the effects of constitutional amendment 29 on the regional allocation of public funds for the National Health Service in Brazil

Abstract This article analyzes the effects of Constitutional Amendment 29 in financing the Brazilian National Health Service, SUS, between 2000 and 2010. The aim was to analyze how the resources that were allocated by the three spheres of government were used on a general basis and in specific regions. Analysis was also conducted on the possible repercussions of the Amendment in the allocation of finances for SUS. The results showed: an important increase in the designated resources that were used by the three spheres of government during the aforementioned period. There was an increase in real terms of 112% in consolidated spending and an 89% increase in spending per capita by the three spheres. There was also more participation from the States, the Federal District and the Municipalities in financing the system. However, in spite of the increase in the use of financial resources, regional inequalities, in relation to spending per capita, remained practically unchanged.


introduction
The lack of financial resources is one of the main problems in the implementation of the Brazilian National Health Service (hereafter SUS).This is the case even though these resources have become more available recently and especially after the approval of Constitutional Amendment 29 in 2000.With the approval of Amendment 29 both the States and the Municipalities had their participation in the financing of their health services linked to their tax revenues.The Amendment required them to ring-fence and use 12% and 15% of their general tax revenues and specific federal tax revenues.In the year 2000, the minimum percentage to be used was fixed at 7% from general tax revenues and specific federal tax revenues from the States and Municipalities.This was due to differences in relation to designated percentages in Amendment 29 that required there to be an increase of a fifth every year [1][2][3] .With reference to the Union (the Union is a legal entity that is made up of the Brazilian states, municipalities and the Federal District) the financial resources for SUS were not connected to the amount raised by taxes but rather they were linked to the growth of the country's Gross Domestic Product (GDP).The starting value in 2000 was the value for 1999 with an increase of 5%.From 2001 the minimum value was set based on the year 2000 value given by the Health Ministry, with an adjustment due to variations in the GDP in the then previous two years.
It was noted that after the approval of the amendment, there was a significant increase in designated financial resources for the National Health Service.There was also an increase in the participation of the States, the Federal District and the Municipalities in financing SUS.Nevertheless the impact that the Amendment made, differed from region to region, which will be shown later in this paper.
One of the least explored areas of research was the repercussions of the regional allocations in public financial resources that occurred due to Amendment 29.Analysis was done to see whether the implementation of the Amendment, based on data from 2000 to 2010, had an effect in reducing the inter-regional inequalities with reference to the allocation of public finances for health.This was the aim of the text.Thus the purpose of this paper is to show that the implementation of Amendment 29 was effective in increasing spending on health.The same cannot be said with reference to reducing regional inequalities.This reduction would only be possi-ble if the Union were to increase its participation through providing finances to cover the expenses for health care services (public health care services hereafter will be referred to as ASPS) for the poorest regions in the Federation.This study noted that although the above occurred in the period that was analyzed, namely an increase in spending on ASPS for all of the regions in the country, during the same period there was a reduction in the Union's involvement with respect to financing.In the north east region the Union's involvement fell from 67% to 43% and in the northern region there was a drop from 54% to 32%.This meant that the increase in the ASPS expenses was financed by financial resources from the States and Municipalities.This in turn weakened the idea of reducing inequalities.
This paper is divided into five sections with the introduction being the shortest.The second section covers the methodology used and the sources used for this article.The third section provides details of the public spending that took place on public health care services and activities in Brazil between 2000 and 2010.From this we noted that the financial resources which came about due to Amendment 29 provided greater stability for health care financing through designating a continuous increase in this resource.It also instigated more involvement from the States and Municipalities in the financing of SUS.The fourth section will show that there were regional distributions of specific public funds for health care between 2000 and 2010, but despite the increase in the availability of financial resources, there were no significant changes in the positions of some regions.The fifth section gives our closing considerations.

Methodological aspects
The government, through its three branches, has consolidated public spending and works with the idea that spending should occur on ASPS in accordance with the fifth, sixth and seventh directive in Resolution 322 from the National Health Council published on 8 May 2003.This government's imperative, with slight changes, was ratified by the Federal Act 131 that regulates Amendment 29/2000.
Data relating to populations in evaluating public spending throughout different regions and macro-regions, was taken form DATASUS which provided specific demographic and socioeconomic information.
Budgetary information from the Secretary for Planning and Budgeting, which is a part of the Health Ministry (SPO/MS), was used to quantify federal spending.This information, that was used to bolster the government's goals for health care, excluded finances spent by the Health Ministry on ex-military personnel that were no longer active but received benefits from the government.It also did not include: internal and external services and debts, and expenses on the Fund to Combat and Eradicate Poverty (which would, after a few years, be an expense for the Health Ministry).In order to understand the weight of these exclusions, in 2009 for example the Health Ministry's total expenditure was R$ 62,919 billion Brazilian dollars.If you exclude ex-military personnel (that receive benefits) and the aforementioned debts, the value was R$ 58.27 billion Brazilian dollars.This equates to the expenditure on health care services by the Health Ministry.
In relation to the States, the Federal District and the Municipalities, the data that was used was taken from the Budgeting System for Public Health (hereafter SIOPS).The system contains information from the entities that make up the Federation.It was created to monitor and document the expenditure of the three spheres of government in public health care services.For the municipalities, we used data on their own municipal expenses which was taken from SIOPS.In other words this was data declared by them.The data from the municipalities was consolidated by state and has very good coverage.More than 96% of the municipalities had given their data to SIOPS between 2000 and 2010.For the states, the value of their expenditure with their own financial resources that was declared in the SIOPS was collated with values revised by technical reports drafted by the Technical Department in SIOPS.These reports that were drafted by specialists to be scrutinized and analyzed by the SIOPS Technical Committee, were a collection of information that had been declared by the states.On occasions small discrepancies were found in some of the values that were given.In these cases only information from the technical reports were used.This practice was used for the years 2000 to 2008.For 2009 to 2010 the data that was used, was that which had been given to SIOPS, since the analytical work stopped being done by the SIOPS team.
The consolidation of the expenditure for the three spheres of government in public health care services, was done by the states.The expenditure values for the states were later aggregated with the five largest regions.In every State there was a consolidation of financial resources that from the Secretary of State for Health with their own resources.There were resources from the municipal secretaries for health that were used with those allocated by the Health Ministry.This was done either through transferences to the states and municipalities or direct transferences were made from the Health Ministry in the state.States and municipalities using their own financial resources complied with the use tax revenues as stipulated in Amendment 29.
In relation to federal expenditure, not all the Health Ministry's expenditure on public health care services was regionalized.In 2009, for example, the value for non-regionalized expenditure was R$ 17.4 billion Brazilian dollars, which was almost 30% of the Health Ministry's expenditure on ASPS.This means that only 70% of federal expenditure in ASPS was regionalized.In 2009 expenditure on ASPS, which was under the national rubric, was on active people (41% of the total) and on expenses related to centralized purchases and acquisitions (for example for the purchase of medication).The purpose was for the distribution of what was bought by the Health Ministry to states and municipalities.
Using this methodology it was possible to regionalize the percentages that varied between 92.5% in 2000 and 87% in 2010.This was in relation to total spending by the three spheres on ASPS.The remaining finances that were non-regionalized referred to spending carried out by the Health Ministry done using a centralized register.
In order to analyze whether the approval of Amendment 29, which saw increases in financial resources used by the three spheres of government, aided in reducing regional inequalities in respect of spending, we used as an indicator for progressive percentage variations in public spending per capita done by the three spheres in each region.This was done in relation to the national average spending on public health care services.
To understand these variations we used, as our baseline, the following years: 2000, 2005 and 2010.The identical procedures for comparing percentage variations of regional values in relation to the national average were also used when looking at the allocations of federal financial resources.

effects of amendment 29 on the increases in Health care Public Spending
The approval of Amendment 29 brought with it a large amount of financial resources for SUS.It is debatable whether, even with this injection of funds, this was sufficient to meet the demands of the public system 4,5 .For the federal authorities, Amendment 29 would guarantee the minimum amount that should be spent, with the responsibility resting with the Union.These finances were to be used for public health care services.We noted that from the year 2000 federal financial increases were not proportionally high when consideration was given to additional finances from other federal bodies.This was a reflection of the then growth in the GDP.This clearly showed that there was some financial stability during this period as financial resources increased on a continuous basis 6 .The importance of this fiscal change was noted by the public row that ensued based on restriction on the Union's health budget.The changes meant that the Union had to ensure that a certain amount of finances were raised, ringfenced and save and there needed to be increases in public spending on health care services.There was also a requirement to fund programs aimed at transferring funds to the less well-off 7 .
In spite of the aforementioned, Amendment 29 was successful in meeting its objective of obtaining more involvement on the part of the states and municipalities in funding health care services 8 .In the 1980's the Union designate, on average, 75% of specific public financial resources to health 9 .Then in 1996 this involvement dropped significantly to 63% 10 .In 2000, the year when Amendment 29 was approved, the Union covered 59.8% of the financial public resources allocated to SUS.Since then its financial involvement has been decreasing and was 45% in 2010.In the same period the states involvement have moved from 18.6% to 26.4% and the municipalities increased from 21.7% to 28.5% (Table 1).
Public spending by the three spheres on public health care services increased from R$ 64.8 billon, in 2000, to R$ 137.5 billion in 2010.These were the average figures for the last year.Amendment 29, without a doubt, was the most important measure for changing the level of spending on public health care services which allowed for a real term increase of 112% in spending by the three spheres.
The increase in financial resources was shared between the Union, the States and the Municipalities.The States gave additional financial resources equivalent to R$ 24.zilian dollars during the above period (R$ 38.7 billion Brazilian dollars to R$ 62.0 billion Brazilian dollars, an increase of 61%).
The changes in the spending of the three spheres of government on public health care services (ASPS) can be seen in Table 2, through two indicators: expenditure per capita and expenditure in relation to the gross domestic product.As a proportion of the Gross Domestic Product (GDP), public spending on health went up from 2.89% in 2000 to 3.65% in 2010.
In spite of the above which equated to less than 1% of the GDP and which pushed up public spending to 3.65% of the GDP in 2010, this still does not change the fact that Brazil is behind many other countries.These countries have systems similar to SUS and spend on average 6.5% of their GDP 11 .The expenditure per capita by the three spheres on SUS went up from R$ 381.6 in 2000, to R$ 722.0 in 2010, which was a real term increase of 89.2% during the period.
One way to measure the economic impact of the three spheres of government's financial contribution to SUS is to look at the revenue that that they received.
As was noted earlier, Amendment 29 established specific criteria for the allocation of the minimum amount of financial resources to be used on public health care services.For the Union the minimum value for 2000 was based on the value for 1999 with a real term increase of 5%.From 2001 the minimum values would be defined based on the previous year's values which were corrected by the nominal variation in the GDP in the previous two years.The states and the municipalities ought to have used, respectively, 12% and 15% of their revenue raised from general taxes and specific federal tax revenues.In other words only the states and the municipalities had this situation where their financial resources for health care were based on a percentage of their tax revenue.In 2000, the minimum percentage to be used would be 7% by bodies that made up the Federation.The hope was that the figures would be 12% for the States and 15% for the Municipalities by 2004, reducing the difference between the specific percentages and what had been verified in 2000 by a proportion of 1/5 in the year 1,3,12 .
To analyze the impact of the expenditure by the three spheres of government on ASPS in relation to their revenue, we looked at the average spent on ASPS by the States and the Municipalities in relation to what was raised from tax revenues (general and specific federal tax revenues).We also took into account what was defined as their own revenue and federal spending on ASPS in relation to the Union's total revenue (RCB).
Using data from SIOPS it was shown that spending on health care services by both the states and the municipalities was high in the aforementioned years.In the municipalities spending on ASPS represented 13.7% of their own revenue in 2000.In 2000 they were using 17% of their revenue on ASPS which was higher than the value specified in Amendment 29.By 2009 they were using 21.8% which was well above what was specified in Amendment 29.
In relation to the states spending on ASPS, it was 7.2% in 2000.By 2008 it did not reach the specified value of 12%.It was only by 2009 that the states reached, on average, a percentage of 13.8%, passing the 12% level for revenues from both general and specific federal tax revenues as designated in Amendment 29.It should be noted that these percentage figures are average indicators from the states which is different to the financial contributions from the other federal bodies.
With reference to the Union using the total amount of revenue that it raised (known as RCB) which is not always the best value to use because it includes tax revenues that are later shared with the states and municipalities through FPE and FPM and through pension schemes, we noted that it used 7.2% of RCB in 2000 and 7.5% in 2009.In other words its financial contributions practically remained the same.By the end of the 1990's, before Amendment 29, spending by the Union as a proportion of RCB was higher: be- ing on average 8.37% between 1995 and 2001 11,13 .Therefore as a proportion of its revenue, the Union's spending on public health care services, in spite of having a significant real terms increase, remained relatively stable.

the progression of regionalized public expenditure
Between 2000 and 2010 there was a considerable real term increase in public spending in all of the Brazilian regions (Graphic 1).Spending on public health care services in the regions was the sum of the per capita values which came from the Union, the States and the Municipalities.The spending requirement for the regions, which was spending per capita, practically remained unchanged during the period.The south eastern region had the highest values and the lowest was in the north east and central western regions.The north and south had average values with some changes during the period but it generally kept the same figures.As can be seen in Graphic 1 the highest growth in spending per capita occurred in the north east (88.7%) and the south (81.45%).In the regions, per capita spending, between 2000 and 2010 went from R$ 272.46 to R$ 514.30 in the north east and from R$ 336.71 to R$ 610.97 in the south.These increases, however, were not sufficient to alter the relative positions for each of the regions.
The distribution of public funds in health care is very unequal.It can be seen in Graphic 1 that in 2010 the south east region had the highest public spending per capita in health with a value of R$ 716.93 per inhabitant.After this region came the southern region (R$ 610.97), the central western region (R$ 610.17) and the northern region (R$ 590.79).The lowest spending was seen in the north east with a per capita value of R$ 514.30 which was about 38% less than public spending in the south east.In 2000, ten years earlier, spending in the north east region in relation to the south east was 33% less.
One of the main functions of federal spending was to reduce inequalities in the allocations of public financial resources for health.This objective was made clear in Article 17, of the LC 141 that came into force in 2012.It was also mentioned in item II paragraph 3 of the Article 198 of the Federal Constitution 1988.It is important to analyze the distribution of federal spending which came from the Health Ministry.Based on what can be seen in Graphic 2 there was an important inflection between 2003 to 2005 for all of the regions.This was particularly the case in the central western, south eastern and northern regions.This was due to the way the expenditure accounts were calculated.Federal civil servants started to have their spending on health services registered nationally as centralized spending from the Health Ministry.This meant that it would stop being regionalized.From 2005 the southern and south eastern regions that saw the highest amount of public spending, started to receive the highest funding allocations from the Health Ministry.This took the form of being direct or indirect transfers.The north east region, that had the least amount of public funds spent on it, was in third position with reference to regionalization of its resources that came from the Health Ministry.
The percentage of federal spending on the regions was higher in the north east and southern regions where it was equivalent to 43% of total public spending in 2010 (Graphic 3).When focus is given to the year 2000 when Amendment 29/2000 came into force, federal involvement was high relative to all of the regions.Over the next few years due to Amendment 29, involvement by the States and Municipalities increased in all of the regions.However in Graphic 3 one can see a difference for the States and the Municipalities: for the south east and southern region's spending from the municipalities is higher than spending from the states.However in other regions, principally in the north, state involvement was higher.
In the north and central western regions, as mentioned earlier, financial resources that were used from the states, was higher than from the municipalities.In the central-western region, the fact that public spending by Federal District was considered state funding, may have contributed to this situation.The north east region is the region with the most balanced and equal distribution of state and municipal financial resources.
Another interesting fact in relation to regionalized allocation of resources 14 is that the distribution met what it was required to do which was to reduce the historical regional inequalities in respect of public spending present in Brazil [15][16][17] .Thus it was necessary to see whether the differences in the allocation of resources by the three spheres of government increased or decreased and if in the regional distribution of federal resources this was done with priority given for the less well-off regions 18 , the north and north east.
So when a comparison is made with the years 2000, 2005 and 2010 considering percentage variations in public spending per capita by the three spheres in each region, in relation to the per capita national average, there was a reduction in spending in the north east.In 2000, the regional public spending (R$ 272.46 per capita) was 22.8% less than the national average (R$ 356.96).In 2010 with a per capita of R$ 514.30, the difference in relation to the national average decreased  to 17.97% (the national average was R$ 626.97).
The northern region in turn increased the negative difference in relation to the national average.The southern region is the only region in the last three years of the study, to have public spending above the national average even though its variation had declined.It was 16.82% higher in 2000 with a per capita value of R$ 412.33.It subsequently went down to 14.35%, higher in 2010 with a per capita value of R$ 716.93.Also when doing the same comparison to evaluate the regional variations in federal public spending on ASPS, we noted no signs of a reduction in regional inequalities.The allocation of resources per capita in the southeastern and southern regions was higher than the national average in 2000, 2005 and 2010.The high variation-was decreasing in the south east region and increased in the south.In the regions in the north east, the negative variations continued to drop.In 2000 it was -8.42% and in 2010 it was -5.61% of the national average.In the northern and center-western regions the negative variations in relation to the national average increased.The area that drew the most attention was the northern regions whose per capita was less than the national average by -4.94 in 2000 and in 2010 there was a high negative difference of -18.18%.This showed the inefficiencies of federal spending on ASPS in enabling reductions in the differences in relation to the national average.
Considering that the designated percentage earmarked for health by the Amendment 29 was not sufficient to reduce the inequalities, one needs to consider other methods.The criteria for federal transfers were placed in legislation back in 1990 in the SUS Act (Law 8080/90, Article 35).One of the major problems with Article 35 was that it did not define the role of the federal transfers.It is uncertain whether it should have promoted equality or reward efficiency.A study carried out in 1990 19 made one of the first proposals for implementing this Article.However even with it obtaining formal regulations to back it up, it had irreparable anomalies in that it had an excessive amount of criteria giving different orders.Besides the aforementioned, Law 8080 did not define the role that federal transferences should carry out.Law 8.142, that came into force on 28 December 1990 which was just over three months after Law 8080/90, stated in the first paragraph of Article 3 that while there was no regulation brought in for Article 35 the criteria of proportionality should be used for transferences.It also established that 70% of these resources should be transferred to the municipalities.This Total spending (per capita, R$) 352,43 420,25 590,79 272,46 346,10 514,30 412,33 492,05 716,93 336,71 428,17  349,36 436,60

North
Northeast Southeast South Midwest mandate was not followed by the Health Ministry.Also the proposed criteria was for equitable distribution, thus it did not show any concerns with correcting inequalities in the allocation of resources.
The requirement for the redistribution of federal financial resources with the aim of reducing regional inequalities was finally placed in Constitutional Amendment 29.However the forming and enacting of accompanying regulations for Amendment 29 occurred 12 years after the brining in of the amendment, through LC 141 in 2012.This law reinforced the redistributive role of federal financial resources and it established the criteria that must be used to define the distribution model for the distribution of resources.LC 141 presented an opportunity to create a distribution model for resources that prioritizes equality and having as its aim the reduction in inequalities between the different Brazilian regions.
Therefore the way how inter-governmental transferences are to be handled in relation to the National Health Service (SUS) was one of the issues included in LC 141 in January 2012.The regulating of the transferences from the Union to the States, the Federal District and the Municipalities is relevant for the following reasons: (i) the centralized character that ought to exist for management in public health care services (item I, Article 198 of the CF), (ii) the importance of the federal financial resources (about 45%) for SUS, and (iii) the redistributive character that should be understood in relation to allocations of these resources, in accordance with item II, paragraph 3 in Article 198 of the CF.
It is undeniable that the LC141 defined the general guidelines for the required methodology which transferences should adhere to, being a necessity for the health care of the population.It also has the aim of reducing regional disparities.The lack of the these guidelines and definitions was one of the main problems that was noted in the old regulations.Also LC 141 has a socioeconomic dimension to it which was not present in the Law 8.080.This means that important variables in the definition of health needs can now be taken into consideration.This begs the question how can health needs be both understood and evaluated?
What is unfortunate is that in the approval of LC 141 an opportunity was lost to simply define a proposal that would provide guidance on transferences between spheres of government.Therefore it was necessary to refer to Article 35 of Law 8.080, in order to use it in a more definitive way.A group of 14 considerations ought to be considered in the transference methodology for it to be used on a continuous basis.It is however difficult to do this without taking into consideration political and operational aspects.

Final considerations
In terms of practical results, it is undeniable that the approval of Amendment 29/2000 gave more stability with reference to federal financial resources for SUS.It also legally obliged the States and the Municipalities to provide financial commitments for health care.Spending was seen to have increase strongly.Spending on public health care services had a real terms increase of 112% on its total and an 89% increase per capita between 2000 and 2010.This was shared amongst the Union, the States and the Municipalities.In the last two years there was a higher increase than the proportional increase that occurred with the Union.The States made available some additional financial resources equivalent to R$ 24.3 billion Brazilian dollar, which led to a 200% increase.The municipalities made available R$ 25.2 billion Brazilian dollars in additional financial resources, which was an increase of 180%.Additional financial aid from the Union was R$ 23.3 billion Brazilian dollars in the same period and this was an increase of 61%.
The positive progression in spending on public health care services revealed itself clearly when measure in relation to the GDP.For the same period, spending on health went up from 2.89% of the GDP in 2000 to 3.65% in 2010.In spite of the increase, public spending on health in relation to the GDP continued to be well below the levels found in other countries with similar health care systems, where the value is on average 6.5% of the GDP.In Brazil public spending is approximately 46% of total spending (public and private) on health 20 .With this level of public spending Brazil is in a paradoxical situation: it is the only country with a public health care system that has a responsibility to care for all, but which spends more on private health care.
Therefore, in spite of there being growth in public financing for SUS, the value is insufficient for the system to meet its constitutional responsibilities.Those that argue in the need to increase the amount of financial resources for SUS place all of their hopes in the regulations from Amend-ment 29, that was passed in 2012 through LC 141.The hope was that LC141 would include changes in the involvement of the Union in financing SUS.This would mean that the Union, just as the States, the Federal District and the Municipalities, would commit itself in using the equivalent of 10% of its total revenue.This would increase by less than 1% of GDP, being R$ 40 billion a year in federal financing for SUS and it would be adjusted annual to take in account the changes in its tax revenue intake 21 .This was not included in LC 141.As a result other parliamentary initiatives with the same idea came about and an amendment petition was signed having more than 2 million signatures.This was sent to Congress.It was unsuccessful.The LC 141 did not bring increases in the amount of financial resources for SUS, but rather dealt with the role that federal financial resources should play in financing SUS, as set out in Amendment 29.It also defined the criteria, without operational proposals, for the allocation of federal financial resources for the states, the Federal District and the municipalities (Art.17).
Even with the limitations on the regionalization of spending, it was estimated that the methodology used was able to regionalize between 87% and 92% of total spending of the three spheres of government.The results showed, as mentioned earlier, that in spite of the increase in the use of financial resources, the differences in the allocations per capita between the regions did not change in any meaningful way.The exception was the north eastern region.This showed that although there had been a significant increase in spending in health care by all levels of the government, there was no significant redistribution in the regional allocations of public resources for health.
Finally it can be said that with the lack of regulations for Amendment 29, which touched on the redistribution of federal resources, helps to explain the results found that in spite of increases in the availability of financial resources due to the amendment, there was no reductions in disparities amongst the regions.Therefore better results would only have been obtained by chance and not design.Also in respect of t federal transferences, no regulations came with Law 8.142/90.There were no guidelines specifying its role in the process of decentralization and it was subsequently segmented in an intense way.So one of the possible contributions of this article in relation to what it states in Art. 17 of the LC 141 is that an increase in public resources for SUS does not automatically mean that there will be a reduction in regional disparities in relation to allocations.It can only occur where redistributive criteria for federal resources are adopted.collaborations SF Piola, JRM França and A Nunes had equal participation in the complete development of this paper.
3 billion Brazilian dollars in the period between 2000 and 2010.It went up from R$ 12.0 billion Brazilian dollars in 2000 to R$ 36.3 billion dollars in 2010 (an increase of 200%).The districts gave R$ 25.2 billion Brazilian dollars in additional financial resources.They used R$ 14.0 billion Brazilian dollars in 2000 and R$ 39.2 billion Brazilian dollars in 2010 (an increase of 180%).The additional financial resources from the Union were R$ 23.3 billion Bra-

Graph 2 .
Evolution of public spending (Federal Government) l with public health services, as value per capita, by region, 2000 -2010.Source: Secretaria de Planejamento e Orçamento do Ministério da Saúde (SPO/MS) and DATASUS.
Public spending by the three spheres on public health care services, 2000 to 2010 (in R$ billon from 2010).
Graph 1. Evolution of total public spending with public health services, as value per capita, by region, 2000-2010.
Graph 3. Percentage distribution (Federal, State and Municipal Government) in total spending with public health services, as value per capita, by region, 2000, 2005 e 2010.: Secretaria de Planejamento e Orçamento do Ministério da Saúde (SPO/MS), SIOPS and DATASUS. Source