per capita ranged between Sh 11 in Kigoma to Sh 32 in Kilimanjaro. Similarly, in the town councils, it ranged -_ between Sh 65 in Dodoma to Sh 106 in Dar es Salaam and Sh 113 in Moshi.5/ 1/ Quoted in the Budget Speech, 1969/70, p. 16. 2/ Tax Reform Planning in the United Republic of Tanzania, Tables VIII and XII. 3/ R.G. Penner, Financing Local Government in Tanzania (1970). 4/ Eugene C. Lee, Local Taxation in Tanganyika (1965), pp. 39-LL. 5/ IMF's Report on Local Government Taxation in Tanzania. pp. 8-9. .7 - I - 1.09 While some of the inequity in the local government cost per capta m- ight have res-lted from the differences in tne size and density oi population as well as the size of the area comprising various district or town councils other reievant factors here were the differences in the levels of prosperity of different rural districts and towns and the di-ferences in the rates and structures or taxes which otten determined their respective revenue generating capacities. 1.10 The rates of taxes varied significantly from one district to another. The amount of minimum local (personal) rate, for example, varied from Sh 25 in Nachingwa district to Sh 65 in Kilimanjaro district. The maximum rate also ranged from Sh 60 in Maswa district to Sh 450 in Kwimba district. The levels of income to which the maximum local rates applied also varied from Sh 2,400 to Sh 25,000. The tax treatment of women also differed from district to district. Such a great variation in local personal rates and exemptions from one district to another was obviously highly inequitable and unacceptable in a socialist country. 1.11 Same was also true for produce cess, a kind of local sale tax on marketed produce including exports, levied on agricultural, livestock ana fish products up until 1965 when the Central Government forced uniformity from above in the rates of these taxes. It has been estimated that about 88% or the produce tax revenues werec,ollected from the smallholders and only 12% were attributable to estates.Y1 This, once again, was unacceptable to socialist leadership. 1.12 The burden of local taxes in town council areas also varied from one area to another. On the assumption that the burden of property taxes (i.e., site rate, the municipal house tax and the urban house tax) was shifted to tenants and lodgers, the burden per household, independent of size, was estimated to range between Sh 35 in Tabora to Sh 100 in Arusha._/ 1.13 It was not only that the local taxes generated and perpetuated inter-district inequities. most of them were also considered to be regressive with respect to income and inelastic with respect to tax bases. Largely because of their regressivity with respect to innome it was fo'nrid that the average burden of local taxes did not vary much between rural and urban areas. The average burden of local taxes in rural districts was Sh 51 against Sh 66 in the towns.Y 1/ Ibid, p. 27. 2/ Ibid, p. 33. 3/ Ibid, p. 36. - 8 - 1.1L The rural-urban inequity, the inter-district as well as the inter-personal inequities implici+t i the local +tx s-+qtp s became the focus of Government's attention particularly after the Arusha Declaration of 1967 setting the co-utlry on the path o+' socilism. The Government now desired a near unit ormity in tax treatment, removal of certain nuisance taxes and the introduction of' progressivity in the +ax burdens. The facts that local authority finances were in a critical situation -- 3L out of 58 rural districts and 8 out of 15 to-wn councils overspent in 1966 running down their reserves to dangerously low levels or ending the year with bank overdrafts -- and that press-ure was mountting through the Public Accounts Committee of the Parliament as regards the quality of local tax and expenditure administration forced the Government to start thinking about the overall reorganization of the system of local governments in Tanzania. Local Finance Reforms, 1969/70 1.15 As a first step, the Government took the follo-wng meas-res at the time oI its 1969/70 budget. 1.16 First, thte local (personal) rate was abolished with effect 1rom January 1, 1970 which was the start of the next financial year for all district councils. 1.17 Second, the produce cess was abolished immediately, i.e., with effect from June 20, 1969. 1.18 Third, it was announced that in due course the Government would a-iM to have a single tax levying and collecting machinery to serve both the local authorities as well as the Central Government. The idea behind such a move was to use the country's tax system as a developmental and redistributive tool in a more vigorous fashion. 1.19 It was estimated that the district councils would lose about Sh 110 million in revenue due to the abolition of the local rates and produce cesses. To balance their budgets, the Central Government decided to take over some oi' their more expensive functions, particularly the maintenance expenditures on rural health centers, rural water supplies, district roads, and the payment of salaries of categories A and B primary school teachers. At the same time it was also decided to make compensatory payments of about Sh 32 million to the district councils. 1.20 To finance the extra burden of these compensatory payments and their newly acquired functions from local authorities the Central Government decided to levy a national sales tax on a variety of imported and locally manufactured articles. By abolishing the two local taxes, which were primarily paid by the farmers, and levying a national sales tax, which was expected to be paid primarily by the urban people, the Government immediately shifted the tax burden fra rural to urban population and corrected the past rural-urban tax inequities resulting from local taxation. -7- 1.21 During the period 1970/72, rural local governments thus operated at a substanTially truncated level. They were stripped of all of their taxing authority though they did continue to collect fees, market dues and certain other charges yielding relatively small amoun't of reve,nues. On the expenditure side their activities in relation to rural health, roads and water supplies were completely taken over. As f-ar as the town councils were concerned their functions and taxing powers were not changed up until early 1974. 1.22 As a result of these changes, the revenue pattern OI dibtrict councils changed markedly. In 1971, 80% ot their revenues came from the Ce±itral Government as compensatory payments and grants, 7% from primary school fees, 5% from market dues and other fees, 4% from liquor licenses and another 4% as income from rented property and investments. On the e=enditure side now close to 60% of their receipts were spent on educa- tion, 15% on administration and 25% on all others.L/ 1.23 All these were obviously trarzsitional arrangements which, as far as local governments were concerned, could not have lasted for long. In effect, they paved the way for the abolition of local governments as separate entities and the decentralization of (Central) Government administration in May 1972.2/ It must be pointed out here that the process of decentralization is still continuing. For ex-ample, all urta, councils were brought under decentralized administration early in 19714/ and two new regions of Dar es Salaam and Rukwa were formed on January 1, 197L. Decentralization: Administrative Structire 1.24 The main highlights of the decentralization announced in 1972 were as follows: IMF Report on the Rationalization of Taxes in Tanzania, prepared by R.N. Malhotra (June 29, 1973), -op. 5a-57. 2/ Even as it was a two-tier Grovernment performing largely the same kinds of tasks it was considered by many too costly for a countrv with limited skilled personnel. 3/ The 197h/75 budget also aholished the ennnm-l ii4rha nlnd and property taxes accruing to the urban councils and replaced them by a single charge called t.he "Lnnd Rent and Serice Charge" yielding a similar amount of total revenues to the Central eAchequer= - 10 - (a) First, the rural district governments were abolished right away and their assets and liabilities becaml, the assets and liabilities of the Central Government.. Their existing staff now became Central Government employees subject to the same salary structure, personnel policies, and conditions ol service as the existing employees of the Central Ministries.,/ A decentralized structure of ad- ministration replaced the former local governments. This structure consisted of the Central Government, Regional Administrations and District Administrations. The structures of the last two are described in the attached charts (pp. 12 and 13). The whole country was initially divided into 18 (now 20) regions and about 65 (now 73) districts. (b) Second, agricultural and rural development as well as the development of small industries and commerce were made the primary responsibility of the District ancd Regional Administrations. To make this ef'fective, nine Central Ministries viz. Agriculture, Lands, Housing and Urban Development, Natural Resources and Tourism, Health, Education, Water Development and Power, Commerce and IndustrY, Communicationsand Works, and the Prime Minister's Office (responsible for Regional Planning) were decentralized. As Table 1.1 shows, Regional and District Administrations were made responsible immediately in 1972/75 for spending about 52% of the recurrent budget and 10 of the develop- ment budget of the decentralized Ministries. (c) This also meant that their ' taff was now dispersed into the Regions and Districts.2 These staff were now to receive their direction from and be responsible to the Regional Development Director rather than anv Central 1/ Section 8(b) of' the Decentralization Ant. 2/ Section 9 of the Decentralization Act, 3/ Tt is estimatAd that. nearly 'i-nnn h1gh_-manlit c+ntral empoyees have been dispersed into regions since 1972/7B. Many of them had been prrnmntpd to higher posit-ions as an "i-centive"% In addition, they had to be provided with housing in the regional headquartver rnr araz Al + n~n .n^. - -sua headnunr+-s o t21 reas ^11these facto,rs "I'A V have resulted in substantial budgetary costs but then they are considered i-.s expecteda- on co maison deetra tble tbenen.its whi te ratton is expected to reap f'rom decentralization. Table 1.1: TANZANIA: DEGREE OF MINISTERIAL DECENTRALIZATION SINCE 1971/72 (in millions of Shs.) _Recurrent: Eenditure Development Expenditure 1972/73 1973/74 1974/75 i)X773 1973774 1 '74 (i?A) (AE) (BE) (PA,) (AE) (BE), 1. Office of the Prime Minister 84.2 81.0 60.6 17.7 69.1 87.3 2. Agricalture 88.8 200.4 103.1 102.2 146.4 319.6 3. National Education 192.0 224.4 308.1 59.6 53.8 63.3 4. Commerce and Industry 10.6 14.4 20.7 - 26.3 47.3 150.5 5. Cbmmunication and Works 74.6 90.2 99.0 359.2 374.3 202.9 6,. Lands, Housin,g and Urban Dev. 9.4 9.7 12.0 8.5 30.5 30.4 7. HEealth 95.1 96.2 116.2 5. 0 35.5 47,7 8. National Resources ancd Tourism 39.3 34.7 41.8 39 .9 54.6 49.7 9. Water Resources and Power 42.9 32.9 52.8 189.6 348.7 380.1 10. Nine Ministry Total 6,36.9 783.9 814.3 808.0 1,160.2 1,331.5 11. Regional Total 700.4 773.5 976.7 92.4 194.5 297.5 1:2. 1(10) + (11) 1,337.3 1,557.4 1,791.0 900.4 1,354.7 1,629.0 Ratio of 10 47.6 50.3 45.5 89.8 85.6 81.7 PA = Provisional Actual AE = Approved Estimate BE = Budget Estimate SOURCE: Budget Estimates, Various Years [ REIGIONAL ADMINISTRATION 1 Regional Commissioner: C hairman j _ _ __ ._ ___--_ POLITICAL BODY: Regional Developmernt EXECUTIVE BODY: Regiona.l Detvelopment Council Te am Regional Commissioner: Chairmai iRegiorLal Development Director: Chairman _____ _ _ jegional Development Director: Secretary Regional Regiona] Regional Planning Finance Personnel Officer Officer . Officer _~~~~~~~~~_ _ __ _ _ -_~~~~~~~~ - _ 1 _ . _ I -_E fl i TANU Regional Members of Parliament _ Regional Regional Regiona:L Regional C.hairman: Memer from the 'Region: Members Agriculture Health Water Land Dev. Off.icer Off icer Officer Of'ficer .~ ~ ,z _ -I ___ - l _ _ _ _ _ _ FTANU District Area District 1 Regional Regional Regional R egional Chaircmen :membersL Commissioners Devealopmient. Com Wrks Natural EducatioDn Commerce -Members Directors: Off licer Resources Officer Industry Members _ Officer fLf ficer _ District Development Directors [TThSRICT ADMINISTRATION . POLITICAL BODY: District Development Council [ EXECUTIVE BODY: District Developmaent and TANU: District Chairman: _Chairman _ Planning Coummi.ttee Area CommiLssiciner: Chairman )is-tr ictD~e-velopmejnt I)irec:tor:7Se-cretai7y__ L2istrict Development _ _irector: ._creta~Wj I Districit Development Director: Secretary, ________ | -and Chairnnan, District Dev. Team lected lemberIs1 Members of Parliament . LE_from_the_District_ Me Members of 1h4th Elected -i~m the District: Members I Parliament Members - II _______ ____ ~I rom the Dist_ . Brea Comiissioners: Staff of the District Developraent ____ ex-off icic, Member an.Planning Team: E,x-of ficio Meiab~ers '= _ Distr:Lct District District Planning Finance Pers onne:L Officer Off icer Officer D-is triLct District Di trict f Dis trict7l Agriculture Healtlh Water Land Dev.j Officer Officer OfficerJ O Officer J DiLstrict District District Dl)istrict |Colmm. Works Nat. Res. Education f ComME!rce & Officer Officer Officer Industry -____ - LOfficer _ - ±4 - Ministry.!/ The remaining staff of the nine decentralized Ministries now played the role of generally laynrg down- the policy framework and plan guidelines, fixing user charges, and giving technical assistance and guidance whenever necessary. They were not to monitor and control the work of the officials employed in tne Regional anrd District Administrations. (d) Fourth, the Regional and Area Commissioners who were to be appointAd by the President were to be like [uLunisters in Regional and District matters. Similarly, the Regional Development Directors ana District Development Directors, to be appointed by the President's Oifice, had the same position as principal secretaries to ministers. Tnus in their own jurisdictions the Development Directors were given real power and responsibility for management, includ- ing the administration of regional (and district) budgetary f'unds. (e) Filth, political bodies and executive agencies were fully integrated under decentralized administrations. For example, the proposed plan of a district or a region is prepared by the executive agencies but it cannot be forwarded to the higher agency until it was approved by the TANU District Executive Council or the TANU Regional Executive Council. Decentralizatiaa: Plarming System 1.25 The most important impact of decentrali7ation was, however, on the planning process and the implementation of development plans. With effect from July 1, 1972 the responsibility for planning and control has 1/ Yet, as mentioned earlier, the personnel policies have now become centralized. The hiring, firing and promotion of onal staff remain a central responsibility. This system create problems of "dual loyalty" of the decentralized f who might now be spending too much time writing reports nded to make an impression at the Center and perhaps too le on actual development work. been dezenttralized to the Regional and DiMSrict Admiistrations. They are now re¢%pcnsible for carrying out, t'ae plans as -cell. T'he role of the Primie sini-ter's Office is imply to provicde the overall plan guidanlinez, momitor the physical perfornrance of tle plan, coor&inate the phasing of' the projects, and to provide aid in eff.ective cor-ct1ve actions once t'-.e. Tower level administrations havs identi:ied potential and actiial pcthlems or bottlenecks. l 26 Although the system of decetitralized planning is still evolving, efforts are be%ng made to inculcate the following two elements into it: FlrstJ,lW the Plar.ning process must really start at the village and ward lewels. As it, stands now, district is the first level at which the formal plrn.s orizti.ate and are prepared. The initial sugeestions made by the people for the development projects in their area are compiled and codified by thle district functional officers and used as a basis for drawring up the district plans. These draft plans are then approved by the people through their District Develonment Committees as well as by t.he TANIT Dir trict Executive Committees before they are sent. upwards to the Ragional Acbdmnistration to be recon.eInd with plans of other districts and finally to be forwarded to the Center. 1.27 In formilating their plans, the District Administrations are expected to .nPciftv not only how Ymch each project will cost or what type o' manpower, equ:Lpment and materials will be needed but also how ma-ny, peopl-e I I benefif f. ?rnm the pro n n.- arA ,P.a TAl_ 1 'ha Ai phases of the physical progress of the project. Ln this way it is P-eqo,tM.a tk.a+ olpri+.s ^P a pr,, ll e,es 4nl-nei r;+h- its .p- andr a I-- p11 -s..~ 6- - - -L% ---t -tS b . V - - planning control mechanism will be automatically established. decentralization with Regional Administratlions more or less acting as J..; _- 'J J^J.;.LU*, L4jt UVsV ± ~±U CO _1Lj, ..~ C.U ±e OUA Is _vv-zV' J-JAL '45UVUA LILAM CLLs& s AlcU. DLLIC kotMlUlt;.W 1.28 SecondAS3 Vhe oVt:I--ll pla I-c21 o (;L)-4LJationfJ- J carriU eUUI out >,, te Prime Minister's Office where the "Plan Guidelines" are drawn. These ing the overall sectoral irrvestment allocation which each region should .,7 * kva:sAJ O DLIJJUtk>, V.Lt.;VU.LOt,U U -%Pi UWII 0,VUK_ fLdc1 .L ;1DAG Besides, for the benefit of the Plannng Officers in the regions, the '"G-uid.elines" also specify the standard (materia) cost estimiates of various categories of projects. 1.29 The 1974/75 "Tlan Guidelines"., for example, suggested that 'nrc+.~7 j *J'.p.L4vduJctV nLIIV0V IIsIt.,sUV , VUc,UL±UILLC ±fILL. aI.dLLC..UJrLLU soc,al aLl- frastructuire investments should receive inve6tible funds of each region ;n t,he proportion- of 3V 16 and 50%, respectiveLy. his ov°eVall guideliLne could, hoaever, be modified only in the interest of inter-regional balance 4" ~~~I'W,,.;^ f'4...- 4.U. - .~~ I ..- - 1 _ I' - 8> PSF1C 9W JWSO JYnt,h naiorXl bige3 o *u-cation, healthv water :nipply. housing, etc., those regions which were below the "national av7rer3.5e" in rue,ect of a partic-ular service were oncouraged to spend more on it. The achievement of inter-regional equity in public services and taking the whole country towards the longrt=erm nat tiarne+.s with regard to different public service seem to have been two paramount considerations in the allocation of development funds since decentrali- zation. 1.30 Under Tanzania's decentralized structure, annual planning and control cycle goes through five stages: (i) Reviewing Performance (July-August). At this stage both District and Regional Administrations review the performance for the completed plan year and determine changes in the current and next (fiscal) year's plan. (ii) Establishing Plan Guidelines (October-November). At this tage the 'lower level adminIstrations review the prelififnary guidelines proposed by the higher level administrations. The regions review those proposed by the Center and the districts review those proposed by the regions. Only after their interaction the final regional and district guidelines are prepared by the Center and the regions, respectively. (iii) Preparing Plans (December-February). During tnis period, the District Adilinistrations identify programs to be included for the next two years, select and dociment specific projects for the first plan year and prepare the draft district plans. The Regional Administrations simultaneously document possible regional projects, integrate regional and district project proposals and prepare a draft regional plan. For each project a stage- by-stage "Action Plan" is also expected to be formulated. (iv) Reviewing and Approving Plans (March-June). During this period, the Center approves the regional plan for the next fiscal year and authorizes its implementation and similarly Regional Administrations approve district plans and authorize their implementation. (v) Controlling Implementation: Over the plan year regions and districts are required to produce "Quarterly Physical Progress Reports", "Quarterly Financial Reports", and an "AnMal Programmes Performance Review" for the higher levels of administration. These reports, along with any "PrrohlAn Report" which might be generated during the course of the implementation of a plan are used to monitor and control the plan implementation. - 17 - 1.31 The Government recognizes certain serious constraints on an effective planning under the decentralized system and these include a seriou1s lack of trained and skilled personnel, the inadequacy of even basic information necessary for district level planning, and the existence of severe phY sical com-nrication problems between the districts and regional headquarters as well as between the regions and the Center. With-in these constraintsj however; Ta nzania Government is doing as best as it possibly can; it certainly gives the impression of having generated a great deal- of official enthu+.siasm about and conmitment to the new experiment. Decentralization: Financial Arrangements 1.32 The financial arrangements between the Center and the Regional Administrations can be discussed Wnder various subc-heads The general questions of taxing powers of the regions as well as the methods by which the levels and pattern of their recurrnt and development expenditures are determined are discussed in this Chapter and Chapters II and IV. The questions of expeniLtur-e discretion, tender and procurament proced-res, etc., are discussed in Chapter III while the working of the financial controls is described in Chapter V. Taxation and User Charges 1.33 Under the decentralized system, the Regional and District Administrations have no taxing powers of their own. As was mentioned earlier, gradually the taxing powers or the former local authorities were taken away in the late sixties and early seventies and vested in the Central Government. Tne existing sy-stem of local taxation was found to perpetuate rural-urban tax inequities as well as inter-personal inequities. The centralization of taxing powers, it was thought, wo-Uld permit the Government to pursue a tax policy with a heavier equity orientation and also get rid of some taxes levied on the very low income earners since pre-independence days. In addition, it was argued that it would economize on the tax administration which was in short supply in Tar.zari±a. 1.34 The lower level administrations in Tanzania consequently receive all their funds from the Center and whatever little revenues they do collect they do so on behalf of the Center. As is clear from Table 1.2, they collect less than 2% of all revenues, primarily through urban water charges and rents of public houses and buildings. Even the rates of these charges are determined by the relevant ministries at the Center. 1.35 Although it is the responsibility of Regional and District Administrations to provide primary education, health services and rural water supply, it is the policy of Tanzania Government not to charge for these services; consequently no such revenues show up in the regional budgets. - 18 - Table 1.2: TANZANIA: REVENUES TO BE COLLECTED BY REGIONS, 1974/75 (BUDGET ESTIMATES) (rounded to '000 Shs.) Hire Rent of From Game Agri- Medical Plant, Govt. Forest and culture and Sale Vehicles Quarters Water Royal- Fishing Related Sewerage Dental Market of and and Region Charges ties Licenses Cnarges Disposal Charges Dues Fish Craft Offices OtMere* Total 1. Arusha 3,010 300 75 445 11 250 - 700 600 380 1,118 6,889 2. Coast 270 100 26 70 - - 50 25 - 30 329 900 3. Dodoma 1,120 125 25 556 90 55 260 10 500 150 854 3,745 4. Iringa 835 45 38 255 - 30 30 - 84 164 523 2,004 5. Kigoma 495 210 100 223 - 13 - 140 - 125 318 1,624 6. Kilimanjaro 3,626 632 186 244 501 19 490 15 16 196 1I211 7,136 7. Lindi 661 168 165 213 6 22 165 - 201 400 934 2,935 8. Mara 375 12 175 210 5 30 500 25 400 350 61 2,143 9. Mbeya 700 125 210 1,007 - 83 488 - 404 100 454 3,571 10. Morogoro 2,055 900 420 262 5 110 5 - 500 700 881 5.R38 14. Mtwara 770 180 60 48 1 40 115 12 550 305 471 2,552 12. Mwanza 3,545 80 291 115 105 10 430 150 500 500 1732 7,458 13. Ruvuma 690 5 26 145 - 11 10 - 5 45 351 1,288 14. Shinyanga 882 30 16 219 2 32 166 2 132 60 416 1,957 15. Singida 375 5 10 393 - 15 160 - - 250 341 1,549 16. Tabora 806 600 5 347 6 100 205 5 405 55 331 2,865 17. Tanga 3,650 300 151 5 20 400 2 20 500 500 559 6,107 18. West Lake 180 150 31 189 5 25 125 25 230 280 472 1,712 19. Dar-Es-Salaam - - 558 8 1,000 . 600 500 - 400 4,502 7,569 20. Rukwa 153 133 7 149 3 28 142 3 200 264 308 1,390 TOTAL 24,198 4,100 2,575 5,103 1,760 1,274 3,943 1,632 5,227 5,254 16,166 71,232 Percentages 34% 6% 4% 7% 3% 2% 6% 2% 7% 7% 22% 100% Of the Total Source: Financial Statement and Revenue Estimates, 1974/75, Vol I. Bulk of these are "Miscellaneous Receipts" from the Sale of Goods and Services and other Related User Charges as well as Rents of Public Property other than Government Quarters and Offices. - 19 - 1.36 Tanzania Government Is rationale of, not charging for these services even though bulk of government expenditure goes to finance these services is not spelled out in any official docurment. Yet it seems that government revenues could soon, and in fact already have, become a constraint on the country's rate of capital formation. An attempt mnut, therefore, be made to understand this policy. 1.37 The principal argument here appears to be that once a country has established national "targets" of universally providing social services like education, health and water supply the way Tanzania has (see Chapter IV below) it does not make sense to ration them through the price mechanism for that way it will be within the reach of certain people and beyond the reach of others. It is this principle which seems to have guided Tanzania's policy regarding "user charges". 1.38 It is the policy of the Government to provide community tap water within easy reach to everyone, hence there is no charge for it. However, if an individual wants "'special treatment" and wants to receive "piped" water linked up to his residence he has to pay for it. This explains why urban water charges appear in the regional budgets but no rural water charges. 1.39 Once again, it is the policy oI' the Government to provide universal primary education by 1989 (now revised to 1977). Consequently, primary school fees, which ranged between Shs 10-40 in different districts, were abolished last year. The secondary and university education are also free in Tanzania but there the Government tries to recover part of the cost of education through its strict incomes and salaries policy (including "low"incomes during the national service period of two years) and its policy of progressive income taxation. 1.40 The regional budgets do seem to contain some, though insig- nificant, receipts from medical charaes. They, once azain. reflect charges for "special treatment" which some patients receive in the form OI "better than average" or class A food or room accommodaticn. Otherwise medical services too are generally free in Tanzania. 1.41 It is obvious that once the taxing powers of the Regional Administrations have been taken away, and theV are not given. bv virtue of political decisions, any powers to collect "user charges" for most costly ot the services they provide. thev have to denend unon the Cen'er for almost all the funds needed to carry out their activities. An understanding of how the levels and sectoral allocations of regional expenditures are determined thus becomes relevant. Level and Pattern of Development Expenditures' 1.42 The manner in which the pattern of development expenditure for any narticla-r region. is I_ been -n d - == i s primarily dictated by the needs of the people and the overriding criteria of inter-regional equity. The needs ol' the people are partially, if rioL ful'ly, incorporated in the final a" ocations nsofar as the regiors are asked to speciIly their priority projects, along with their costs, should they receive only 50% of the previous year's allocations or only 752 or 100,? or 125%. This way the Prime Minister's Office determines what the regional priority projects are -Iithout actually speci-Yring the eXact allocation of' investible funds to a region. How a decision is made as to how much o0' developmemit htnds wrill a reglon receive in a partI.Ic lar year is, however, less clear. 1i.h4 The regions receive two separate kinds of development funds -- an iunrestricted" Regional Development Fund (RDF) and a "pro-ect/prograume earmarked" Development Budget. The process by which RDF and development expenditure allocationo are made is described now. Regional Development Fund 1.44 The Regional Development Fund (RJDF), which is like an !uncon- ditional" grant was set up in December 1967. Initially it allotted a sum oa hal -million shillings to each region to carry out locally initiated projects. Tlhis amount was then raised to one million shillings from 1968/69 onwards. The Presidential Circular 1 of 1968 recoummended that RmDF shoula be mainly used for development projects which involved communal actlvities and which were related to economic productivity. In particular, it recommended thai it shold be utilized to (i) increase productivity of crops, (ii) proNide better storage and marketing facilities, (iii) establish cottage industries and (iv) improve transportation facilities. o ever, in actual operation, it has been founa that bulk of this sum has gone to f'inance social services and not productive investments. The size of the Fund has been growing since its inception. It was 24 million shillings in 1970/71, 50 million shillings in 1971/72 and 45 million shillings in 1974/75. Since 1971/72 Norway and other Scandinavian countries have been supporting RDF on an increasingly larger scale. Their financial contribution has increased t'rom 6 million shillings in 1971/72 to 15 million shillings in 1974/75. 1.4) The criteria oI' allocation oa. RDF amongst regions has also undergone change over time. During 1968/69 and 1969/70, each region was given a flat sum of 1 million shillings. This absorbed 18 million shillings and the rest was kept in "contingencies". In 1970/71 for the first time, the criterion ot "Tneed" was brought into consideration and size o0 the population of a region became the sole basis of allocation of RDF. A fixed sum of 2 shillings per capita was allocated to each region with the result that the region with the largest population received the highest allocation. More recently, the "need" criterion has been further developed and the RIF is now allocated on the basis ol population ow a region, size of its area and the inverse of' the per capita output of' that region. - 21 - Development Budget 1.46 As far as development budget of a region is concerned, some six months before the fiscal year is to begin, the 'ireasury, Development Plan Ministrv and the Prime Minister's Office carrv out a forecast of the level o± development resources which would be available tor the next fiscal year. On the basis of the past experience and taking into account the pipeline of ready projects, the Central Government indicates the "CGuideline" allocations for Gentral Ministries. narastatals3 and regions. For 1974/75, for example, th; respective shares proposed in the "Guidelines" were h4k for the lentral MinistTries 43A for the narastat1s.q nn 23f for the regions. While the latter proportion does not appear to be very large at nresent it has grnwn from 14L in 1972/73 and 20% in 197A/7)_ and the intention is to raise the regional share even further as the latter's administ.r2tive nannhili+y- andt absorptinvr apc+exad 1. A7 The "Pla C-n uidel4hes a4ther lay doiin "sectoral allocations" as a guide for regional planning. Table 1.3 gives the 1974/75 sectoral estimates of 1973/74. It highlights how the Central Government very much desires that region.s ej+and their investment in directly productive sectors and reduce development expenditure on economic and social 4"frastructure. 1.48 How far has the Govexmnment succeeded in achie-v-ng thib or ls likely to succeed in this respect is discussed in the next chapter. For the piresent, it needs to be noted that the Prime indicates to the regions only the total development allocation lor all regions~~~~~~~5_~ &nal theC; secvra UU-14.e butJ dosro-tl yregionJ -whatL thJe level of its aevelopment allocation would be. This has become a cause of compla.t ;t V-." certa4Xr. reg on, parti UL.-cL 'al.y ULIe bet;UteL-oVLi regio_Lns with large absorptive capacity, who often put up a well-prepared plan only to -ind in due course that they au-e n,ot goin8g to get the desired amount of money. 1.49 In the last analysis, the goals of inter-regional equit;y of pubic sece .^ theL "neds oi a region seem uo play an i-mpOrtanl. role in determining both the level and pattern of development expenditure for a region. The available pipeline of projects or the absorption capacity o± a region is less of a determinant at this stage though it cannot be denied that when it comes to act-ual spending these facors do play an important role. Experience shows that richer regions quickly spend their allocations and act-ualy come back to seek development sup- plementaries while the poorer regions are unable to spend even the buugeted amo-ults. For example, the available data show that whereas Arusha (per capita income Sh 762) had spent almost 90`1 of its allocation \..iU±LuC±-ig 1- ouppieumeriuary of 7 million shillings) during the first nine months o± the fiscal year 197/724 Tanga (per capita income Sh 443) hiAd spent onlPy 4OS,- and Doaoma (per capita income Sh 415) only 52%. For all regions taken together the ratio of actual spending to total al- locations was 65 for the sane period. Table 1.3: TANZANIA: SECIORAL GUIDELINES FOR REGIONS, 1974/75 1973/74 Budet E:stimates 1974/75 Guideline Allocat:ions Amount % Amount %_ (in million Shs) (in million Shs) I. Directly Productive Investment 46.6 26.3 86.0 34.4 Agriculture 28.2 15.9 43.0 17.2 Natural Resources 15.3 8.6 :30.0 12.0 Commerce and Industry 3.1 1.8 :13.0 5.2 II. Economic Infrastructure Investmlent 37.8 21.:3 40.0 16.0 Lands 4.3 2.4 5.0 2.0 Comwcrks 33.5 18.9 35.0 14.0 III. Social Infrastructure Investment 92.7 52.4 124.0 49.6 E'ducatiorn .22.6 12.8 33.0 13.2 Health 23.2 13.1 34.0 13.6 Water- 46.9 26.5 57.0 22.8 IV. TOTAL 177.1 100.0 250.Ci 100.0 Source: Prime Minister's Offire. - 23 - Level and Pattern of Recurrent hSxpenditure 1.50 Whereas in the case of development expenditure, a region does not know in advance as to how much it is likely to receive, in the case of recurrent expenditure it knows the "ceiling amount" which has been fixed for it. Only now it does not know how it has been derived. Within the ceilings, they are free to determine their sectoral allocations without any interference from the Center. However, as generally about 80% to 90% of their recurrent expenditure is already committed (e.g., about 66% of Tanga's recurrent expenditure in 1974/75 is 'Tersonal IXmoluments") there is not much flexibility left anyway. 1.51 The Treasurv Circular No. 1 of 1974, for example, inuormed the regions or their recurrent expenditure ceilings for 1974/75. Table 1.h compilps this information and relates it to the 1974/75 budget estimates which were finally approved as well as the budget estimates of 1973/74. ihe ceilings nPrmittPd a growth of about i to 7X with the excention of Kigoma and I.beya and the two regions of Dar es Salaam and Rukwa which hadi been newly formAd on TJnlAnrv 1, 1974i nuite obviouslv. the Treasurv had not been very realistic in imposing its ceilings on the regions. A. it turrned out later the regions had to be rgiven more than 251< increase over their last year's allocations. The recurrent expenditure estimates of regoins and ministries taken together were 311 higher in 1974/75 than in 1973/74. In part this large increase reflects the rise in mnnlmm.lm. wages and general mala1v inereasj ~ ~annoCenfed in Anril;- I974t rising prices oi goods and services purchased by the Government, and the e-rnPn.q in-ro1vpr i-n f,,illr .qt.nft9np t.h _.PntrnIi 7.-d administratinns and expanding their activities. 1.52 In addition the growing recurrent burden of the social infra- strulcturen "roiects,* ori w.vhich vrdormn4 n +the 1 e-,mx+. h,ie-tia+. of 'PM-nz!ni also explains part of the rapid increase. The system as it exists now is that-,. t 4-l an,,-r4- nfN9- 4% r - .e.a s n f n ,-v,.a t ar met by +he ace,, Jr LJ~C L LL*i . %.LI4.4.O. L 4. LJ*L 4. ~'4~Aj.4. -.I~ Ui . t k-. i j L LI C. - JUL. L/ Li U. ' A . L .Ot,- -jL4 from out of its block sub-head called 'Expenditure Arising from Five Year ralop n-mn-n+ Vlnnfl buti+ evnIjir forv +the fi _+ -ran" ri-i +oIn. oper'ti + of t.-he project. In the succeeding years, however, it must become a normal part OIr a aln-le-cyS bua+get TVhe _-roea nan,a a m Of Sh 11.7 I 4I17 to this sub-head in 1972/73, Sh 12 million in 1973/74 and Sh 15 million in 1 071, /7C5 1.5, The uoven-s,e. has beer cor.sc-ous ot' it ra- 2 -ro1 .1.. ).) Lk~ t)u V entAIR4. bL L1 LVI=~L %ttJ .~.LLU. 'J. L/V) 4".V~.LLL.j E,.L- .LA recurrent expenditure and, as stated in the Budget Speech for 1973/74, cosLUV.±-0 Lk, -MV0U.LIhL;VJLy -.uJIkjV'IOLi.LVU IA. u.0.rL U%A.LUjIId C.L IIIULZ~U± Liro considr it- zsltl !le-tivtot'e additioa ,esres to contro this rapid increase in recurrent expenditure which is so markedly in excess oYL . . . (thle r-a'ke ofL) economilc gro-wth. VYLIether uh GVQ-eIIrIUneIUli as the capacity or willingness to achieve this goal and what possible impact -wo-uld he decentialization of adminisbration nave on this is discussed in Chapter V., Table 1.4: TANZANIA: REGI9O4 REWRRENT EXPENDITURE ALLOCATIONS, 1973/74-1974/75 ,:-n m.___. on 6 Rate of Growth in Rate of Growth in 1973/74 1974/75 1974/75 Ceilings 1974/75.Budget Estimates Region Approved Estimate Ceilings Budget Estimate over 1973/74 (Z) over 1973/74 (2) (i) (ii) (iii) (iv) (v) (vil) 1. Arusha 49.3 52.7 59.8 6.9 21.3 2. Coast* 47.0 48.9 38.3 4.0 _ 3. Dodoma 43.7 46.2 54.9 5.7 25.6 4. Iringa 37.2 38.7 42.4 4.0 14.0 5. Kigoma 28.2 36.9 38.2 30.8 35.5 6. Kilimanjaro 45.6 48.9 63.1 7.2 38.4 7. L1ndi* 32.1 34.9 37.4 8.7 16.5 8. Mara 31.3 33.2 36.9 6.1 17.9 9. Mbeya* 49.0 55.6 53.5 13.5 9.2 10. *Murogoro 45.8 47.8 58.9 4.4 28.6 11. Mtwara* 36.2 38.2 42.5 5.5 17.4 12. Mwanza 56.6 59.0 62.4 4.2 10.2 13. Ruvuma 33.4 34.5 41.7 3.3 24.8 14. Shinyanga 35.4 37.4 47.4 5.6 33.9 15. Singida 31.6 32.6 43.0 3.2 36.1 16. Tabora* 40.2 42.9 44.0 6.7 9.4 17. Tanga 56.0 59.8 73.3 6.8 30.9 18. West Lake 44.5 45.9 49.2 3.1 10.6 19. Dar-Es-Salaam* 20.8 44.5 59.8 113.9 20. Rukwa 9.5 29.7 30.2 ;12.6 - TOTAL 773.5 868.3 976.7 12.5 26.3 i The 1974/75 and 1973/74 data cannot be related for these regions as they were reorganized affective January 1, 1974 only to form two new regions, of Dar-Es-Salaam and Rukwa. Source: Treasury - 25 - CHAPTER II. DEMOCRATIZATION OF EXPENDITURE DECISIONS: IMfPACT ON SBCTORAL ALLOCATIONS 2.01 One of the major objectives of decentralization was to decentralize and demcratize the expenditure decisions. It was thought that the previous system under which the Central Ministries in Dar es Salaam decided on the projects to be carried out in each district or region was not conducive to public participation. It was hoped that by letting people decide through their local political representatives and administrators as to what they wanted to spend their share of the country's tax revenues on. bureaucratic elements in expenditure decisions would be replaced by democratic elements. 2.02 The object of this Chapter is to suggest on the basis oI some verY preli-minarv data that this does not see to have hannmned to any significant extent. of course, the available data are highly inadequate to propose it as anything mrore than a mere suggestion at this stage. This hypothesis is based on two grounds: Development Expenditures 2.03 First, the Prime Minister's Office, guided by its estimate of the bud-eta v resources of the Crverm.ent, restricts veJr severelr the amount of development funds which each region receives as a share. However f m 4,I +~ ~,1 aTD1 v C.4AD1v",' Heer.^. viat is sig-.,ifi;<|^+- i4s that, through. i4ts UU 4. 4.4S t it also regulates and steers the sectoral allocations of development funds for a region as well. 2.o4 As a. evidence of the latter exaM-ne Table 2.1 whLich shows how the regional development allocations for 1974/75 did not change Mn - T44 a C-. --Tterinl bt etw-"er Stage I(-ilGu-de1ines% and Stage III. (r-enal Allocations). This seems to suggest that the Center's strategy oi al- lo a'tng larger sums to directly productive sectors and red-uclrg the share of social infrastructure investments, whatever the region's own desires, ^co4.'d weu-ll be LWvJrking . UL"U.LA. 1ur L orura Ly L.Lo d dataare zv-iailabLe regarding the actual regional development expenditures sector by sector n.d 4 -hismakes n eva uavion. o L,h WLjOat 'Ls hLLAp ULIrL r difficult. 2.05 Theoretically, of course, it is arguable that through a continuous pocess of inter.auctinw bet-L weer,W thIe pujb'Li, IOJ.o d- aftlin-isXrators and Central officials, the 'Plan Guidelines" come to reflect the "true" wishes of the people so that a simlarity between final development a±locations to regions and the initial "Plan Guidelines" issued to the regions cannot be tak-en to be indicative of the large influence which the Center wields in sectoral allocations for development. However, the probability that this would have happened in just two years after decentralization of the administration is rather small. In case such an assumption is granted, Table 2.1 would seem, prima facie, to suggest the hypothesis that perhaps "true" democratizationrwith respect to expenditure decisions has yet not taken place. - 26 - Table 2.1: TANZANIA: REGIONAL DEVELOPMENT EXPENDITURE BY PLANNING STAGES, 1974/75 (,J- miLL±L.Lon Sh.LLS /1 __2/ Stage I- Stage II-/ Stage III/3 Amount % Amount % Amount % I. Directly Productive 66.1 29.2 78.0 35.1 84.0 31.8 Aariculture 30.1 13.3 54.9 24.8 64.3 24.3 Natural Resources 16.4 7.2 12.4 5.5 12.9 4.9 Commerce and Industry 19.6 8.7 10.7 4.8 6.8 2.6 II. Economic Infrastructure 38.7 17.1 33.0 14.9 52.6 19.9 Comworks 32.5 14.4 27.5 12.4 47.8 18.1 Lands and Urban Dev. 6.2 2.7 5.5 2.5 4.8 1.8 III. Social Tnfr;astrutiiutire- 121.4 53.7 110.8 50.0 127.5 48.3 Education §1.4 18.3 38.9 17-9 43.8 16.6 Health 25.5 11=A 2S.1 11 32.8 12.4 Water Supply 54.5 24 1 468 21.2 50.9 19.3 TOTA1 226.2 9 1f 991M 100.0 926 .1* in /1 Stage I: As per'initial' Guidelines /2 Stage II: 'Revised'G-uide1nes following Regional Inputs and Cabin-et Comittee's Recommendations. Higher allocation to agriculture is largely due to the latter. /3 Stage III: 'Final' Allocations * Exclusive of Sh. 33.5 million of block funds earmarked for urban district and sub-district programs. Source: Prime Minister's Office. - 27 - 2.06 There is also a second way of examining the validity of this typothesis and that is to compare the requests of development funds made by different regions with the allocations they finally received. Should the two differ substantially not only in terms of levels (which would easily be understandable in view of the overall resource constraint) but also in the sectoral allocations this could be suggestive of the fact that the regions, after all, would not be able to spend the money the way they really wanted to. Here it must be pointed out that even the regional "requests" themselves are normally the outcome of Regional Planning Officer's review of the District Plans and local people's "desires" in order to accommodate the "Plan Guidelines" as much as possible. 2.07 To consider this ouestion; the relevant data were collected for three regions: Arusha, Dodoma and Coast. These data are presented in Tables 2.2-2.1. 2.08 Table 2.2 oives the relevant data for Dodoma region. I.n 1974/ 75, this region asked for 38 million shillings but got only 14 million shillings. This might be quite iinderstnabh1e. However, in this context it is interesting to note that Dodoma wished to spend only 16% of its develoDment funds on '"direArt1- prnouiiive" seotrs bu, t as allocated 32 of its funds for these sectors. similarly, it wanted to spend 73% of its funds on ".social infrast-cture" Yhfle the share of these sectors in the final budget has been reduced to 56%. 2.09 A similar story unfolds in the case of data provided by the Coast reg'ion which are given ri rable 2.3 .F 1 -I71./7, 4i4 reei-iA only 12 million shillings against 37 million shillings it requested. Tn its renupist .was inclrid a mu of 1c5 million shillings for the const;-* tion of the regional headquarters necessitated by the formation of Dar district. as - 'ar,,, - --1- -aV reA;^n bu or32U 2 I 5, -11J., -we .inLy allocated in the budget.1, 2.10 Then it wanted to spend about 49% of its development funds on "Pronom4oinrs-c- ' but these se^or weres 'eda do--n A.- 28% of its development budget. Besides, it would now be spending 22% of its developmen.t fnds on "directly productive" sector-s agaist-15% which it really wanted to spend. In addition, it would now be spending a higher proportion of its developme.t buduet on "isocial in.frastructure" than it really intended to. 1/ It is not clear whether this reduction in the budget allocation ref41ects a scaling do-wn of the project or merely a rephasing of the expenditures. Table 2.2: TANZANI:A: DODOMA REGION DEVELOPMENT EXPENDITURE REQUESTS AND BUDGET ALLOCATIONS, 1973,/74 AND 1974/75 (excluling District Development Corporations) (in million Shs.) 1973/74 _ _ 1974/75 Alloca- Alloca- .Request % tion Re aest tion __% I. Directly ProducctiNre 11.2 20.0 1.0 5.2 6.2 16.4 4.5 32.1 Agriculture 1.;7 3.0 0.6 3 3.2 8.4 4.4 31.4 Ujzamaa and Cooperatives 7.7 13.8 - -0.4 1.1 Natural Resources 1.6 2.8 0.4 2.1 2.6 6.9 Cl. 1 0.7 Comnerce and Industry 0.2 0.4 - - - _ - II. Economic Infrastructure 4.8 8.6 1.0 6.3 4.2 11.0 1.6 1L1.4 Conrworks 4.'5 8.0 1.0 5.2 4.2 11.0 1.0 7.1 Lands and Urban Development 0.:3 0.6 0.2 1.1 - _ C0.6 4.3 III. Social Infrastructure 40.0) 71.4 16.9 88.5 27.5 72.6 '7.9 56.5 Education 6. L 10.9 2.1 11. 0 4.9 1l2.9 3.2 22.9 Health 6.:3 11.3 0-8 4,.2 4.7 12.9 1.0 7.1 Water Supply 27.6 49.2 14.0 73.3 17.9 47.3 2 .7 26.5 TOTAL 56.0 100.0 19.1* 100.0 37.9 100.0 14.0 100.0 * Includes a major supplementary for a rural water supply scheme Source: Of fice of Dodoma Regional Development Director. Table 2.3: TANZANIA: COAST RE;GION DEVELOPMENT EXPENDITU'RE REQUESTS AND BUDGET ALLOCATIONS, 1973/74 AND 1974/75 (exClLlding District Development Corporations) (in maillion Shs.) 1973/74* _ 1974/75** Alloca- AiLloca- Retques %_ tion % Leque st % tion % I. Directly Productive 4.4 26.8 1.8 18.9 5.C5 14.8 2.5) 21.4 Agricultuare 0.6 3.7' 0.4 4.2 1.51 4.0 0.7 6.0 Ujarnaa and Cooperatives 1.2 7.3 - - 1.2 3.3 - _ Natuiral 'Resources 2.1 12. 8E 1.0 10. 5 2.51 6. 7 1.2 10.3 Comrerce and Industry 0.5 3.0 0.4 4.2 0.3 0.8 0.6 5.1 II. Economic Infrastructure 2.4 14.6 0.9 9.4 18.2 49.1 3.3 28.2 Comworks 2.0 12.2 0.9 9. 4 18.0 48,5 3.2 27.4 Lands and Urbtan Development 0.4 2.4 - - 0.2 0.6 0.1 0.8 III. Social Infrastructure 9.6 58.6 6.8 71.'7 13.4 36.1 5.9 50.4 Educationa 2.8 17.2 1.9 20.0 3.0 8 ,1 2.3 19.6 Health 2.3 14.0 2.6 27.4 3.1 8.4 1.4 12.0 Water Supply 4.5 27.4 2.3 24.3 7.3 19.6 2. 2 18.8 T'OTAL 16.4 1.00.0 9.5 100.0 37.1 100.0 11. 7 100.0 * For five districts includiLng Dar-Es-Salaam. ** For four districts excludiLng Dar-Es-Salaam. Source: Office of Coast Regional Development Director. Table 2.4: TANZANIA: ARUSHA REGION DEVELOPMENT EXPENDITURE REQUESTS AND BUDGET ALLOCATIIDNS, 1973/74 AND 1974/75 (in million Shs) 1973/74 _ 1.974/75 Provisional Alloca- Actual % Request % tion % I. Directly Productive 5.7 31.7 14.5 50.7 9.5 45.7 Agriculture 5.6 31.1 12.9 45.1 8.7 41.8 Ujamaa and Cooperatives, neg. neg. 0.2. 0.7 0.1 0.5 Natural Resources 0.1 0.6 0.4 1.4 0.2 1.0 Commerce and Industry - - 1.0 3.5 0. 5 2.4 II. Economic Infrastructure 1.2 6.7 3.0) 10.5 5.3 25.5 Comworks 1.0 5.6 2.0 7,,0 4.8 23.1 Lands and Urban Development. 0.2 1.1 1.0 3.5 0.5 2.4 III. Social Infrastructure 11. 61.6 11.1 38.8 6.0 2.8- Education 4.9 27.2 5.9 20.6 2.4 11.5 Health 2*.3 12.8 2.1 7.4 1.2 5.8 Water Supply 3.*9 21.6 3.1 10.8 2.4 11.5 TODTAL 113.0 1.00.0 28.6 100.0 20.8 100.0 Source: Office of Arusha Regional Development 'Director. TaLble 2.5: TANZANI,A: SECTORAL ALLOCATIONS OF' (BUDGETED) REGIONAL DEVELOPMENT EXPENDITURES, 1'373/74-1974/7' _ 13973/74 1974/75 Amount _ Amount _ (mill. Sh) (mill. Sh) I. Social Infrastructure (a) Education (Primary and Adult) 22.6 13.0 43.8 16.6 (b) HEtalth 23.4 13.5 32.8 12.4 (c) Water SuppLy 46.9 27.0 50.9 19.3 Sub-Total: 92.9 53.5 127.5 48.:3 II. Economic Infrastructure (a) Comworks ('Roads; & Buildings) 35.8 20.6 47.8 18.1 (b) Lands and Urban Development 3.5 2.0 4.8 1.3 Sub-Total: 39.3 22.6 52.6 19.9 III. Directly Productive (a) Agriculture and Crop Development 10.0 5.8 51.6 19.5 (b) L:ivestock Development 15.2 8.7 12.7 4.8 (c) Natural Resources (Fisheries, Forestry, Beekeeping, Game Parks) 14.1 8.1 12.9 4.9 (d) Small Scale Industries 2.2 1.3 6.8 2.6 Sub-Total: 41.5 23.9 84.0 31.8 TOTAL:; 173.7* 1(0.0 264.1** 100.0 * Exclusive of Sh 3.4 million expenditure budgeted for district development corporations and regional coopei-atives. ** Exclus;ive of Sh 33.5 mi'llion of block funds earmarked for urban dist:rict and sub-district programs.. Source: Prime Minlister's Office - 32 - 2.11 Arusha's final allocation for 1974/75 does seem to be fairly close to the amount it requested. Perhaps it is because Arusha, being a better equipped region in terms of staff, has well-prepared projects. But it too, like others, would have a development outlay pattern very different from what it requested. As an extreme example, for Comworks it has been allocated about 5 million shillings in its 1974/75 budget while it actually requested only 2 million shillings. 2.12 The evidence presented above is admittely inadequate at this stage, yet the little bit which is available seems to suggest that regions have neither been able to get the total amount of development funds they wanted nor did they receive the sectoral allocations they initially requested for. The Prime Minister's Office seems to have been able to control and regulate the level and pattern of regional development spending. Table 2.5 at once highlights the success the Center is having in its strategy of emphasizing "directly productive" sector investments and de-emphasizing "social infrastructure" investments. What the final outcome of this effort will be will not be known until appropriation accounts for 1972/73 and more recent years become available. 2.13 While the direction given by the Center must be considered quite desirable from the point of view of objective of achieving a certain target growth rate of the economy, inter-regional equity, and of containing the possible heavy public pressure for social infrastructure, at the same time it does indicate that the expenditure decisions have perhaps not yet been as much democratized as one tends to imagine by reading President Nverere's speech announcine decentralization. Recurrent Exnpnditures 2_11 As was mtentinned in rhqnter T. the "ceilings" on regional recurrent expenditures initially issued to the regions by the Treasury turned nit tn he pmite "unrealistA&'? in the rourse of time. The final allocations of recurrent funds to different regions were frequently more than what were originally proposed by the Center. Against a "ceiling" amount of 46 million shillings, for expample, Dodoma region received 55 million shIllIngs. Similarly, Arusha received 60 million shillings against its "ceiling" of 53 million shillings. Coast region, however, received less h... itbs "ceili,r ht ther. the eynThnation for this is that in early 1974 Dar es Salaam was carved out as a separate regilor. by itself,4 --n *__ so ->a th ios regio was left ^,nh o-l fo-llr I. &~AJ W.J A. uO _..L. 0'. so I1 tlua UL1 k UCLJO J I US.LU'JA IVQ .L ~-. ~ ----- districts instead of the original five for which the "ceiling" amount 1-. IhCLA b'eer. prescILLb eUd e2.15 As fa sh Le fr---, exper.-c4-m 4 s con.cert-d Tables 2.6 to 2.8 reveal that both Dodoma and Arusha (Coast regions figures are not strictly comparable) received -substantial 4-creases $ni their allocations for 1974/75 over the previous year. Dodoma's recurrent allocations are something like 30% higher tnan last year's whElile ushLa are more than 20% higher. - 33 - Table 2.6: TANZANIA: DWMA REGION RECURARENT EXPENADITURE nEQUESTS AND BUuGET ALLOCATIONS (in million Shs.) 1972/73 1973/74 1974/75 Budget Actual Approved Requested % Allocation i I. Administration 3.1 5.8 9.1 12.1 7.4 13.5 II. Directly Productive Agriculture: 3.3 4.5 6.1 8.1 6.7 12.2 Agriculture - Administration ) 0.4 0.7 1.4 1.1 A.ricult.re - Productive Services 0 .1 0. 3 0.2 Agriculture - Crop Production J 1.9 2.3 4.2 3.4 Ujamaa and Cooperatives 0.9 1.4 0.2 2.0 Natural Resources: 1.1 1.0 2.9 3.8 1.2 2.2 Natural Resources - Games Operations 0.3 0.3 0.8 0.3 Naturn. l Resourcs - Fishere 0.3 U.2 0.2 0U. Natural Resources - Forestry Operations J 0.5 0.5 1.9 0.7 Commerce and Industry Neg. 0.1 0.2 0.3 0.1 0.2 Sub-Total 4.4 5.6 9.2 12.2 8.0 14.6 Education: 11.5 14.3 19.0 25.3 14.1 25.7 Education - Administration 1 0.3 0.8 0.8 0.7 Education - Primary 9.9 11.9 15.8 12.1 Education - Adult 1.2 1.5 1.9 1.0 Culture o.; o.i u.s 0.3 Health 5.6 6.4 16.3 21.7 11.7 21.3 Health - Curative Services j 4.7 4.0 8.5 5.8 Health - Preventive Services | 0.3 0.3 1.5 1.4 Health - Rural Health Centers 0.6 1.2 2.7 1.8 HeAl-th - Dispe ...and Clinics N.A. 0.9 3.6 2.7 Water Supply: 6.0 4.0 - 9.2 12.2 6.5 11.8 Water - Rural Water Supply 5.2 3.7 7.8 5.4 Water - Urban Water Supply 0.8 0.3 1.4 1.1 Sub-Total 23.1 24.5 44.5 59.2 32.3 58.8 IV. Ecor.omic Ifrfastruct.re Comworks: 5.9 5.1 10.1 13.4 6.4 11.6 Comworks - Roads Maintenance 2.9 2.0 3.7 2.6 Comworks - Buildings and other Establishmenrt,l 3.0 3.1 6.4 3.8 Lands: 0.5 1.2 2.3 3.1 0.8 1.5 Lands and Urban Development j 0.3 1.0 1.8 0.5 Surveys and Mappings 0.2 0.2 0.5 0.3 Sub-Total 6.4 6.3 12.4 16.5 7.2 13.1 TOTAL 37.0 42.2 75.2 ino.n Si 9 10nn0n Source: Office of Dodoma Regional Development Director. Table 2.7: TANZANIA: ARUSHA REGION RECURRENT EXPENDITURE REQUESTS AND BUDGET ALLOCATIONS (in million Sha.) 1974/75 Budget Approved Requested % Allocation Z I. Administration 7.7 10.2 16.3 9.2 15.4 IT. Directly Productive Agriculture: 8.3 7.5 12.0 7.5 12.5 Agriculture - Administration i.3 1.3 1.3 Agriculture - Productive Services 4.2 4.3 4.3 Agriculture - Crop Production J 0.3 0.3 0.3 t.amea and Cooperatives 2.5 1.6 1.6 Natural Resources: 2.3 3.1 5.0 3.0 5.0 Natural Resources - Games Operations i 1.1 1.7 i.7 Natural Resources - Fisheries Operation 0.3 0.3 0.3 Natural Resources - Forestry Operations J 0.9 1.0 1.0 Commerce and Industry 0.1 0.1 0.1 0.1 0.2 Sub-Total 10.7 10.7 17.1 10.6 17.7 III. Social Infrastructure Education: 11.8 15.4 24.7 14.1 23.6 Education. - Administration 1.0 0.9 0.9 Education - Primary [ 9.3 12.4 12.0 Education - Adult 1.4 1.7 0.8 Culture J 0.1 0.4 0.4 Health: 7.8 14.6 23.4 14.2 23.6 Health - Curtive Services 5-5 9.2 8.9 Health - Preventive Services 0.4 1.9 1.9 Health - Rural Health Centers 1.0 1.9 1.8 Health - Dispensaries, Clinics 0.9 1.6 1.6 Water Supply: 3.9 3.8 6.1 3.8 6.4 Water - Rural Water Supply 2.7 2=9 2.9 Water - Urban Water Supply 1.2 0.9 - 0.9 Sub-Total 23.5 33.8 54.2 32.1 53.6 IV. Economic Infrastructure Comworks: 6.4 6.7 10.7 6.8 11.5 Comworks - Roads Maintenance 3.2 3.9 3.9 Comworks - Buildings and Other Establishments 3.2 2.8 2.9 Lands: 1.0 1 0 1.7 1.1 1.8 Lands and Urban Development 0.6 0.5 0.6 Surveys and Mappings __0.4 u_ 0.5 Sub-Total 7.4 7.7 12.4 7.9 13.3 TOTAL 49.3 62.4 100.0 59.8 100.0 Source: Office of Arusha Regional Development Director. - 35 - Table 2.8: TANZANIA: COAST REGION RECURRENT EXPENDITURE REQUESTS AND BUDGET ALLOCATIONS, 1974/75 (in million She.) 1972/73 1973/74 1974/75 Budget Actual Approved Requested I Allocation I I. Administration 4.0 7.8 7.9 11.-6 5.8 15.1 II. Directly Productive Agriculture: 2. 4.4 5.9 A,7 4,2 11.0 Agriculture - Administration 0.3 0.6 1.2 0.5 Agriculture - Productive Services 2.2 1.7 1.7 1.0 Agriculture - Crop Production J 0.1 0.5 9.0 0.7 Ujamaa and Cooperatives N.A. 1.6 2.1 2.0 Natural Resources: 2.2 2.9 4.6 ±& 3.1 8.0 Natural Resources - Games Operations 1.2 1.4 1.9 1.4 Natural Resources - Fisheries Operations 0.6 0.6 1.4 0.9 Natural Resources - Forestry Operations J 0.4 0.9 1.3 0.8 Commerce and Industry 0.1 0.2 0.4 0.6 0.1 0.3 Sub-Total 4.9 7.5 10.9 16.1 7.4 19.3 III. Social Infrastructure EducatIon- 1.6 7.8 19.2 28.3 8.2 21-4 Education - Administration J 0.7 0.7 0.8 0.7 Education - Primary N.A. 6.6 7.5 6.7 Education - Adult 0.9 0.5 7.8 0.5 Culture J Neg. Neg. 3.1 0.3 _Heaith 2.6 5.7 5.7 8.4 6.9 18.0 Health - Curative Services 1 2.1 4.0 4.2 4.5 Health - Preventive Services 0.2 0.5 0.8 0.5 Health - Hural Health Centers 0.3 0.3 0.4 0.5 Health - Dispensaries. Clinics J N.A. 0.9 0.3 1.4 Water Supply 2.3 2.5 3.0 4.4 2.3 6.0 Water - Rural Water Supply 2.3 1.9 2.6 1.9 Water - Urban Water Supply N.A. 0.6 0.4 u.4 _ Sub-Total 6.5 16.0 27.9 41.1 17.4 45.4 IV. Economic -frw-trture Comworks: 14.0 13.6 19.7 29.1 6.5 17.0 Comworks - Roads Maintenance 4.4 5.2 11.1 5.1 Conworka - Buildings and Other Establishmentj 9.6 8.4 8.6 1.4 Lands: 1.0 1.4 1.4 2.1 1.1 3.2 Lands and Urban Development | I 0.4 - 0.5 1.0 0.3 Surveys and Mappings J 0.6 0.9 LA - _ Sub-Total 15.0 15.0 21.1 31.2 7.7 20.2 TOTAL 30.4 46.3* 67.8* 100.0 38.3** 100.0 For five districts including Dar-En-Salai. ** For four districts excluding Dar-Es-Salaam. Source: Office of the Coast Regional Development Director - 36 - A2.6 What i ii portant iS that none of t'h regions really got c at it asked for. Dodoma, for example, asked for 75 million shillings but got. --'I- 55C r.4'11ion w..l rsA n4OsistA thn+ 62 .?.4114ion shill-gsn" was its "rock bottom" reqaest after being reduced from 72 million shillings (I.t.s aLruali -,fge `1 I a~'so is- 42-I. r,4 11ior- Mhl4"-g) I.. goi --- -- -0r U tL OLIU A YC.L UC LtLJ. .~~~. L& fl'+ tL.'l 1Sf--Lt.rS O %I LJU V million shillings. For its original five districts Coast region asked ior_ 68Q .,'li on shli but,4 w-1A prashv,gotten no m-ore 4-9ar IUUl UU IIL_L.._LL± 4"I L.L.L1IO LULAV WW'JLL " rWJ ~ ~U U~L1L4 U& LA± 55 million shillings. 2.1.7 While the Center effected cuts in recurrent expenditures dJl arouindU ±U WC-L ±s flt Lby VLheV .5. LL)AO VL=U that ` O".Ler cu;tLUs -U ere ruaude in personnel emoluments and social service expenditures and relatively n ___ _s~ __ _A :~_:J __ _A n __4_r-3_ _X - _ 1/ 'Larger cuUsi were z adue i5m1 aitnuice cuii LLULO.lacemenu.ofuV ±e 1i1,U1 Vi ili.l maintenance and other expenditures for buildings as well as transport aIi-Lwarices for ULIe employees. 2.1v The three regiLuns vIsLted by the mi8ssin all complainea ol the inadequacy of budgetary funds and insisted that, as most of their U ]Ur Wev 1 _* X_g I _ v _ Uey A t _ L n Ve T, recurrent e:xpenditure Ccoumiments we rigid, they WoujA have to seeK "supplementary" funds during the course of the fiscal year. After all, they hrad been doing this sInce decentrailzation. During 1972/73, for example, Arusha sought 28 million shillings supplementary and for 1°73/74 it estimated this figure to be 40 million sni lings. Similarly, Dodoma got a supplementary of 6 million shillings for 1973/74. 2.19 The sectoral allocations in their recurrent budgets were, nowever, not signlficantly affected by the severe cuts in their desired levels. As Tables 2.6 to 2.8 suggest, the patterns of recurrent ex- pendi-ture finally received by all of them have been very similar to what they originally requested. As it stands now, almost 55% of their recurrent budget is spent on "social infrastructure", about 1;% on "acrinistration", another 15% on "economic infrastructure,, and another 15% on "directly productive" sectors. 2.20 The explanation for this, of course, is very simple. It has been estimated that about 60% to two-thirds of their recurrent budgets are spent on "personal emoluments" and another 25% for the maintenance and rmning of plants, vehicles and establishments. Both these facts obviously impart a certain degree of inflexibility to the level and pattern of regional recurrent expenditures which neither the regions nor the Center can manipulate in the short run. I/ And this despite the fact that more than 50% of all Government owned vehicles have been found to be unserviceable for use at any time compared with 10%, which proportion is regarded as "acceptable". - 37 - Conclusions 2.21 To sum up: the little evidence which was available to the mission seems to suggest that since decentralization the regions have not gotten what they requested either with respect to the level and pattern of development expenditure or the level of recurrent expenditure. If anything the Center still seems to influence all these decisions very substantially. The patterns of their recurrent expenditures do match with their original requests but then this might well be because of the rigidities inherent in these expenditures. 2.22 On the whole, then, it appears that the democratization of expenditure decisions has yet to take place in Tanzania and perhaps it is too early to expect this to happen anyway. This is not to deny the larger discretion that regions and districts have been granted since decentralizatian in implementing and executing their budgets. An assessment of this forms the sub iect matter of the next Chanter. CHAPTER III. FINANCIAL AUTONOMY UNDER DEC1WNTRA LIZATION AND SETF-HET.P TN DEVE:TflPW-. 3.01 Tn theorv, the financial autonomv of the lower level of government is determined by, firstly, the degree of independence it enjoys in raising and spending monev! and. secondly. the disrrpetln it has in implementing and altering, if so needed, its financial plans. 3.02 Chapter I which described the financial arrangements under decentralization indinated that the lower level (regffional nnd strict) administrations in Tanzania have no independent taxing powers whatsoever. Nor do they have any horrovnw ng rn p-sf their- r%m, This muQn that at least in one sense the financial autonomy of the present lower level administrationsnl/ has heen nmtailed vis-a-his their predecessors, or, the former local governments. 3.0i Chapter II then showed that the Regional and District Administa- tAons were not entirely free to determ;e their level of spendig either. They seem to be constrained and regulated by the Center (Prime Minister's Offic-e and the Treamsury) in determin4ng their level anrd the pattern of+ recurrent and development spending. 1/ The question of 'tfinancial autonomy" of the District Development ,wa.s iove.-ed 4- ..4sL rULd4V±. - 38 - 3.04 A closer examination of the financial arrangements under decentralization reveals that the little autonouW that has been given to them relates: (a) To initially proposing what the composition of their budgets would look like; but, (b) More importantly, in implementing the budget once it has been approved in the National Assembly. It is these aspects that the present Chapter highlights. &xoenditure Discretion j.05 An important aspect of decentralization relates to the amount of financial discretion which the lower level administrations now enjoy. 3.06 Whereas before decentralization the local governments (for many of their functions) and regionally located Central officers (for all their functions) had to get the sanction from Dar es Salaam for each and every procurement they wanted to make or any funds they wanted to disburse, this situation stands changed now. The Regional and District governments possess their own sDending authority and have been given some, though a still limited, amount of discretion in carrying it out. 3.07 Once a region's budget is approved by the National Assembly, its officials are not reauired any more to seek day-to-day authorization from the Center to undertake particular expenditures. In fact, regions receive quarterly releases for their recurrent as well as development buugets from the Center more or less automatically and the only requirement here is that thev must regularly send the quarterly financial and other progress reports to the Center. Not only that, should a region require m_re than its quarterly allocation in a particular quarter it can easily make a special request which is normally granted. All these make for some degree of flexibility in budget implementation at the regional level. The same kind of financial relationships seems to exist between District A4m4n-strat±ons vis-2-vis their Regional Administrations. 3.08 Atlthough the gen-eral budgetary rule that regional funds must be used for the purposes for which they were originally approved by the National Assembly still applies, the Regional Administrations have also been given some discretion to deviate from their approved budgets. 3.09 With regard to their recurrent budgets, the Regional and ELstrict governments have the authority to reallocate recurrent funds from one sub-head to another under the same sub-vote in cases of "surpluses" and "sL-orUt-f&_ s. .Ln the p-ast, however, they. cold not a-fec-t this reallocation between "Personal Emoluments"t, "Replacements" of office furniture, equipment, vehicles, etc., "Special Expenditures", and "Sub- - 39 - ventions, Grants, Contributions and Subsidies" unless they had a written authorization from the Minister of Finance. This placed severe limits on the expenditure discretion of Regional Administrations and, as a matter of fact, slowed down the regional activities many times. 3.10 Since 1973/74, this reallocation rule has been modified slightly. Now regional accounting officers have been given authority to reallocate funds between "Replacements" and "Special Expenditures" provided these reallocations do not exceed Sh 20,000 in any one financial year. No reallocation is still permitted between "Personal Emoluments" and "Sub- ventions, etc.", and this largely to avoid the possible abuse of financial powers and the creation of unwanted jobs or the granting of unwarranted subsidies by Regional Administrators. 3.11 As far as the development budget is concerned, here also the Regional and District Administrations have been given limited discretion. They can now utilize the "savings" from one project or program for spending on another project or program contained in the budget or the nlan of the region nrovided prior consent is sought from the Center. Even a new project or program, not included in the regional plan, may be undertaken through such "savinps" nrnvided the annroval is received from the Prime Minister's Office, Dev Plan and the Economic Comittee of the GahbnAt.= The latt.Ar renuiire-ment is eonsidArpd bv manv as "nominal" because, in reality, such approval is quite easily given. 3.12 On the whole, then, it seems fair to say that the decentralization of admin1Tratn+on has hben acompniedou by limid4+ted vœmarrt;.rn discreon, and the word "limited" must be emphasized. Some important changes hav, h4-Anera,4 +aken pac sir.ce Aaar,-.t4zno44+ion 4in ha++vter rlt;ng to tendering and procurement procedures to which the next sections are Tende.r Pr-oced-es 3.1 T'ho RPgiorsvtal (n-AA Dlstri,ct) A +43r-l AAd-4ni 424a-r aA ir. seeking their own tenders has been extended considerably since 1972/73 and this has significantJ+'y reduced the delays .rin car-ying out po-,ects and programs and improved the qeed of government spending. 3.14 It might be pointed out here that Regional (and Area) Tender Board ztdats eir'haqas e-S- ever befreth 1072 decentralization of administration but they carried so little authority that they exsted "re 4n name thar. in reality. 'IAy' were not pe ,1tV' to seek and award tenders on their own for new construction or jobs e-n-eed-i4n Sh 1000 value 4Ti --I - U -Mes that f - most _- their re- LI*f i J..Iw.L .LLA.~L UAL&4J,L5 ULA~ U 4L'J.L LUUJOL % J.L ULILeiL re quirements, local authorities had to go through the Central Tender Board (CTB.LJ/U) to %N% se LeLrV.dr. rJistries .Lf.LV0 D. eJ.ence aLdLL EjuLLcavio hL , andIU still have, their own independent tender boards. Since the decentralization oA *e an U_ea.dr+ _j_wetn pu-hority pofLA thi owr.. 3.23 Prior to decer.tral_zati on, the major procur.ement problems related to inadequate funding. The 21 million shillings Revolving Fund which finances Gartral supplies has remained *"hre s4ince io6L. * o.e^e. there has been a constant overdraft of some 30 million shillings per anu ,m . Other procuremrent problems before decer.tralizaton -cluded inadequacy of professional staff capable of doing inventory planning and tech>ntcal C-ecifcation. Of mate 4al purchases, the lengtUhy Prouess, taking sometimes six to eight months, to go through foreign exchange acquisttioni nd get customs qrran.ce, shipping and harbor co-qestion nd the lack of storage capacity._ 1/ Because of all these problems one estimate puts the government stOres out-of-stock i'tems at an "average !! Of 21%, far too nigh by any standard. - 42 - 3.2-1, The Aecen+xaz1ti c of administration seems to have accentuated mary of these problems even further. Inadequacy of staff at the regional level has creat.ed Aiffimi1+tPe with resnect to budgeting and planning for their material requirements. They have failed to appropriately t-ime their 7nqeral procruremPents for each nrolect and have sometimes forwarded wrong material plans thereby locking up capital. The storage capacitv in the regions was highly inadequate even earlier; this has been made worse with decentralization and the expanded activities of the Regloa LA-m i+st+4i^ns. The liri+.otl arnilah1 nrocnrement staff had also to be spread too thinly with decentralization with the result that pMcUrM1ert _n^1-+cton is now be4ig perfo-rmmed by manV untrained officers. The Government is, of course, very conscious of these problems and has st aed A .t,-4g aL var-us solurt s Oef of r e+hsce iwjhi-h J.q heiina LIOJ LA,4 ULLIL..f..LL OLvw.Av U. V~A .LJUW ~ ~~S given serious consideration is that of strengthening the Regional Trading Com,'Lp.es in order t4o LLCLLL" tV&he gr.er 'I -eqir-.e.sM- tE. egoa and local administrations. Hopefully this will speed up deliveries and w-wi. avu'du tlhe nee-dA fo.L- Whe dup'J'cionofJ.%J1 U sLtaf 'A' t s- facilities. 3.25 To sum up: since the decentalization of administration, the -lower level governmenLts hla.ve beer.Al giV 1.LLJ'...L larger authorit over their own expenditures. However, they still are subjected to a high degree of financial control by the CerntLer. They have to get many of the "deviations" from their authorized budgets "approved" by the Treasury or the Prime Minister's Office and somet-imes ev-en the Econom-,,ic Committee of the Cabinet. Their authority over tendering has certainly expanded and this is considered a great improvement. The situation.- with respect to the coordination for procurement has certainly become more difficult with the decentralization of administration. 3.26 Wnat impact this limited financial autonomry has on the iumplementa- tion ol development budget is hard to say. The Government still does not have the audited figures on actual development expenditures incurrued by different regions since 1972/73.TAs such, therefore, a major objective of decentralization, i.e., the improvement in tne speed oI spending is hard to quantify. However, by talking to the regional officials one gets the impression that it has improved. As an illustratim, a poor region like Dodoma sought a development "supplementary" of 11 million shillings in 1973/74 in addition to its "original" estimate of 8 million shillings. Also, the Prime Minister's Office has provisionally estimated that for 1973/74, regional development expenditure would be about 225 million shillings against the original budget estimate of 177 million shillings. They would indeed be a great improvement over the past. Self-Help Contributions 3.27 At the time of introducing decentralized administration it was anticipated that it would improve local participation in development projects and would encourage self-help contributions towards development effort. 3.28 Prima facie, the self-help contribution should have improved if Lor no other reason because the rate OI- development spending by the regions seems to have improved following the decentm lization and in the regional development budgets in Tanzania only funds for material costs and hired (skilled) labor inputs for each project are included. For example, it is estimated that the full cost of building a classroom is 20,000 shillings but then as 50% of it is in the form of unskilled labor it should be mobilized voluntary. Consequently, the budget estimate of a classroom would be only 10,000 shillings. In Tanzania this costing procedure is adopted for practically all development projects which means that "voluntary" labor is built into the development estimates as a com- plementary input. It is the policy in Tanzania not to finance unskilled labor unless there is a reason to assume, as in the case of education projects in the Masai areas, that public participation will definitely not be forthcoming. It would seem, then, that higher rate of regional development spending alone should be an evidence of higher participation rate by local people in the development effort. 3.29 However, discussions with the Regional Administrators slagest that the conclusion that the local people are now more eager to contribute labor cannot be taken for granted. Whereas in a poor region like Dodoma, it was felt that people's contribution to development effort has improved after decentralization. and a huge water supply project carried out through voluntary labor was given as an example' the administrators OI a rich region like Arusha thought that there was actualy1 a deterioration in self-help contribution after decentralization. Nevertheless, all Regional Administrators felt that it is easier to get voluntary labor for water supply, health and classroom projectsl7 or the projects which benefitted people directly. On the other hand, people have been found to be very hesitant to contribute voluntary labor for building teachers' houses or maintenance of roads or construction of godovns and such 4n= frastructure. As generally voluntary labor is a necessary ingredient of the nlan for all the latter projects, amiisrators have had to foc labor out of local people for these projects through the use of social pressures and sometimes even hy threatening people with a "fi-e" under some traditional rules which require all adult males to work a given day of the week on covmmanal proiects. The compulsion, a.d the threat used to seek "self-help" in the implementation of the latter kinds of prolects have sometimes created problem.s w-hn "self-help" is needed for other "desirable" projects. l/ hven though sometimes it has proved to be inefficient and independable and could not be+ 4t4egrated -wel into a tgnt construction schedule for a project. 3.30 Some field researches undertaken in Tanzania have found that 1j sJo ntlaneous" self.-help bas n.ot ---Ae -;+,ficar.t' ..* VS4 'eenL'. za- tion excepting in regions which had a strong tradition of voluntary lao.rL Uco1b.UtU.LgnLull Vv I -l- W±Jit or -"IL1U developmer,t bu.gets co,,- very largely of social infrastructure projects in the rural areas. £'1WJs' v r-gi.oL p±d-U-LuJ1 U±officers h-ave b .eenUrouUd to eL.U p h-asX ze t.Ld.a,t eA .g self-help for development projects is their major headache particularly for the projects w-hlch hiave not been "generated by the peop'e andu th includes majority of "economic infrastructure" and "directly productive" projects. CHAPTER IV. IN'IER-REGIONAL EQ=UY IN SOCIAL SERVICES 4.01 Since the Arusha Declaration of 1967, Tanzania has set before itself the long-term goals of socialism and self-reliance. One major element in the achievement of this goal, and this was adequately stressed in the Second Plan, is that of "social equality" implyirg an even spreading of benefits of development amongst the masses and redressing of the regional imbalances. In 1971, TANU further directed the Government to give top priority in the allocation of resources to three essential social services -- health, education and water supplies. The Government has, therefore, set up long-term national targets for these services in quantitative terms. The purpose of this Chapter is to highlight: (a) the regional imbalances which exist in the provision of social services in Tanzania; (b) the fiscal feasibility of achieving the quantitative national targets which have been laid dovn by planners; and (c) the extent to which the Center is attenpting to modify the short-run allocation of resources in attaining an inter- regional equity in social services. Inter-Regional Imbalances 4.02 Table 4.1 presents the data on the crucial socio-economic indi- cators for 18 regions (excluding thc two regions of Dar es Salaam and Rukuwa which were formed on January 1, 1974). The differences in the availability of education, health and water supplies from one region to another are evident. In matters of primary education, Ruvuma, West Lake, Tpnga, Tabora and Kilimanjaro seem fairly well ahead of others with 60% or more enrollment ratio while Iringa, Mbeya, Mtwara and Shinyanga severely lag behind with 40% or less enrollment ratio. 4.03 The health services are also veryv unevenly distributed. Much of the nation's health services seem to be concentrated in Dar es Salaam. The city has only 3% of the country's population but still has 7% of - 45 - Table 4.1: TANZANIA: REGIONAL SOCIO-ECONOMIC INDICATORS, (FORMER) EIGHTEEN REGIONS AROUND 1973/74 Proportion Primary Population of Educatior. Populatior. per Population Enrollment per Population Rural Served by Feeder All Roads Area Population Ratio Hospital per Health Water Roads Roads per Km Region ('000 Km2) ( 000) (Standard I)T Ad Dispensary* Center* S,upply (in K.) (in K.) Area 1. Arusha 82.0 743.8 52 650 10,780 148,760 28 1,910 2,830 .0357 2. Coast (including Dar-Es-Salaam) 33.7 996.7 42 1,700 6,000 183,950 12 1,558 2,478 .0734 3. Dodoma 41.3 809.7 47 1,200 8,890 96,660 38 2,087 2,952 .0714 4. Iringa 56.9 827.3 38 850 9,200 420;000 8 1,398 2.649 .0465 5. Kigoma 37.0 516.1 53 1,030 6,840 125,000 20 399 9,590 .0427 6. Kilimanjaro 13.2 791.0 60 640 11,730 113,000 30 974 1,894 .1429 7. Lindi 65.6 464.2 50 770 6,630 114,000 18 1,435 2,214 .0337 8. Mara 21.8 666.4 48 1,460 9,340 128,860 10 765 1,219 .0560 9. Mbeya 90.1 1,143.3 40 1,060 10,210 228,660 10 1,780 3,319 .0368 1-. Horogoro 73.0 768.6 52 670 6,600 97,880 15 1,007 2,368 .0324 11. N4wara 16.7 731.0 40 830 6,630 125,800 15 1,326 1,917 .1147 12. Mwanza 19.7 1,232.3 44 1,330 11,010 100,040 10 1,821 2,493 .1221 13. Ruvuma 61.3 461-3 73 440 15,000 92,250 19 1,478 2,178 .0355 14. Shinyanga 50.8 1,043.8 40 1,800 9,800 110,000 6 1,964 2,522 .0498 15. Singida 49.3 496.3 62 900 7,690 100,000 17 1,466 2,082 .0421 16. Tabora 12a2 635.2 66 720 8.360 127,000 35 2,359 3,668 .0300 17. Tanga 16.8 899.8 60 670 7,820 99,970 17 1,470 2,826 .1054 18. West Lake 28.7 739.1 68 560 7,800 120,000 25 848 2,099 .0730 * Rounded to nearest zero. Source: Prime Minister's Office 1.j4 - L46 - hospital beds, 38% of government doctors and 35% o0 Grade A nurses. Trt 1070,/71, M.i"m.bl .op+lreevd,4eoertrgfnsta .1-* --/ I Ir.., ~ ~ ~ I. F. v- U 1A IA'. V LkIJV.L 0. LI-LLA L~ .1. U D. ULId.UI 10 regions combined. Of course, She Dar hospital does act as a teaching hospi+t' tn. ha a national -4pi"- =oTv er, e f f-ct4-.In ---eea ----vwa West L -1 Tanga, Kilimanjaro, once again, have largest number of hospital beds per capita, 'L1tj3,k Sh04 nga, TDoAma, ,Yara- --A a..d -d& A. LL4B a r v long way to go in this respect. The same is also true of the availability of dispensaries or r-ual health centers anxngsb regLons. h.oh In terms of water supply, Shin-yanga is the poorest region with only 6% of its population served by tap water and Iringa 8% while Tabora has 35% o its popuIlation enjoyimg this service and Kilimanjaro 30%.. Dodoma leads the nation in tap water supply but then it lags behind in ever-y oth.er service. Nation-l Targets 4.05 The Government has set up the following national goals with respect to essential services which it wants to achieve by around 1990,1/ Health: (1) One rural health center per 50,000 people; (2) One dispensary per 10,000 people; (3) One hospital per administrative district and one bed per 1,000 people; and (4) One medical doctor per 22,500 people. Primary Education: Universal primary education by 1989.Z./ Water Supply: Easy access to adequate and safe (tap) water for all population by 1990. 4.o6 The present (mid-1974) population of Tanzania is estimated at about 14.3 million and assuming it keeps on growing at 2.7% per annum, the population to be served by the above mentioned services in 1990 would be around 21 million. 1/ Taken from various publications of the Government of Tanzania and the "Plan Guidelines, 1974/75". In addition to the targets mentioned in the text, there is a national target of providing all-weather feeder road access to all uilamaa villagpe of over 1,000 people by 1980. 2/ This was the initial target date which the Government has now changed to 1977. The calculations given below are based on the "initial" and not the "revised" target date. '.7 4.07 The achievement of the national targets would then require that by 1990 there should be 420 rural hel th centers, 2,100 dspensaries, 73 (which is the present number of districts in Tanzania) district hospitals (with 21,000 beds), and about 1,000 medicl docvors. 4.08 As far as (primary) educatiop is con-cern-.ed, in 1971/72 primary school (Standard V-VIII) enrollment was 982,00021 which is about 37% of the relevant age group o01 2.65 million (i9.5% of the totLal pQpul.ai.oUn. L±L L no major changes in the proportions of (primary) school age children, in the total population, the total number oI schooi piaces required by 1990 to achieve universal primary education would be approximately 4.1 million. This means that new classrooms would have to be built for some 3.1 million children between now and 1990. On the assumption that each class- room will hold 45 students and that one teacher will be abie to handle two classes,_/ there will be a need for approximately 69,000 additional classrooms, 35,000 additional teachers and the same number of teacher houses. 4.09 Presently about 11.5%Y of the rural population, or about 1 million rural people, are served with adequate and wholesome water. Assuming that by 1990, thp ratio of urban population to total population grows from the present 8%!V to 20%, about 17 million people would still be living in the rural areas. This means that about 16 million people in the rural sector would need to be provided with rural water supply by 1990. Fiscal Implications of National Targets 4.10 To get an idea of whether or not these national targets can be financed with the available fiscal resources, ignoring for the present the serious problems of finding adequate numbers of engineers 1/ Economic Survey, 1971/72, p. 130. 2/ This might be a generous assumption. 3/ Economic Survey, 1971/72, p. 2. 4/ The estimated populaticn of 10 large towns in Tanzania is less than 1 million for 1974. Second Five Year Plan, Vol. I, p. 182. - 48 - ar.d +e-hnbicrians to imonlpment them.i/ Dro.ections ot likely revenues oi the Government by 1990 must first be made. Assuming that the government revenues grow at the same rate as GDP during 1960-1971, i.e., about 5% per annum,27 the present tax revenues Of 3,000 million shillings would reach a level of 6;600 million shillings (at constant 1974 prices). This means that the Government would have about 27,000 million shillings additionally {at cntnnt prices) over the next 15 years to spend. 4.11 Tnzanina Government has develoned over time what it calls the standard (capital only) cost estimates for different projects. These are based on certain asormptions regarding the self-help labor in the construction of these projects. Due to rapid inflation many of these estimates hnvi b a ltteV out. not date. Table h.2 presents the standard cost estimates as well as "more realistic" cost estimates provided bmy the regi-on-r -authorities . 4. 12 T, ,e ^ costs (at rcnstant nrices) of achievin2 the LI. LC - i -- - - - . .C.3. f -- aLLJV.A IA' J'Jb- -- -- 1 national targets are then estimated in Table 4.3. 4.15 It appears, then, that the construction of rural water works COLonAe accorA--- 4o the ,-nat+iorl ^-n 7 tar&et. yrld absorb niit 8% of' the dJLJLI~ A A. " J 6 IA.J U1 i. & - -1 - ,-- --- - __ __ - ___- accumulated increase in revenues. The construction costs of health facilities, prima r education and rural water facilities all tJtogether would just about absorb some_15 a of accumulated incremental revenues, leaving a surplus or -meeting the CostS of fiurnishing and rminnng these facilities. This is projected without making any allowance for the expa sion of any other activitvy or attemptIn in. to rer the unevenness in the distribution of other services between districts and regions. 4.14 What the recurrent expenditure implications of the heavy investment outlays i-ul 4ici 4in _ar,ets mn be i"rJ t__ now. The rate of growth of recurrent expenditure resulting from the achievemaent of national targets w-ol ddepend uponr: (a) the "co-efficient" of recurrent (normal running and maintenance) expenditure consequent 1/ An official report (of September 1973) points out that only se-ven of the ~18 r-1n h-ad at 4t.-4t ti.'.e qualiiedRegona ,SVL101 UIJLL -± .L-~Ut iIV LICLU 0.1. VIJ 16.L W klJJ.. Ii L- - Engineers and that most regions lacked the required number of inspectors and foremen. 2/ The GDP growth rate was 4j% durng 1960-1967 Uhile it is esti,ated to have grown at 5.4% per annum during 1964-1971 (at constant i966 prices). Taole 4.2: Tc%NZANIA: COST ESTIMATES OF CAPITAL PROJECTS Standard Cost Estimate Actulal Costs (in Shs. per project) 1. Rural Health Center 400,000 700,000 2. Dispensary 35,000 45,000 3. District Hospital (without self-help) 2,000,000 N.A. 4. Regional Hospital (without self-help) 4,000,000 N.A. 5. Investment per doctor 100,000 N.A. 6. Classroom 10,000 12,000 7. Teacher's House 12,000 20,000 8. Rural Water Supply Station 1,000,000 N.A.* 9. Water investment per capita 140** 200** * The actual cost of providing piped water supply has been estimated at 175,000 Shs. per village (without self-help) at the very minimum. ** According to the 1974/75 'Plan Guidelines' investment in rural water supply should not exceed 140 Shs. per capita and in no case 200 Shs. per capita. (These estimates closely resemble the estimates derived in a research paper by Dennis Warner, "The Economics of Rural Water Supp'y in Tanzania," Economic Research Bureau. University of Dar-es-Salaam.' Paper No. 70.19, pp. 16-17.) But then according to Economic Survey 1971-1972 (p. 96), a rural project serves on an average approximately 2,700 people, which will yield "standard cost" of a rural water supply station per capita of 40 Shs. only. Source: Mission Estimates. - 50 Table 4.3: TANZANIA: CAPITAL COST ESTIMATES OF NATIONAL TARGETS Actual Cost Present per Total Target Availability Unit Estimate (in Shs. per (in mill. Shs.) project) A. HeaiLth: 1. Rural Health Center 420 94 700,000* 230 2. Dispensaries 2,100 1,440 45,000 30 3. Hospitals 73 126 2,000,000** 40 4. medical Doctors 1,000 400 100,000 60 Sub-Total 360 B. Primary Education: 5. Classrooms 69,000*** - 12,000 828 6. Teacher's Houses 35,000*** - 20,000 700 Sub-Total 1,528 C. Rural Water Supply: 7. Population to be served 17,000,000 1,000,000 140 2.240 Sub-Total 2.240 GRAND TOTAL 4,128 * cf. William P Mayvr, "Tmnlempentation and -ontrol Under Decentralization: Tanzania's Rural Health Programme," A Research Paper of the Department of Pnlitical Science TTniversity of Dar-es-Salaam, June 1974, p. 45. ** on the assametion that 9a dist-icts s--t1 ned hos-itals. *** This is the 'net addition' estimate. S M ~4..'.JLL Estimates an' Ta I e 4 .2 - 51 - upon a shilling of investment in each of the rural water, health and educational facility and (b) the size ot' annual capital expenditure reouired to achieve the national tareets in relation to the existing level of recurrent expenditure. 4.15 It is estimated that the (annual)recurrent cost co-efficient of rul water schemes in TanzaLnia is ahout 0.20 of the origiiial eanital expenditure (in constant prices) while the corresponding co-efficients for rural health ar.A ducat4on are 0.Ln and O.cvO pt 1.fn compared with the experience of some other countries, these co-efficients do% ppear a little high bout then. this might be due to the fact that the project capital costs in Tanzania are somewhat lower than elsewhere due to the self-help coponent which is not paid for by t-he +3ver.-w.ent 4.16 TP +Ihe vnis-4onts ps+4vieS (give;wn on pags50 are sec-et-4, .+.- i#T4 4._ WI& 4,w" tW g,4wv~ e.. .. ~ Tanzania Government would have to invest an "average" of about 275 .. 1140n sh54ll4n6s per annr.s (J:. const-V pnices) for +he I,t 15yer MA.i.~ ~L A J' L AA_kJJ "L&UMA .Lk WC&Au 1J 'V J UL a&Jv L/" Y'.n in order to achieve its stated national targets. Of this 25 million s.;:'lIun-s wou.3d. go towards th-u&e crealt-ior. o f x-rq-z" ' 1 he - nth f5 tt 100 million shillings towards the provision of education and 150 million 314hL.L.gEs to-P.lW s r-ar.-LUL -wtoer scOujrues. r* 1 use of the co-effiCivers given above, the mission estimates that these investments would generate ne-w recurr-ent expenuditure co"niunent o' about 90 million sbLi 'ngs every year, in tVe aggregate totaling 10,800 million shillings over the next , , In aggregate -I 15 years.- (Thais represents t'ne sum of 9'0 + 180 + 270 . . + 1, million shillings.) This would mean that the present level of recurrent expenditure of 3,660 million shillings would expand to about i,41S60 million shillings (in constant prices) in 1990 due to additional recurrent expenditure co=mitments of tne national targets for health, education and water alone. 4.17 There is perhaps another way of looking at this. Even if Tanzania Government were to get all the money it needed to finance the capital costs of its national targets by way of outright grants from foreign donors and international agencies, the recurrent expenditure bill of 10,800 million shillings (in constant prices) generated by 1990 'would very substantially curtail the capacity of the Government to finance the existing and other (target unrelated) activities with its built- in increment of tax revenues growing to a level of 3,600 million shillings (in constant prices) during the same period. 1/ This amount may be compared with projected additional revenues of Sh2 7 billion (refer para. 4.10). -52- 4.18 In fiscal terms, then, the national targets of essential services seem unrealistic unless foreiLgn aId and non-tax revenues (including parastatal savings) expand at a much higher rate than can presently be envisaged. Equity-Oriented Ailocations 4.19 This, however, does not muean that the Goverment should not attempt to correct the inter-regional inequities or to shift the allocation of resources to the poorer regions; all that this implies is that perhnps the Government does not recognize what it has committed itself to. 4.20 As pointed out earlier, the Government has decided to adopt the goal of inter-regional equity in essential services and prvi-ding these services free and hopes that the abolition of the local tax systems and local budgets and replacement of tnese by the nationall1y approved budgets of the decentralized administrations would help it to achieve this goal. 4.21 One major limitation of regional budgets in the puru,it of this objective should be highlighted right away. As Table 4.4 shows, regional budgets still do not amount to more than one-fifth of the national total (Central and Regional combined) budget. At present, about 27% of Government's recurrent budget and 13% of its development budget pass through regional channels. These proportions would certainly be much smaller if the resource flow through the parastatal and rest of the (non-government) public sector could be taken into account. 4.22 Consider now the extent to which these limited budgets are being used to correct regional imbalances since decentralization. The manner in which the Regional Development Fund (RDF) is allocated was described in Chapter I. Table 4.5 now gives the data on development ex- penditures per capita and total (recurrent and development) expenditures per capita and compares them with per capita incomes of the regions. The evidence is fairly mixed. Whereas, on the one hand, richer regions like Arusha, Dar es Salaam and Kilimanjaro continue to receive higher than "average" development (as well as recurrent) funds per capita, some poorer regions like Ruvuma, Tabora, Singida and Lindi have also been allocated substantially higher than "average" resources. One gets the feeling, then, that a serious attempt is being made to break away from the past in this respect and larger national resources are being diverted towards poorer regions. This is clearly a result of the deliberate policy of the Government. 4.23 Not only are larger sums being allocated to poorer regions, attempt is also being made, though with somewhat lesser success at this stage, to reduce the inequality of social services between regions. Consider Table 4.6 which shows the shares of (regional) sectoral allo- cations received by each region and the level of existing public services - 53 - Table 4.4: TAIZmt.t.: SHARE or REJGIONAL BUDGEL S IN oaE /^TION-1L BUDGET, 1972/73-L974/75 (in million Sh_.) Region Recurrent Revenues Recurrent Expenditures Development Expenditures Total Expenditures 1972/73 1973/74 1974/75 1972/73 1973174 1974/75 1972173 1973/74 1974/75 1972/73 1973/74 1974/75 (A) (A) (B) (PA) (A) (B) (PA) (A) (B) (PA) (A) (B) 1. Aruska 6.2 5.7 6.9 37.5 49.3 59.8 5.2 14.9 20.9 32.7 64.2 80.7 2. Coast 36.0 14.3 0.9 40.1 47.0 38.3 4.9 9.6 11.7 45.0 56.6 50.0 3. Dodoma 6.4 4.0 3.7 46.8 43.7 54.9 12.4 9.1 18.1 59.2 52.8 73.0 4. Iringa 5.1 2.9 2.0 38.4 37.2 42.4 3.7 11.0 16.8 42.1 48.2 59.2 5. Kigoma 3.2 2.5 1.6 22.6 28.2 38.2 4.1 5.8 10.1 26.7 34.0 48.3 6. Kilimanjaro 7.9 5.4 7.1 41.7 45.6 63.1 4.8 12.0 17.4 46.5 57.6 80.5 7. Lindi 3.0 3.2 2.9 26.6 32.1 37.4 2.6 10.8 15.2 29.2 42.9 52.6 8. Kara 4.5 3.6 2.i 32.3 31.3 36.9 5.2 14.0 12.5 3/.5 45.3 49.4 9. Mbeya 8.0 5.6 3.6 51.3 49.0 53.5 7.4 13.6 17.1 58.7 62.6 70.6 10. Morogoro 6.2 5.5 5.8 40.7 45.8 58.9 4.7 8.5 15.7 45.4 54.3 74.6 ii. Mtwara 4.4 2.7 2.60 J4.8 36.3 42.)5 J./ 12.) 17.j J 38.5 48.B 59.8 12. Mwanza 9.0 7.4 7.5 47.0 56.6 62.4 6.9 11.8 16.4 53.9 68.4 78.8 13. Ruvuma 4.0 2.6 1.3 34.9 33.4 41:7 3.9 9.9 11.4 38.8 43.3 53.1 ;4. Shinyanga 6.5 2.6 2.0 34.9 35.4 47.4 4.5 11.8 12.7 n 3.4 41.2 60.1 15. Singida 5.3 3.9 1.5 34.9 31.6 43.0 5.8 8.7 10.0 40.7 40.3 53.0 16. Tabora 6.8 5.1 2.9 39.7 40.2 44.0 3.6 7.6 13.3 43.3 47.8 57.3 17. Tanga 8.7 6.2 6.1 52.5 56.0 73.3 4.2 13.7 iS.2 26.7 69.7 9i.2 18. West Lake 4.2 2.1 1.7 43.6 44.5 49.2 4.9 9.4 10.2 48.5 53.9 59.4 19. Dar-es-Salaam - - 7.6 | 20.8 59.8 - - 23.1 | 20.8 82.9 --a. = - = ). - 9.5 30.2 * =19.5 39.7 TOTAL rcTOTc in S 7. I I - 77RC 9767 92. 19. C507. C8 709 96S.5 0 1.2774.2 Approved TOTAL NATIONAL 2,188.2 2,613.3 3,664.2 3233. 2.613.1 3.661.5 |1,101.0 1.715.7 2.199.2 1Li432. 3.728.8 5.860.7 TOTAL REGIONS/ TOTAL NATIONAL M , _y 3. 2. - , 1i,, 23.1 26.0 21.7 Totals may not tally due to rounding A = Approhed _stiates PA = Provisional Actual B = Budget Estimates * Includes opproim-itely..fl 3l3 illi4,n nf bl,c1k (yet fun-ctio.ally ,.allocated) funds for urbann district n.,d sub-district programe. + Includes Sh 3.4 million for the district development corporations and regional corporations. SOURCE: Budget Estimates - 54 - Table 4.5: TANZANIA: REGIONAL EXPENDITURES PER CAPITA, 1972/73-1974/75 Per Developjient Expenditures Total Expenditures Capita Population* Per Capita Per Capita Income Region Mid 1972 Mid 1973 Mid 1974 1972/73 1973/74 1974/75 1972/73 I973/i4 1974/75 in thousands ) ( in Shs. ) ( in Shs. 1. Arusha 719.7 743.8 768.4 7.2 20.0 27.2 45.4 86.3 105.0 762 2. coast 551.9 560.7 569.8 8.9 17.1 20.5 81.5 100.9 87.8 617 3. Dodoma 791.6 809.7 828.2 15.7 11.2 21.9 74.8 65.2 88.1 415 4. Iringa 802.5 827.3 852.6 4.6 13.3 19.7 52.5 58.3 69.4 519 5. Kigoma 508.6 5i6.1 523.9 8.; ;1.2 " 19. 52.5 65.9 92.2 500 6. Kilimanjaro 766.1 791.0 817.5 6.3 15.2 21.3 60.7 72.8 98.5 679 7. Lindi 456.2 464.2 472.4 5.7 23.3 32.2 64.0 92.4 111.3 460 8. Mara 644.3 666.4 688.9 8.1 21.0 lS.l 8.2 6S.0 7!.. 400 9. Mbeya 797.2 818.7 840.6 9.3 16.6 20.3 73.6 76.5 84.0 544 10. Morogoro 753.6 768.6 784.0 6.2 11.0 20.0 60.2 70.6 95.2 566 11. Mtwara 711.1 731.0 751.2 5.2 17.1 23M0 54.1 66.8 79.6 460 12. Mwanza 1,200.5 1,232.3 1,264.9 5.7 9.6 13.0 44.9 55.5 62.3 627 13. Ruvuma 449.0 461.3 473.8 8.7 21.5 24.1 86.4 93.9 112.1 410 14. Shinyanga 1,017.8 1.043.8 1,070.5 4-4 11.3 11.9 38.7 45.2 56.1 604 15. Singida 489.4 496.3 503.4 11.9 17.5 19.9 83.2 81.2 105.3 439 16. Tabora 553.5 565.0 576.5 6.5 13.5 23.1 78.3 84.6 99.4 443 17. Tanga 876.5 B99.8 923.5 4.8 15.2 19.7 64.7 77.5 99.1 730 18. West Lake 724.6 739.1 753.9 6.8 12.7 13.5 66.9 72.9 78.8 504 19. Dar-Es-Salaam 404.9 439.3 476.6 - - 48.5 - - 173.9 3,838 20. Ru-kwa 383.0 394.9 406.8 - - 23.4 - - 97.6 - TOTAL 13,602.0 13,969.3 14,347.4 6.8 13.9. 20.7 58.3 69.3 88.8 650 * Based on 1967 Census and growth rate 1948-67. SOURCE: Prime Minister's Office - 55 - Table 4.6 : TANZANIA: REGIONAL DEVELOPMENT ALLOCATIONS FOR SOCIAL SERVICES, 1974/75 Proportion of Primary Population Population Education per Served by Water Enrollment Hospital Water Region Education 7 Health Supply Population X Ratio Bed Supply (mill. Sha) (mill. She) (m.11. Sha) (mid 1974) (Standard I) 1. Arusha 2.4 5.5 1.2 3.7 2.4 4.7 768.4 5.4 52 650 28 2. Coast 2.2 5.0 1.4 4.3 2.2 4.3 569.8 4.0 42 1,700 12 3. Dar-Es-Salaa^ 2.8 6.4 8.3 25.2 1.0 2.0 476.6 3.3 - - - 4. Dodoma 3.1 7.2 1.0 3.1 3.7 7.3 828.2 5.8 47 1,200 38 5. Iringa 2.0 4.6 1.4 4.4 2.0 4.0 852.6 5.9 38 850 8 6. Kigoma 1.5 3.3 1.2 3.6 2.1 4.2 523.9 3.7 53 1,030 20 7. Kilimanjaro 2.4 5.4 1.4 4.2 2.3 4.5 817.5 5.7 60 640 30 8. Lindi 1.5 3.4 1.0 2.9 2.1 6.0 472.4 3.3 50 770 18 9. Mara 2.4 5.5 1.4 4.3 2.3 4.6 688.9 4.8 48 1,460 10 10. Mbeya 2.6 5.9 1.3 3.9 2.5 5.0 840.6 5.9 40 1,060 10 11. Morogoro 2.9 6.6 1.2 3.8 2.6 5.1 784.0 5.5 52 670 15 12. Mtwara 2.2 5.0 1.4 4.2 6.3 12.4 751.2 5.2 40 830 15 13. Mwanza 3.3 7.5 2.0 6.1 3.5 6.9 1,264.9 8.8 44 1,330 10 14. Rukwa 0.8 1.9 0.9 2.8 1.6 3.1 406.8 2.8 - - - 15. Ruvuma 1.9 4.3 0.7 2.2 1.4 2.8 473.8 3.3 73 440 19 16. Shinyanga 1.7 3.9 2.8 8.3 2.8 5.4 1,070.5 7.5 40 1,800 6 17. Singida 1.8 4.2 1.0 3.0 3.4 6.7 503.4 3.5 62 900 17 18. Tabora 1.3 3.0 0.8 2.6 2.2 4.2 576.5 4.0 66 720 35 19. Tanga 2.7 6.1 1.2 3.6 2.0 4.0 923.5 6.4 60 670 17 20. West Lake 2.3 5.3 1.2 3.8 1.4 2.8 753.9 5.2 68 560 * TOTAL 43.8 100.0 32.8 100.0 50.9 100.0 14.347.4 100.0 - - - SOURCE: Budget Estimates. presently available in each region. In case a region's share in regional ecacation expenditure is higher than its share in total population, it would suggest that it is receiving a higher than "average" allocation per capita. 4.24 For education, then, the fact that regions like Coast, Dodoma and Morogoro with low enrollment ration have been allocated relatively larger education expenditure per capita "than the average" is welcome. Conversely an attempt seems to have been made to reduce Tabora's share as it already has a high enrollment ratio. Similar conclusions can be reached with respect to health and water supply allocations in the regional budgets. Yet a closer perusal of Table 4.6 suggests that results once again are mixed, i.e., the negative correlation between sectoral alloca- tions for a region and the existing availability of public services is iar from perfect. Parastatal Expenditures 4.25 The foregoing conclusions are based on regional government expenditures which are only 20% or so of the total Goverment expenditures. Very little is known about the impact of Central Government expenditure or the rest of the public sector expenditure on the inter-regional equities. One can only make a guess here. If one assumes that bulk of the productive activities of the Central Government have not yet been decentralized or that the activities of the major crop authorities or the credit policies OI the rural and industrial development banks have by and large remained unaftected by the adoption of inter-regional equity as a national objective, one can surmise that these sectors perhaps would not be contributing to inter-regional equity in any significant manner. 4.26 In fact, a study carried out in the Dev Plan has concluded that "there is still strongly uneven distribution in parastatal investment in the country." The gross per capita parastatal investment for Tanzania (mainland) was estimated by this study at 31 shillings -- with Arusha, Coast, Mbeya ana Morogoro receiving at least 25d above this average while Lindi, Mara, Mtwara, Ruvuma. Shinvanga and Tabora received 25% below this average. 4.27 Many recent big Central Government projects have been found to have benetitted the relatively richer regions. TAZARA, for example. which became a national priorit.Y project following the UDI by Rhodesia in 1965, costing over 3.000 million shillings, is likelv to benefit regions like Coast, Morogoro, Mbeya and Iringa more than others. Kidatul, a hydroelectric national pro ect. costing about 800 million shillings, is also situated in Morogoro region. Inves-ment in tourist industry has led to the construction of hotels in the national parks oL Arusha and Coast regions and the location of the Kilimanjaro International Airport in Arusha. The same observation couild be made with respect to the location oI the General Tyre Factory in Arusha, fertilizer and steel - 57 - rolling mills in Tanga, textiles factorv in mNwnza a cffee Arocessing factory and sugar estates in West Lake region and several other large national pro.jects. Conclusion 4.28 To sum up: Although rInvArnment appears to have start+ed using its expenditure allocations to improve the regional balance in the availability of social- servi-es; the resumlts are still very mixed. The "pockets of development" and inter-related social and economic in- frastructure generated by the- Annort-nrien+tat±on of the developmen.t in the past still continue to have their pull on budgetary allocations. Besides, the facts that regional budgets are still vpev qmAll arxA that parastatal investment decisions are still made on financial considerations rather than on the basis of the onhip-Mtve of inter-regionl eqity leave on.e with the feeling that, though the start has been made, the end is still verv far. h.29 Anothn r oonnlunion of this Chapter is that the prospects of financing national targets oI social services with likely fiscal revenues are highly ucertai. TPhe projected revenue e w, ,b so large that national targets for 1990 will probably have to be modified even if major new revpeni effor+-s (inclduing tation and levy of user- charges' are undertaken. Gn - I W wrv-Xm> OF IUZ r_UU1tlYaJtrLcAln yL OflilOb =alm 5. I '--.- -l.n .a Siuuation wlere decen -alze aU,irLwdIjions do not nave to generate their own resources and the Center undertakes to finance all their activities, it is relevanl to ask: now does the Center ensure that the regions (and districts) control their budgets or "effectively" utl4ize the fLUndE whLich they receive from the Center? The purpose of this Chapter is to describe the mechanisms which have been instituted since decentralization to affect cornbrols at the time of the formulation of the budget and difrink the fiscal year when the budget is being implemented. toer the Chapter '-.dicates ho-" these control mecnanisms seem to be working in reality. Control Mechanisms at the Formulation Stage 5.02 The major instruments of controlling a region from making n "ndue" demand on the developmeAnt buaget in general or to "overspend" on any particular development project are the "Plan Guidelines" and the "Stand&-d Cost!! estU,tes of a project. 10The TlcGLi uuidel±ens', as was mentioned earlier, limits the aspirations of a region and forces it to decide upon its sectoral priorities. -I Ue f.inal analysis, as was shown in Chapter II, it is - 58 - 4the C, -r ..1,4lc-. A -s -1, Aana- d lvp.er+ fNi, a rmgior. t,11 receire ,,lle u e1iv ML". LiJ%A';- 1 % LAi.WN 51~.L~' ~ ~ - or wha'. they will be spent on. 5.oh. To ensure that all development projects will be "reasonably" costed by ihe re.g±Uion at the n tim of t Uhe 1ILU-ati.on of .LiLt UvelU.et budget, "standard cost" estimates of different projects are issued to regions. These esti.Uates take into acco-aUnt only the CenLer's eUpUectLation as to the amount of voluntary labor participation which would be forthcoming transport costs and other such economic factors. The application of !'sUandZIUdVU cuost!!j -5 LUVO 5stUmaIA) se, tb a aJU. .L4 t-.UUM11=L.U of. uV.1Ir,-LLLLU project finance in Tanzania. 5.05 On the recurrent budget side, the major control mechanism used at the ','ume of the fo0rulrlatJ.±5on o0f thLe Uuuget is whe arinuLuncemzII. by the Center of "ceilings" on each region's budget. These "ceilings" are noi-nally arrived at by tak-ing into account the likely revenue position of the Government, previous year level of recurrent expenditure of a region, "normal creep" as well as tne recurrent expenditures resulting from recently completed development projects and the size of the development budge. GenerUaLy, the "cei igs" are so utsignied as to imtprebs on the regions the need for iiscal and financial discipline. 5.Ob To further tighten the control upon the recurrent budget, the Center insists that the regions should not frequently reallocate their funds from one sub-vote to another or transfer the funds earmarked for one sub-head to another sub-head. Only recently some discretion has been granted to them in matters relating to their recurrent budget. Never- theless, as was indicated in Chapter III, funds a1located for major components of the recurrent budget like "personal emoluments" and sub- ventions" still cannot be-reallocated without proper authorization. 5.07 It seems then that the "Plan Guidelines" and "Standard Cost" estimates act as two control mechanisms at the time of the preparation of the regional development budgets and the "ceilings" and limited "reallocation powers" work as two control mechanisms at the time of the preparation of the regional recurrent budgets. A further control mechanism lies in making it difficult for the regions to get "supplementary" votes. Conditions for "Supplementary" 5.o8 To ensure that regions act with adequate fiscal discipline, the Center has made the grant of supplementary allocations to them as difficult as possible. 5.09 Under the Financial Orders, no ministry ojl region is entitled to a "supplementary" vote for its recurrent budget excepting: (i) "in cases of dire need where public interest will be jeopardized;" - 59 - ii) "r'y 4 itca, be s h-M, tat te e xp-er."tecol no- %have been ±±J_. " U J _.±J ..4 L t I v.L U 1 OI WL I ,J U U.LIL J4J1LL.LVIU.Lq %, Lk.LLA V1 LI . foreseen at the time of formulating estimates"; and also, (iii) if it c annot bue met W.LI.14W I., rJ.UL1 0 -h the .oL..LUs reL'Jo 4 caJtiI 0 . GVer.L.LyLJ NIJ io.i' about the expenditure indiscipline which decentralization might encourage, ',he grant of su-pplemenares -was-a en madre dVeVf' i ir U -L I .U. u .L L 7 1972/3) by postulating: (iv) that all requests for such funds must have the approval of the Economic Committee of the Cabinet. 5.10 Similar conditions are also imposed ior the supplementary requests for development budget which a region might like to make. Control Mechanisms During Implementation Stage 5.11 The major control mechanism here is the rate of disbursement of tne funds to a region which is tied to Center receiving from the region its Quarterly Financial and other Reports. 5.12 All regions receive quarterly disbursements. While they automatically receive a quarter of their development and recurrent budgets in the beginning of the fiscal year, i.e., July 1, further quarterly disbursements are not so automatic. 5.13 As far as disbursements for development budget are concerned, they are based on the region submitting to the Prime Minister's Office: (a) Quarterly Physical Progress Report comparing the actual physical progress of a program for the quarter and year to date with the plans as well as indicating a forecast of probable physical accomplishments in the quarter ahead; and(b) Quarterly Financial Report comparing the actual development and recurrent expenditure of a particular project or program with plans. Besides, the region must submit a "Problem Report" whenever an immediate action on a major actual or potential problem relating to a project is required in the region or district and which cannot be resolved there itself. Should a region fail to submit the former two Reports reqularly, release of development funds can be withheld. 5.14 For their recurrent expenditures also, the regions are expected to submit to the Treasury a Quarterly (previously Monthly) Regional Expenditure Summary Report showing actual recurrent expenditure (in- cluding outstanding amounts or comuIEiiEs already made) by districts and by sub-votes against the budget allocations. 5.15 Under the Financial Orders, the national budgetary accounts must be prepared by October 31 and then audited and presented to the Parliament by December 31. With decentralization, therefore, it has been expected that the regional accounts would be ready by September 30 or three months after the close of a fiscal year. The regions are, thus, expected to prepare Annual Appropriation Accounts which should include an Annual Programmes Performance Review indicating how each region or district performed over the last year by comparing, for each development project or program, the actual benefits, physical achievements and costs against the approved plans. - 6o - WNTorking of the Control Mechanisms 5.16 The available information on this is extremely limited, hence what follows must remain an impressionistic appraisal. 5.17 To begin with perhaps two factual statements can be made categotrc, 1 lyv. 5.18 First, the audited budgetary accounts of Government of Tanzania (including Regional Administrations) since decentralization was 4ntrodu&ced in 1972/73 have still not been presented to the Parliament. This means that data on actual expenditures by regions are not yet avai able even though it is about two years that the final accounts were due. 5.19 Second, most regions have asked for supplementary votes in the last t-wo years and are expect-ed to ask for them in the current fiscal year as well. Failure to Produce Final Accounts 5.20 A variety of explanations have been given for the failure OI b[the Regional A.;nistrations to prnoduce final accoplntso D)rng 1972/ 73, it is argued, the district and regional functional officers were still i,n the process of moving into +heir new Jobs with the result that they could not devote much time to the compilation of financial reports expected of 'fiem. A second explan.a+i.J for this i9 givpn in ters of the computerization of government accounts which has been in existence in Tanzania since thLLe early sixties. As ln as the nation,' b'dget related to centralized ministries in Dar es Salaam, computerized accounts could be easily prep-ared. Once the adinistratIon was decentralized, it proved extremely difficult, primarily because of long distances and sever-e com-wai'ncation constraints, 4o feed regional and di strict e.i,enditure vouchers and primary accounting returns into the Central computer on a daily basis. The Government has now decided to decomputerize government accounts with effect from the 1973/74 year with the exception of government employee salaries. 5.21 In addition to all these, it is suggested that the Regionl and District Administrations lack the necessary and properly qualified accounting staff (there has also been a rapid turnover in their staff) and this has made the prparation of reliable accounts extremely difficult. Even a relatively developed region like Arusha has found this to be a severe problem. It has 13 accounting staff of which only five are formally trained or experienced. It would like to have, if possible, 17 accountants of which 10 must be trained or experienced. The staffing situation in districts is certainly much worse. No district seems to have more than 50% of "idesired" level of accounting staff. - 61 - 5.22 IT is not oeny that preparation of budgetay accou^nts has suffered following the decentralization of Administration, the flow of 4r-o-mati required Po r -o.+rol mech.ism Whic * ere A escibed above is also simply not keeping pace. For example, the files of the rlme Minister's Office snow that Quarterly Physical Progress ReporTs were received from only six regions for the fourth quarter of 1972/73, eight regiors for tJhe first quarter, 16 regions aor the second quarber and 10 regions for the third quarter of 1973/74 and those are the last ofL the quarterly reports on plan projects received by the Center. It appears then that the Center has not been taking a very serious view of uhins problem and in this senae perhaps some of the financial control mechanisms have yet to become effective. Problem of "Excess Votes" 5.23 Under the financial rules of decentralization (which are also applicable to the ministries in general) it is only under exceptional circumstances that a region can be granted a "supplementary" vote. In fact, one of the listed "-sacred" duties of the Regional (and District) Finance Controllers is to see that "Voted expenditure is not exceeded." Yet the fact is that many regions have been "overspending" on recurrent account (see para. 2.18 above) and some have "overspent" on development account too. 5.24 Once again, various explanations are given for this situation. Firstly, it is suggested that the "ceiling" amounts suggested by the Center and even the "final" recurrent allocations approved by the Par- liament are highly unrealistic. The Center appears to have made severe cuts in the recurrent budgets proposed by the regions, particularly their expenditures on procurement and maintenance of vehicles, traveling allowances, and maintenance of roads, buildings and other establishments. Dodoma region, for example, asked for 25 vehicles in 1974/75, 13 as replacements for older vehicles and 12 new, but finally got only 15. Arusha region complained that the Treasury has permitted it only 7,500 shillings as the "standard (running and maintenance) costs" for each landrover and lorry while it has actually cost it over 12,500 shillings. Arusha requested 6.3 million shillings towards the cost of maintenance of roads, bridges, etc (excluding personal emoluments), but got only 1.3 million shillings. All these have left the regions with a handle that their budget estimates of recurrent expenditures have been unrealistic anyway; consequently they do not seem to attach much significance to their budget estimates. 5-25 There is yet another factor which accounts for "supplementary" requests for their recurrent budgets. The recurrent expenditure con- sequences of recently completed development projects, and it is well known that most of them are social infrastructure projects with parti- cularly heavy recurrent expenditure commitments, are not often anticipated at the time of the budget formulation with the result that they end up - 62 - becoming "isupplementary' reques'us as the fiscal year comes to a close. It is extremely difficult, in view of the complete absence of Appropria- tion Accounts, to substantiate the significance of this factor. 5.26 On the development budgaet sLde, as Chapter "V shows, t,he "standard cost" estimates are out-of-date. This also means that development supplementaries cannot always be ta&-en to be a re'ection 04' 44e ,--deAd "real" development effort by the regions during the course of a fiscal year. In part, as was shown in Chapter III, "standard costs" ca. prove unrealistic if the designated self-help component does not come forth and the region or district has to use paid labor in ±.txu Vo Lo. 5.27 Whatever the reasons of supplementaries, th.e fauct' I,tha' wite are more than isolated requests by one region or another suggests that Tanzania may have more than a marginal problem here. There i9 in fact some evidence that regional overdrafts on the Bank of Tanzania have become a serious problem and this suggests that the control mechanisms described above may not be working. 5.28 One of the explanations for such a situation also lies in the fact that with decentralization in 1972 the politics and administration of Tanzania have been more closely interwoven, with the former having a final say in decision-making. In addition, all the political weight has been placed behind the national targets of social services, as well as the ujamaa villages which automatically are assured of social services, with the result that administrators at the lower levels do not feel as "guilty" as long as they are overspending only on social services and not on other services. For many regional administrators recurrent ex- penditure descipline simply does not carry the same degree of urgency as it does for officials at the Center. It is precisely because of this that budgetary allocations which do show a declining share of social ser- vices in regional expenditures cannot be relied upon to predict what the reality would look like once actual expenditure data become known. 5.29 Another argument here is that any serious attempt by the Center to strictly control the budgets of the lower level administrations could be seen to be against the very spirit of decentralization and this would not be politically acceptable at this stage. This, of course, makes the task of administrators at the Center extremely delicate when it comes to enforcing financial controls over the regions. Is it a surprise, then, that no action is taken when quarterly reports are not submitted on time or when it is found that a region is likely to end up seeking an "excess" vote! Conclusion 5.30 To conclude: for Tanzania's system of decentralization to be successful it is imperative that financial controls are made to work. It seems that they are not working yet; in fact certain forces mentioned abov-e seem to be in operation which are likely to impede their effective- 4n- . o v tll. A solu+tio- to thi- prorbl mm rmist be folu,n if decentralization is not to become a fiscal liability. r' TTA fl(M'D Trr CTM1IA DV AUTh r'(lT T TTCQTtTKTC USIP" VI. v Jurl-.. . j OOiI UL 'SO 6. 01 . It i.s to.LJILt mekW UI.J r" t t AJL. be VLA' of th p .+s ..ade - - his report. It must be stressed again that inadequate data availability doesU .LV- pe.-IILL. atd thisL Ut e the LLV e-L4vation. o+ ar.y fi.,.er concl-orUSiO.ns. Hopefully as more data become available their validity can be tested and ~~~~~~~~~~~_1_ . _ _P _ _n - __ _ _L 4 - I !t - 1- A _- P an overt_ _l ge n l tL _ USXLU U16 OUCCXSS OL UCLLJC" 0 Lrar D _ * L 0 A decentralization made. Tanzania's Own Brand of Decentralization 6.02 Tanzania's system of administrative decentralization differs significantly from wnat is usually defined as "decentralization" of government. A decentralized government usually consists of multi- (often two or three) levels of government with each one having its functions ana tax authority to finance those functions clearly spelled out and even constitutionally defined. Each level of government enjoys autonomy in decision-making with respect to those functions and has its own administrative set up to make those decisions and own civil service to help make as well as implement those decisions. In doing all these, the lower level governments might, and often do, legitimately seek financial and other help from the higher level governments. 6.03 What Tanzania has actually done is that it has abolished local governments altogether. It has centralized their tax powers and has absorbed the employees of the former district, urban and city councils into a central cadre and offered them the same salary structure and conditions of service as those of the Central Government employees. It has also simultaneously centralized the final decision-making authority with respect to how much will be spent and on what. 6.o4 In its place it has created lower, regional and district, level administrations (quite distinct from governments) and delegated some functions or authority. For example, government activities in relation to "rural development" have now to be performed through lower level administrations. Besides, lower level administrations have been made responsible for initiating the ideas, but not finally deciding, as to what should be done in a region or district in the field of rural development. Then, the regions have been "authorized", on behalf of the Center, to carry out and implement the budgets and plans "approved" by the latter for which they are "answerable" to the Central Government and the National Assembly. The regions have of course been given some limited discretion to "deviate" from the centrally approved budgets and plans. Also they are not required to seek day-to-day authorization from the Center in whatever they want to do. - 64 - 6.o5 Thu-s the qvstem of decentralization adopted in Tanzania is somewhat of a unique system -- it is more akin to deconcentration of central aLthorit.y -- and perhaps the most important single characteristic of it is that now the public administration system (or the "executive branch") has been fully integrated for decisiai-making with the political party (or the "political branch"). Goals of Decentralization 6.o6 Theoretically, there is no reason why the taxing authority as well as -pening decisioc-makinP eannot be centralized while at the same time the spending activity is actually decentralized. In fact, the goals of ung ta policy as a redistributivP device or public exmenditure policy to bring about inter-regional (and rural-urban) equity in public ser-vices can perhaps best be achieved after the centralization of tax policy and public spending decisions have been brought about and this was one of the manor motivarions behind the 1972 reformse Similarly, the major advantages of decentralizing spending activity would be to bring about speedier implementantion of projects and programs as well as encourage local participation in development. 6.07 But, then, clearly there are no mechanisms inherent in the system of decentralizatioLn which would automatically encourage the spending units to minimize the costs of services that they provide or to control their expenditurecs Within the authorized levels. Also, as a result of the integration of politics and administration particularly where the form-uer has spelled out its targets for social services LLke education, health and water supply much more clearly than for economic prodUction decentralizationL of a-Inisturation in.creases the danger that productive sectors might get ignored in favor oI' social sectors. Generalizations 6.08 In appraising the results oI decentralization, with admittedly inadequate data avafable at this stage,afew geralizations ght be made. Achievements 6.o9 Decentralization appears to have succeeded in achieving the following: (a) It has speeded up spending and rem-ioved a serious 1n- stitutional bottleneck on local project implementational capacity. However, to the extent this capacity was constrained in the past by the lack of trained staff or availability of local and imported building materials or because of other procurement difficulties, decentralization has done, and in fact could do, very little to remove them. If anything, it has aggravated some of these problems. (b) It has made the pursuit of the goal of inter-regional equity easier for the Government. There is some evidence su.geotinP that the Government is now imnlementing this goal vigorously. (c) It has improved the communications between the villages, wAds, districtsj and regions- The loal nopl now know whom to approach for making suggestions or getting t.hings donn The decent.rnlized syrsft.em has, t.herPfore been successful in conceiving of development projects based on local "needst" and the Regional and District Administrations have found no difficulty in initiating n4 Adeas 6.10 Thmere apper tole cor.siderab-le scope fPor ir,.rovem,ent -4 h 5.. *.L'.1 £IV Li ppears.J t A IV 5..JL4 V. L 0...LV ...bV 4 5 L.JJ 4IlyJ 5. OISVJ L YV.i. respect to the following: (a) The quality of projects proposed by regions or districts las [LU i.a-[roved madrk-edl,y * I L . mThe decentralization of administration has not resulted in any significant inter- functLonal cooperationL at the regional or udlSbrict levels (this is partly because salary levels and conditions of service of different functional officers still remain far from uniform). Rarely any cost-benefit studies are done before a region proposes a project or even before the Center accepts it which, of course, is not unique to Tanzania alone. (b) The greater "cost-consciousness" does not exdst. The major tool of ensuring cost constraint on a development project, i.e., "Standard Costs" seems to have become unrealistic with heavy inflation in material costs and rapidly rising salaries and labor costs. A review of this question now seems to be in progress. (c) Decentralization has not resulted in a perceptible improve- ment in the motivation of people to participate in devel- opment projects. Overall voluntary labor contributions have probably expanded but this is because more is now being done, particularly in the social sectors, than before rather than that decentralization has generated an urge in the people to participate in development projects and contribute voluntary labor. (d) The expenditure decision-making has not been democratized which means the Center still dictates what the overall allocation of development funds should be. It has - 66 - attempted to place a lid on people's desire to have more and more social services at the cost of productive in- vestments but an answer to the question whether or not this strategy is succeeding in reality will have to wait until final appropriation accounts of the Regional Administrations become available. Nevertheless, one thing is obvious that much of the government productive activities are still outside of th, Jurisdictions ot' the Regional Administrations; they either belong to parastatal organizations or the Central Ministries. (e) The financial control mechanisms are not working and there is a real danger that regional expenditures, particularly heavy recurrent expenditures due to fast expanding social services on which regions are not allowed to levy any user charges, might soon get out of hand. 6.11 Of course, it is too soon to know for sure what, in the final analysis, the net benefits of decentralization would be. What has been decentralized so far is pnly nine Central Ministries and only about 20% or the national budget.2 The share of decentralized administration would be even smaller if the resources of the central parastatal enter- prises are taken into account. The uovernment does not appear to be rushing towards regionalizing its budget or its total administration. It would seem advisable to review the successes and failures of present "limited" decentralization before taking the next step but, then, this exercise will have to wait till more data become available. 1/ The "limited'' amount of decentralization which seems to have taken place up till now partly reflects the limited project preparation. oroject designing and imnlementiation canacity of the regional/local administrations. In part it is also due to the insistence of foreipn donors who nrefer to deal wi+th a single centralized agency even though the program may have regpional_/local character (e-g-; 5an-i, nassis,tance for health services).