The disparities in regional budget allocation in the Philippines have widened over the years leading to an increasing imbalance in resources. This steady growth of inequality has resulted in significant gaps in funding, with some regions receiving substantially less support despite having greater needs.
Budget allocation across regions in the Philippines remains highly unequal. In 2018, the Davao Region saw an increase in funding during President Duterte's Term, while two other regions suffered budget cuts[1]. This pattern of disparity continued in 2024, with CPBRD reporting that some of the poorest regions such as Zamboanga, Bangsamoro, and parts of Mindanao received less than 1% of the DOTr's P 181 million budget, while 85% was concentrated in Metro Manila (NCR)[2].
Evidently, for over three decades, underinvestment as reported by EDCOM in 1991 has contributed to a steady decline in education quality here in the Philippines. Although Senator Gatchalian noted that DepED consistently receives the highest budget allocation, distribution remains uneven[3].
In this project, we will use data science to analyze the budget allocation of various government departments in the Philippines. Through this, we aim to find potential inequalities or potential biases in the said budget at the regional level. In addition, we aim to establish correlations, should they exist, between the allocations and several key indicators such as population, poverty, and GDP.
The data set is a mix of different data sheets all grouped per region or province including but not limited to the budget allocations per department, crop and fishery production, enrollment data, population, poverty incidence, etc.
All regional and provincial offices’ budgets were collected apart from the BARMM. This is because of the relatively new Bangsamoro Organic Law which provides fiscal autonomy to the region. The law provides BARMM with the powers to allocate their own budgets and limitedly tax people within their own territory. Hence it is hard to measure the inequality with the budget of BARMM in relation with the other executive departments.
For each department, their budget allocations will be grouped by Region (excluding BARMM). Given a time span of 2020-2024, we have a total of 1360 Observations.
The main dataset includes the budget allocations of the 17 departments listed below.
Data regarding the budget allocations and specific projects of the DPWH was gathered from the documents and files publicly released by the Department of Budget and Management. Only the years 2020-2024 were collected as they were the only ones with available Excel files of the General Appropriations Act, the law passed by Congress each year which provided specific details on the budgets of each government agency for that year.
Data regarding crops, fisheries, and farmland, which was gathered to correlate with the Department of Agriculture and Department of Agrarian Reform, was gathered from OpenStat PSA website.
Data regarding the enrollment data of DepED was found from a multitude of sources. Certain years were provided by DepED itself in its website, meanwhile others were taken from images DepED provided to news outlets, and others were taken from posts made by the official DepED account on Facebook.
Data regarding population was gathered from the Projected Mid-Year Population from the PSA website.
Data regarding the annual per capita poverty threshold and poverty incidence was gathered from the 2023 Full Year Official Poverty Statistics Tables from the PSA website. Only the years 2018, 2021, and 2023 were collected as they were the only ones available in the Excel file.
Data regarding GDP was gathered from the PSA and OpenStat PSA websites. For the Regional Data, only the years 2020-2023 are available. For the Province/HUC breakdown, only the years 2021-2023 are available.
To understand our data better, we plotted the budget alloted by the Department of Education when adjusted to the number of schools and enrollees.
When adjusted for the number of enrollees, we can see that throughout 2020 to 2023, the region with the most budget allocation per enrollee has been the Cordillera Administrative Region (CAR). However, by 2024, MIMAROPA will have taken this lead. On the other side of the spectrum, Calabarzon consistently had the lowest per-enrollee budget allocation throughout 2020 to 2024.
An increase in the budget allocation of DepEd can be observed in the latter years, but the disparity between the regional allocations also becomes more apparent because of a wider range of colors present in the map.
Looking at the per-school budget allocation of DepEd, the National Capital Region received the most budget from 2020 to 2024. NCR is seen to have this bias since it has a relatively high budget despite having the lowest number of schools compared to the other regions, which can also be observed through the colors of the map, with NCR being the only region receiving the upper range of the budget.
This bias is not seen when taking into account the number of enrollees, as NCR has a proportionate amount of enrollees with respect to its budget. CAR, on the other hand, has had the least amount of budget allocation per school throughout 2020-2024, despite having a relatively high budget allocation per enrollee throughout the entirety of 2020-2024.
To test our hypothesis we applied the Kruskal-Wallis H Test to assess if there exists a statistically significant difference between regions in terms of the allocated budget for DepED, after adjusting for the number of enrollees and the number of schools.
For the budget adjusted to the number of enrollees, a p-value of 0.00934 was produced by the test.
For the budget adjusted to the number of schools, a p-value of 1.78304 x 10^-9 was produced by the test.
Given that both tests produced a p-value less than 0.05, the Null Hypothesis is rejected, suggesting that
To understand our data better, we plotted the budget alloted by each department per region when adjusted for their respective needs.
Across all departments, the spending varied wildly when related to each department's respective demands. In terms of agriculture, Regions in Luzon saw a greater allocation per hectare or metric ton of production, particularly in the Cordillera Region. Mindanao saw consistently less allocation despite its large agricultural land and high production of agricultural resources. In terms of social services such as housing, ICT, and employment assistance, certain regions with high populations saw significant underfunding compared to other regions with smaller populations. These regions include NCR, CALABARZON, Central Luzon, Central Visayas, and Western Visayas. However, that could be because these regions could be compensated by the national programs to be enacted by the Department’s Central offices. Furthermore, funding towards housing, ICT, urban development, and employment programs could have been prioritized by the government towards less populated or less developed regions.
In other social services, particularly those related to the law, such as judicial and public safety, there was a great inequality in the funding between the regions, where the National Capital Region received a considerably larger allocation compared to the other regions. In particular, the DOJ and DILG had the NCR receive a larger portion compared to the other regions, as their maps show such a great disparity. This imbalance in funding just reflects the centralization of the judicial and police forces within the Philippines, where most of our funding towards jails and law enforcement is centered in the capital.
In social welfare and ayuda programs, the poor in NCR saw higher allocation compared to other regions. This reflects the greater concentration of social welfare programs for the urban poor, especially in the NCR.
For transportation and public works, certain regions are funded less. These regions typically include regions within the south of Luzon, Visayas, and Mindanao. The CAR saw that it was consistently funded well as its population and GRDP are quite low, and yet their funding has not been that far from the funding of other regions.
For health, the NCR, CAR, and Davao regions saw consistently higher funding compared to the other regions as which could reflect the lower population count for the case of CAR and the higher number of public hospitals and medical services in NCR and Davao compared to other regions.
For finance, trade, and economic development, certain regions in the North, and Mindanao and Visayas, saw higher funding, especially when compared to richer regions such as the NCR, CALABARZON, Central Luzon, Central Visayas, and Davao. This goes to show that despite certain regions having a greater contribution to the country’s GDP, the funding in the departments remains relatively equal throughout all the regions. Interestingly, MIMAROPA remained constantly less funded in the DOF.
Despite our initial guesses of over-concentration of funds in the National Capital Region, when excluding the budgets of the Central Offices, we found that numerous departments underfunded the NCR, especially when considering its large population and GDP. These departments include: DHSUD, DICT, DOLE, DOST, DPWH, DTI, and NEDA.
To test our hypothesis we applied the Kruskal-Wallis H Test to assess if there exists a statistically significant difference between regions in terms of the allocated budget for each department, after adjusting for demand.
The summary of the results of the test for each department and their corresponding demands can be seen in the table below.
| Department | Demand | p-value |
|---|---|---|
| DA | Volume of Agri-Fish Production | 5.18339e-6 |
| DAR | Hectare of Land | 1.25720e-9 |
| DENR | GRDP | 5.02694e-10 |
| DHSUD | Population | 0.04960 |
| DICT | GRDP | 1.96159e-7 |
| DILG | Population | 1.51195e-9 |
| DOF | GRDP | 4.24245e-10 |
| DOH | Population | 1.20957e-8 |
| DOJ | Population | 3.17324e-10 |
| DOLE | Population | 0.01326 |
| DOST | GRDP | 7.40774e-10 |
| DOTr | Population | 5.06934e-8 |
| DPWH | Population | 1.48457e-8 |
| DSWD | Population in Poverty | 1.86210e-7 |
| DTI | GRDP | 0.00016 |
| NEDA | GRDP | 2.57199e-10 |
Since all the tests produced a p-value less than 0.05, the Null Hypothesis is rejected, suggesting that
Excluding allocations that could possibly originate from the DepEd Central office, certain regions are found to be significantly underfunded compared to others in terms of funding for schools and regional offices. In particular, CALABARZON, with its large student population, falls way behind other regions in terms of funding. Sadly, only 1 region in 2024, MIMAROPA, ever got sufficient funding across all the regions and the 5-year period. Regions with larger student populations such as NCR, CALABARZON, and Central Luzon noticeably are underfunded more than other regions.
We performed k-means clustering using data from the Department of Education across the 16 regions. Features included are the Year, Number of Enrollees, Number of Schools, DepEd Budget, Budget_Enrollees, and the Budget_Schools.
Principal Component Analysis was used to flatten the number of variables to 3 to visually represent the clusters.
The clustering resulted in 3 main clusters being identified. Cluster 0 has regions with the highest budget on average in the Department of Education with around 50 billion pesos. Cluster 0 also has a lot of enrollees and schools with around 2.5 million enrollees and 800 schools on average. Cluster 1 contains regions with a lower average DepEd budget of 25 billion pesos, and has around only 1 million enrollees and 500 schools on average. Cluster 2 meanwhile contains just the National Capital Region which has a relatively high education budget of around 45 billion pesos for around 2.5 million enrollees and 300 schools.
Bar Graphs were added to display the Average of each relevant feature per cluster.
.png)
.png)
.png)
Cluster 0 seems to represent high-population regions with a greater number of schools and higher budgets. Interestingly when comparing the regions in cluster 0, the inequality index computed in the nutshell plot dictates that it contains regions that are underfunded and insufficiently funded.
Cluster 1 seems to represent lower-population regions with less schools and less budgets. These regions in the nutshell plot with the inequality index appear to be better funded due to their lower number of enrollees and schools. These regions range from being adequately funded to decently funded.
Cluster 2 is really an outlier which is the National Capital Region. Its outlier status could be explained that despite its high population and high budget, it lacks a lot of schools compared with the regions in cluster 0. The inequality index also signifies that the National Capital Region isn’t really fairly funded for its education demands, just like in cluster 0, but its bigger problem is really the lack of schools. This shows that the schools in NCR are overcrowded and teachers are responsible for larger class sizes which is really a problem in an educational setting as it is harder for teachers to effectively advise and give proper teachings to a larger number of students all at once.
Throughout the various departments, the adequacy of the budgets across the different regions and each department's respective demands varied significantly. Despite our initial guesses of over-concentration of funds in the National Capital Region, when excluding the budgets of the Central Offices, we found that numerous departments underfunded the NCR, especially when considering its large population and GDP. However in other departments especially those related to law and welfare, NCR was overfunded. For education, the NCR was an obvious outlier due to its low school count yet high budget and student population. Other regions were classified as underfunded and relatively well funded, notably with CALABARZON being significantly underfunded and MIMAROPA sadly being the only region to reach the adequately funded mark.
It is imperative for Filipino tax payers to know where their hard-earned money is going, especially in terms of national budgeting. The national budget lays the roadmap for social programs and infrastructures that affect our daily lives. Hence, we must know where our money is going and that it is being used with national economic growth and development in mind.
Misuse of public funds for fraud and corruption should be dealt with accordingly. Being mindful of such activities through constant monitoring of said budget is a stepping stone towards leveraging transparency. We can also communicate with channels that have massive reach in forwarding concerns about the national budget.
Our national budget and where they're going is ultimately decided by those in power--the same set of people we elect during voting season. To ensure that our hard-earned money is allocated equitably, we must elect leaders not out of popularity but by examining their track record in public service.
A major limitation which was repeatedly mentioned throughout the analysis and data gathering processes was the exclusion of the budgets of the Central Offices from the various departments. This was justified by the fact that most of the money the Central Offices spend are on national programs meant to benefit all Filipinos from across all regions, however in reality that might not really be the case. To determine a better understanding of how fair and adequate the budgets are for each region means we would have to consider these very large budgets from the Central Offices, however as the budget files of the government do not signify which regions the Central Offices spend on, as they are automatically set to NCR, it would be hard to fairly distribute the real expenditures from the Central Offices. Certain data points are also mislabelled, especially in cases like the DPWH, where projects in other regions are automatically set to the NCR, which can make assessing the budget allocations hard.
For future studies and further research, it might be better to take the Central Office budgets into consideration either through requesting more detailed budget plans or receipts from each government agency or through investigative/journalistic initiatives. Other government agencies may also be considered for the study such as the Department of Energy, Department of National Defense and others. It might be better to also study the budgets allocated for government corporations and organizations as they are allocated quite a significant amount in the government budget.
For determining how fair or unequal the fundings are for each department across each region, it might be better to study more deeply on which demand each department really tries to compensate or fund for. Is it truly the population of a region that matters to that department? Is it the Gross Domestic Product? Or is it some completely different variable which must be considered? By utilizing a better demand metric, fairer findings might be derived and better insights towards the budget allocation of the Philippines.
Hello there! I'm Gav, a BS Computer Science student at the University of the Philippines Diliman. I'm interested in playing video games, singing, and programming — and I recently discovered my passion for full-stack web development. 😎
Hello! I'm Bopie, a sleep-deprived penguin 🐧 with occasional back pains, BLACKPINK's first line of defense, and a bit of a Computer Science junior at the University of the Philippines-Diliman 🌻.
My mother wanted me to become a doctor, but I hated biology. I ended up becoming someone who loves bringing software and design ideas into life, so I guess that still counts!
Hello! I'm GV, a BS Computer Science Student, studying in the University of the Philippines Diliman. I have a great interest in history, economics, finances and programming. I love to play video games, watch animes and K-Dramas and read mangas and light novels in my spare time. 🙂↕️
