White Paper

Determining Municipality Clusters with a Shortfall of Public Schools

For the fiscal year 2018-2019, DepEd has an approved budget of Php 549.5 Billion[1] for basic educational facilities, including the construction of 46,998 classrooms[2]. These new classrooms will accommodate 45 students[3] per classroom in elementary and high schools. This thrust toward infrastructure construction in the education sector invites the question: Where in the Philippines would it be good to prioritize building schools? One way to prioritize the need for education infrastructure is to focus on municipalities with the least schools and the largest number of poor households, given that access to education is a big factor in alleviating poverty.[4]


  1. 1.  Clusters of shortfall municipalities (with few schools in proportion to the number of poor households) were found in Mindanao while sufficient municipality clusters (that fare better in terms of schools to poor households) were found in Northern Luzon.
  1. 2. The provinces of Maguindanao, Lanao del Sur, Lanao del Norte, and Bukidnon have the most number of municipalities classified as shortfall spots and thus should be prioritized for building schools.


POOR HOUSEHOLDS are defined as households whose per capita incomes fall below the provincial poverty threshold that is set by the Philippine Statistics Authority (PSA).

LOCAL SPATIAL AUTOCORRELATION is a way of identifying neighboring geographic areas whose metric values (in this case the PSPH ratio) are similarly high or similarly low.


JOSHUA MIGUEL CORTEZ is a graduate of Ateneo de Manila University with a degree in Applied Mathematics. His interests include Machine Learning and Spatial Data Science.

We define a metric: 

Municipalities with a low PSPH ratio have few schools and many poor households. If, for example, a municipality has a PSPH ratio of 6, it means that there are 6 public schools for every thousand poor households.

Local Spatial Autocorrelation

Using the PSPH ratio, sufficient clusters and shortfall clusters of municipalities were found by computing the local Moran’s I, a measure of local spatial autocorrelation, for every municipality.

Shortfall (in Red)

  • - A municipality that has below average PSPH ratio with neighboring municipalities that also have below average PSPH ratios.
  • - Groups of neighboring shortfall municipalities is a shortfall cluster.

Sufficient (in Blue)

  • - A municipality that has an above average PSPH ratio with neighboring municipalities that also have above average PSPH ratios.
  • - Groups of neighboring sufficient municipalities is a sufficient cluster.

Table 1: Philippine Provinces with Shortfall Cities and Municipalities

The shortfall municipalities need the most public schools. These municipalities and their neighbors face the same problem of not having enough public schools for poor households. Building schools in shortfall municipalities will be best for providing access to education since new schools in a single municipality could be attended by students from neighboring municipalities.

As seen in Table 1 above, there are 80 shortfall municipalities. Maguindanao has the most number of shortfall municipalities while Cebu has only one shortfall municipality. Compared to the Philippine benchmark of 11.5, the PSPH ratios in these areas are lagging behind.

Figure 2: Shortfall Municipality Clusters in Mindanao

The shortfall clusters are primarily in ARMM and in Northern Mindanao. The most prominent provinces are Maguindanao, Lanao del Sur, Lanao del Norte, and Bukidnon. In these 4 provinces, there are 63 shortfall municipalities.

These areas are also known for being rife with armed conflict[7]. ARMM and Northern Mindanao have a poverty incidence of 59% and 40.9%, respectively, based on the 2015 Family Income and Expenditure Survey (FIES).[8] The lack of access to education is intertwined with poverty and peace and order issues, so servicing these areas requires a multifaceted approach.[9]

This analysis could be extended to determine how much educational infrastructure, such as number of classrooms, should be built in the shortfall municipalities, given data on the number of classrooms per school and whether or not the school has double-shifting.[10]

Using that data alongside the school-age population per municipality, we could estimate how many school-age children/teens there are per classroom. Using school-age population instead of number of students controls for municipalities that have many children/teens that are out of school. We could then compare that to DepEd’s standard of a maximum of 60 students per classroom[11]. Then, the number of classrooms to be built should be to the extent that the school-age population per classroom is at most 60. 

This has been an example of how spatial data science can be used to make better informed decisions in public policy. As more government datasets are being released in the recent years[12], citizens and public officials alike will have more opportunities to harness insights from data. The next step would then be to bring these insights into action to improve our communities.

[1] DepEd (2017, Oct 13). Senate approves P549.5-billion budget of DepEd for FY 2018-2019. Retrieved from http://www.deped.gov.ph/press-releases/senate-approves-p5495-billion-budget-deped-fy-2018-2019

[2] DepEd (2017, Aug 15). DepEd proposes 2018 budget for inclusive, nurturing learning environment. Retrieved from http://www.deped.gov.ph/press-releases/deped-proposes-2018-budget-inclusive-nurturing-learning-environment

[3] According to Republic Act 7880, the standard classroom-to-student ratio is 1 classroom per 45 students.

[4] Ducanes, G. M. (2014, July 6). Education, location, poverty. Retrieved from http://opinion.inquirer.net/76312/education-location-poverty

[5] Coordinates and/or city locations of public schools were acquired from DepEd and CHED and processed by Cobena. Data is as of Oct 2017.

[6] The number of poor households per municipality as of 2016. This was obtained from DSWD.

[7] Gavilan, J. (2017, May 28). FAST FACTS: Poverty in Mindanao. Retrieved from https://www.rappler.com/newsbreak/iq/171135-fast-facts-poverty-mindanao

[8] Ibid.

[9] Oxford Poverty & Human Development Initiative Policy – A multidimensional approach.  Retrieved from http://ophi.org.uk/policy/mult...  

[10] Double shifting is a system where two batches of students attend the school in a day, typically one in the morning and another in the afternoon.

[11] DepEd (2012, May 24). DO 41, s. 2012 - Revised Guidelines on the Opening of Classes Retrieved from http://www.deped.gov.ph/orders/do-41-s-2012

[12] iGov Philippines (2017, May 22). iGovPhil collaborates with government agencies for Open Data http://i.gov.ph/igovphil-collaborates-government-agencies-open-data/


This article has been downloaded 78 time(s).

Related Articles

Indigenous Peoples' Access to Education

January 16, 2019
Using the school location data provided by the Department of Education and the Commission on Higher Education, we found that three ancestral domains have no public basic education schools even within 3 km from their respective boundaries.

Connect with Our Industry Experts