BETTER INTERNET FOR ALL FILIPINOS: Reforms Promoting Competition and Increasing Investment for Broadband Infrastructure January 2024 A Policy Note © International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover design: Michelle Lomibao i Acknowledgement This policy note was prepared by the World Bank Philippines Digital Development team led by Naoto Kanehira (Senior Digital Development Specialist), including Mary Grace Mirandilla- Santos (Consultant) and Mitch Abdon (Consultant), with Finance, Competitiveness, and Innovation team including Jaime Frias (Senior Economist), Luis Andres Razon Abad (Senior Private Sector Development Specialist), and Kimberly May Baltao Chandra (Consultant), under the overall guidance of Mahesh Uttamchandani (Digital Development Practice Manager, East Asia and Pacific). Jonathan Marskell (Senior Digital Development Specialist), Biondi Sanda Sima (Consultant), Sharon Faye Alariao Piza (Economist), Irene Jo Estigoy Arzadon (Consultant), Anshuman Sinha (Consultant), Toni Kristian Eliasz (Senior Digital Development Specialist), and Son Tuan Vo (Consultant) contributed to the background research and reports. Samia Melhem (Lead Digital Development Specialist) and Alvaro Gonzalez (Lead Economist) peer reviewed the draft and provided valuable comments. Advice was also received from Feng Liu (Infrastructure Program Leader), Gonzalo J. Varela (Equitable Growth, Finance and Institutions Program Leader), Andrea Barone (Senior Economist), Rajendra Singh (Senior Digital Development Specialist), Paul Phumpiu Chang (Senior Economist), and Davit Melikyan (Senior Public Sector Specialist). The team is grateful for the guidance and support from the World Bank Philippines Country Management led by Ndiame Diop (Country Director) and Dandan Chen (Operations Manager) and from the External Relations and Communications team led by Clarissa Crisostomo David (Senior External Affairs Officer) as well as the administrative support provided by Teresita Fallado Victoria (Program Assistant) and Zoe Adriel Palispis Escobar (Team Assistant). The report was edited by Isabelita Orlina Reyes (Consultant), and the graphic designer was Michelle Lomibao (Consultant). The team sincerely thanks partners and collaborators, including Scott Minehane and Paul Zaman (Windsor Place Consulting), John Garrity and Gigo Alampay (USAID BEACON), Krystal Lyn Tan Uy and Philip Amadeus D. Libre (NEDA) for generously sharing knowledge, contributing to background research, and coordinating initiatives. Table of Contents List of Figures and Tables iv Abbreviations v Executive Summary vi 1. Context: Poor Internet Constrains Inclusive Recovery and Growth 1 2. Problem: Market and Regulatory Failures 5 2.1 ‘Divided’ Market Structure and Underinvestment in Rural Infrastructure 5 2.2 Inadequate and Lagging Policy Framework 10 3. Policy and Regulatory Gaps: Industry-Wide Challenges 14 3.1 Barriers to Market Entry and Investment 14 3.2 Unlevel Playing Field 15 3.3 Ineffective Infrastructure Sharing Policy Framework 17 4. Policy and Regulatory Gaps: Spectrum Management Deep-Dive 20 4.1 Outdated Legal Frameworks 20 4.2 Unclear Procedures in Spectrum Recall and Reassignment 22 4.3 Suboptimal Fiscal and Economic Outcomes of Spectrum Policy 25 5. Ways Forward: Improving Philippine Broadband Infrastructure 27 5.1 Urgent Broadband Infrastructure Policy Reforms 27 5.2 Role of the Government in Bridging Digital Divide 32 6. Conclusion and Recommendations 35 References 37 Annexes 46 Annex 1: World Bank CEM3.0 Growth Ambition and Structural Reforms Tool 46 Annex 2: Affordability Driver Index (ADI) Policy Score Data and Methodology 48 Annex 3: Evidence of Regulatory Reforms Impact on Internet Performance and Growth 52 Endnotes 57 iii List of Figures and Tables Figure 1. Fixed and mobile broadband access, speed, and cost, ASEAN countries 2 Figure 2. Regional comparison of proportion of youth and adults with ICT skills 3 Figure 3. Broadband access trend, across ASEAN and within the Philippines 3 Figure 4. Broadband penetration, per population quintile by municipal population density 4 Figure 5. Impact on GDP annual growth of Philippines fully catching up to Malaysia in 25 years, in percent 4 Figure 6. Telecom and digital connectivity market participants in the Philippines 5 Figure 7. Broadband internet penetration by connectivity types and provinces 8 Figure 8: Top operators’ market share, profitability, and infrastructure, comparison in Asia and Pacific 9 Figure 9: Infrastructure investment and broadband adoption, the Philippines and select peers 10 Figure 10: Affordability Driver Index, individual policy scores, and trends in select ASEAN countries 11 Figure 11: Legislative and administrative landscape for Philippine broadband infrastructure 13 Figure 12: Operators to government payment through main fiscal channels, select ASEAN countries 26 Figure 13: “Theory of Change” for the needed broadband infrastructure market reforms 27 Figure 14: Differentiated interventions fit for market constraints 34 Table 1: Market performance of Globe/PLDT and small ISPs 8 Table 2: Radiofrequency spectrum management – Philippines and other countries 24 Table 3: Global evidence on the impact of key reforms 29 Box 1: Broadband Value Chain 5 Box 2: Policy choices, sequencing, and trade-offs in addressing connectivity gaps 11 Box 3: What is Universal Service, and do Philippine operators have such obligations? 16 Box 4: Infrastructure Sharing – policy trade-offs and practices 18 Box 5: What is radiofrequency spectrum? 20 Box 6: Connectivity in unserved remote area is proven to generate significant welfare impact 30 iv Abbreviations AI Artificial Intelligence CAPEX Capital Expenditure CDN Content Delivery Network CPCN Certificate of Public Convenience and Necessity CSP Cloud Service Provider DICT Department of Information and Communications Technology DSL Digital Subscriber Line EBITDA Earnings Before Interest, Taxes, Depreciation, and Amortization EO Executive Order FOC Fiber Optic Cable FWA Fixed Wireless Access FTTH Fiber to the Home GDP Gross Domestic Product GIDA Geographically Isolated and Disadvantaged Areas GNI Gross National Income ICT Information and Communications Technology IP Internet Protocol ISP Internet Service Provider IT-BPO Information Technology Business Process Outsourcing ITC Independent Tower Company IXP Internet Exchange Point JMC Joint Memorandum Circular LEO Low Earth Orbit LGU Local Government Unit MNO Mobile Network Operator MSME Micro, Small, and Medium Enterprise NBP National Broadband Plan NEDA National Economic and Development Authority NGA National Government Agency NTC National Telecommunications Commission OADT Open Access in Data Transmission OPEX Operational Expenditure PA Provisional Authority PCC Philippine Competition Commission PDP Philippine Development Plan PPP Public-Private Partnership PTE Public Telecommunications Entity POP Point of Presence PSA Philippine Statistics Authority RA Republic Act ROW Right-of-way SMP Significant Market Power USF Universal Service Fund USO Universal Service Obligation VAS Value Added Service v indicated. All dollar amounts are US dollars unless otherwise Executive Summary Executive Summary Equitable access to broadband services is imperative to narrow the digital divide and for more people to benefit from digitalization. Digitalization has the potential to improve efficiency, productivity, and innovation across various sectors, and drive economic growth and job creation. Access to broadband is fundamental to participating in a country’s digital transformation. However, the digital divide in the Philippines is rapidly expanding. In the least populated areas and remote islands, progress in household Internet penetration over the last 10 years has only been a third of progress made in populated urban centers. If the pattern remains unchanged, these gaps will grow wider and inequity in access to opportunities will be exacerbated, leaving most of the poor behind. The government’s key development priorities will be jeopardized if poor broadband access is not addressed. Compared to other ASEAN countries, the Philippines’ internet connectivity lags in affordability, speed, and access, creating an uneven landscape for digital participation. The price of fixed broadband, essential for productivity (e.g., work, learning, micro, small, and medium enterprises’ participation in the digital economy), is 11 percent of gross national income (GNI) per capita, more than four times higher than Malaysia and Viet Nam, and two times higher than the ASEAN average. The cost of fixed broadband converted to absolute USD price is the highest in ASEAN. Philippine household penetration of fixed broadband in 2022 was 33 percent, much lower compared to 50 percent in Malaysia, 58 percent in Thailand, and 76 percent in Viet Nam. For mobile broadband— considered the driver of consumer adoption of e-commerce, financial inclusion, disaster response, and agriculture practices—active subscribers in 2022 were 70 per 100 inhabitants, the lowest among large ASEAN economies. The Philippines is estimated to represent more than 50 percent of the ASEAN population unconnected to mobile broadband. The cost of a mobile broadband basket was 2 percent of GNI per capita, 1.5 times higher than the ASEAN average. Limited internet access curbs digital potential for citizens and businesses, with peri-urban connectivity being critical to future growth. UN data reveals that in 2019, only 2 percent of Filipino youth and adults could use basic arithmetic formulas in Excel, 6 percent could copy and paste into a document, and 7 percent could attach a file to an email. Clearly, broad systemic educational challenges need addressing. However, acquiring ICT skills is not possible if children are unable to digitally connect in the first place. Similarly, for enterprises, weak internet access results in insufficient technology adoption, especially among micro-, small, and medium enterprises (MSMEs). The IT-BPO industry, the driver of employment and service export growth, is still constrained in many locations. The Philippines was the last investment destination among major ASEAN economies for hyper-scaler cloud service providers, such as AWS and Google Cloud. vi The country’s poor broadband infrastructure is rooted in outdated policy frameworks that stifle investment in rural areas and foster a market with weak competition, both of which hinder broadband expansion. Laws on connectivity have remained unchanged despite vast technological advancements, evolving business models, and widening access gap. The broadband market in the Philippines is an effective duopoly. The top two players are vertically integrated: they own international, backbone, middle-, and last-mile connectivity, in the most concentrated, most profitable, and least invested market in the region. According to Affordability Driver Index (ADI) policy scores, the Philippines is among the least favorable in the ASEAN in terms of policy environment for affordable broadband and is among the slowest in the world in promoting any major reforms for affordable broadband. Unlike most other countries, the Philippines neither incentivizes nor obliges its largest operators to expand rural broadband coverage. Hundreds of small internet service providers (ISPs) extend broadband services in rural areas while facing an unlevel playing field. Over the last 15 years, regional and global peers have invested more and rolled out fixed and mobile coverage faster. Over 100 countries during this period have invested at least 1 percent of GDP in telecom infrastructure for at least one year; in many countries, for several consecutive years. The Philippines has invested less, particularly in recent years—from 0.64 percent of GDP in 2018 to 0.44 percent in 2022. Binding constraints underlying the Philippines’ poor broadband infrastructure are inter-related, requiring a comprehensive package of reforms to yield desired entry, investment, and sector performance outcomes. Key regulatory weaknesses include: Barriers to market entry and investment. The political and quasi-judicial nature of franchising and licensing discourages new and small players from entering and thriving in the broadband market. The long and tedious licensing process creates uncertainty and increases the cost of investing in digital infrastructure. The Philippines remains the only country worldwide that still requires a legislative franchise for the construction and operation of networks. Unlevel playing field. High barriers to entry have historically encouraged the sector to consolidate, with the franchising and licensing regime favoring the business models of large, vertically integrated telcos and excluding small ISPs that rely on the telcos for wholesale bandwidth and facilities. There are no regulations preventing anti-competitive interconnection price discrimination, a situation disadvantageous to small ISPs. There are no regulations to curtail dominance of operators with significant market power. vii Ineffective infrastructure sharing policy framework. Lack of a passive infrastructure sharing policy framework, weak coordination among operators and regulators, and the absence of information on existing and planned infrastructure, make network deployment costly and inefficient. Lack of modern spectrum policy framework. Radio spectrum management is based on analog- era laws enacted in the 1930s and 1990s. Unlike in other countries, the Philippines allows an assignee to indefinitely hold spectrum and does not require any service obligation. Procedures in spectrum recall and reassignment are unclear, and spectrum assignment and use non-transparent, causing spectrum fragmentation, spectrum scarcity, and under-collection of government revenue. The Open Access in Data Transmission (OADT) bill is a promising, viable start, among several proposals in Congress. The House of Representatives approved the OADT bill in December 2022, and the Senate is currently deliberating it at the committee level. The bill aims to introduce a regulatory framework to expand Internet infrastructure and improve Internet services by encouraging more investment in broadband infrastructure, especially in the countryside. It will do so by simplifying and rationalizing the entry process for network providers, thereby allowing more ISPs to build and operate broadband infrastructure, especially in communities not served by large telcos. The bill will fast-track and make network rollout more cost-efficient, including for small players and community networks, by promoting infrastructure sharing. Finally, it will democratize access to broadband technologies, both wired and wireless, especially for geographically isolated and disadvantaged areas. Policymakers could build on immediate reforms through the open access bill as an entry point to broader and medium- to longer-term digital connectivity agenda. The next phase of spectrum reforms, such as mobile spectrum restacking, regional spectrum licensing, piloting spectrum auction and revising SUF pricing, could benefit from and/or reinforce, the immediate reforms through the OADT bill. This will serve as an intermediate step to ensure adequate fiscal policy options for longer-term agenda to lower end-user price for broadband connectivity and expand rural coverage while maximizing a scarce spectrum resource. The government’s infrastructure investments (such as through the National Fiber Backbone and Broadband ng Masa) can complement pro- competition reforms in a justifiable, efficient, and sustainable manner. Government competency will need strengthening in terms of design and implementation of policy reforms and complementary initiatives. The cost of inaction—loss of growth opportunity, people remaining unequipped for future jobs, and widening of the digital divide—is too high for the Philippines. Outdated policy and regulations have long stunted the growth of the country’s broadband industry and expansion of digital infrastructure. For inclusive growth through digitalization that benefits all Filipinos, updating Philippine policy to promote competition, encourage investment, and upgrade broadband infrastructure is urgent and necessary. viii 1. Context 1. CONTEXT: Poor Internet Constrains Inclusive Recovery and Growth Digitalization is a key development priority constitutes digital public infrastructure for the Philippines. In his State of the Nation critical for digitalization. The Philippines is Address in July 2023, President Ferdinand also an early adopter of the Cloud First Policy Marcos, Jr. stated, “Digitalization is the call of and a regional leader in cross-border data today; not of the future—but of the present,” flows via multiplying layers of international “government must fully embrace digitalization connectivity, both of which make the country to provide better service to the people,” and a potential regional connectivity hub. “the National ID system will be at the core Filipinos spend 11 hours daily online (Licsi of this digitally transformed network of 2022) and over 70 percent of the population government services.” are digital natives ready to harness these potentials. The Philippine Development Plan (PDP) 2023- 28 (NEDA 2023) emphasized advancing and However, accessing the internet is too expanding access to digital infrastructure expensive for most Filipinos. The state of both and digitalizing transactions and service fixed and mobile internet in the Philippines delivery, while enhancing data privacy lags behind regional middle-income peers and data sharing. The Plan recognizes on access, speed, and cost (Figure 1). For fixed the potential of digitalization to improve broadband, essential for productivity (e.g. efficiency, productivity, and innovation across for remote work, remote learning, micro and government and various business sectors, small enterprises’ participation in the digital which in turn will drive economic growth economy), household penetration in 2022 was and job creation. As the Philippines is on 33 percent, at the lower end of the ASEAN. track to be an upper-middle income country, Fixed broadband basket cost was more than digitalization is key to ensuring that everyone 11 percent of GNI per capita—more than four can benefit from economic growth. times higher than Malaysia and Viet Nam, twice higher than the ASEAN average, and The Philippines is poised to benefit from the highest in ASEAN in absolute USD price digitalization given its industrial ecosystem, —making broadband internet unaffordable ongoing digital progress, and population to the lower-income population. For mobile demographics. Digital industries are driving broadband, often the driver of consumer growth. The Philippine Statistics Authority adoption of e-commerce, financial inclusion, (PSA 2023) estimated that the digital economy, disaster response, and agriculture practices, composed of digital-enabling infrastructure, in 2022, active subscribers in the Philippines e-commerce, and digital media/content, were 70 per 100 inhabitants, the lowest expanded by 11 percent and contributed 9.4 among large ASEAN economies. The country is percent to GDP in 2022. The digital payments estimated to represent more than half of the system, which is more advanced than that of ASEAN population not connected to mobile regional peers, combined with the Philippine broadband. The cost of a mobile broadband Identification System (PhilSys) (registering 81 basket was 2 percent of GNI per capita, 1.5 percent of Filipinos 5 years old and above), times higher than the ASEAN average. 1 Figure 1. Fixed and mobile broadband access, speed, and cost, ASEAN countries Fig 1 – Access speed cost Fixed Broadband Mobile Broadband Access Speed Cost Access Speed Cost Fixed broadband penetration Median download speed Fixed broadband basket Active mobile broadband subscribers Median download speed Mobile broadband basket (% of households, 2022, TeleGeography) (Mbps, Nov 2023, Ookla) (% of GNI per capita, 2022, ITU) (per 100 inhabitants, 2022, ITU) (Mbps, Nov 2023, Ookla) (% of GNI per capita, 2022, ITU) Singapore 111 264 0.6 156 95 0.2 Brunei 108 94 1.2 119 74 0.3 Viet Nam 76 105 2.6 97 47 0.4 Thailand 58 216 3.4 122 41 1.4 Malaysia 50 112 2.5 127 67 0.3 Philippines 33 92 11.3 70 28 2.0 Indonesia 18 28 6.1 91 25 1.1 Lao PDR 12 33 7.2 56 29 2.1 Cambodia 10 26 11.6 102 26 2.3 Myanmar 8 20 15.3 97 23 3.1 Source: TeleGeography, Ookla, and ITU.7 Note: Dotted lines show ASEAN averages. Weak internet constrains broader digital The gap in access to digital infrastructure opportunities. UN data reporting on the persisted and expanded through the COVID progress of SDG Target 4.4.1 (Youth and Adults crisis and the recovery period afterwards. with ICT Skills) reveals that only 2 percent Regional peers have continued to make of Filipino youth and adults can use basic sustained digital access improvements arithmetic formulas in Excel; 6 percent can while in the Philippines, the “jump” in fixed copy and paste in a document; and 7 percent broadband penetration during the COVID can attach a file to an email. Across most of crisis turned out to be merely temporary these activities, Filipinos are by far the least (Figure 3). Lower income households skilled in the region (Figure 2). While broader are increasingly left behind. In the least educational system challenges need to be populated areas and remote islands, tackled, improvements in skills for future progress in household Internet penetration jobs are unrealistic if children are unable to over the last 10 years has only been a digitally connect. Similarly, for industries, weak third of progress made in populated urban internet8 translates to firms’ weak and uneven centers (Figure 4). If the pattern remains technology adoption,9 limiting innovation and unchanged, digitalization will not contribute productivity growth (World Bank 2022, 2023). to poverty alleviation and unequal access to For example, the IT-BPO industry, which has opportunities will be exacerbated. been the driver of employment and service export growth, is still constrained in quite a For faster economic growth, broadband few locations—industry survey participants infrastructure has been identified as among revealed expansion plans in more than 10 the most significant gaps, along with ease cities beyond NCR, Cebu, and Davao with of doing business. For the Philippines, fully limited peri-urban digital connectivity.10 catching up with Malaysia in telecoms infrastructure and business regulations over Unfortunately, the Philippines has been the 25 years is estimated to contribute 2.75 last investment destination among major percent of the 3.21 percent change in annual ASEAN economies for hyper-scaler cloud GDP growth (Figure 5).15 service providers such as Amazon and Google.11 2 skills 2. Regional Figure Fig 2 – ICT comparison of proportion of youth and adults with ICT skills Using basic arithmetic formula in a spreadsheet 40 60 Using copy and paste Transferring files 27 59 tools to duplicate or between a computer 51 54 move information and other devices 25 with a document 46 45 16 25 27 11 21 21 9 6 10 6 2 6 1 2 5 3 7 9 3 16 7 18 8 26 26 37 38 Finding, downloading, 25 45 53 installing, and 58 configuring software Sending emails with 30 attached files 39 Creating electronic presentations with presentation software Singapore Malaysia Thailand Indonesia Viet Nam Cambodia Philippines Source: UN SDG Indicators Database, 2019 data for comparability12 (accessed December 16, 2023). Fig 4 – BB benchmarking trend Figure 3. Broadband access trend, across ASEAN and within the Philippines Access improvement selectcountries, in select ASEAN Access improvement, 2013-2022/23 ASEAN countries Access improvem Fixed broadband penetration Active mobile broadband subscriptions Active mobile broa (% of households, TeleGeography) (per 100 inhibitants, ITU) (% of households 120 70 60 100 50 80 43.2 43.8 40 60 30 40 16.1 20 16.0 8.1 20 3.8 10 7.4 4.4 1.6 1.9 0 0 2013 15 17 19 21 23Q3 2013 15 17 19 21 22 2019 202 Malaysia Thailand Indonesia Viet Nam Cambodia Philippines Source: TeleGeography and ITU. 3 tration Figure 4. Broadband penetration, per population quintile by municipal population density Fig 5 – CEM Malaysia catchup Source: PSA Census of Population and Housing, 2010 and 2020; NEDA PDP 2023-28. Figure 5. Impact on GDP annual growth of Philippines fully catching up to Malaysia in 25 years, in percent 0.1 0.0 3.2 0.1 0.3 1.1 1.7 Business Telecom Financial Labor market Trade Top tax rates Total regulations Infrastructure development regulations openness Index Source: CEM3.0 Growth Ambition and Structural Reforms Tool. See Annex 1 for data and methodology. 4 2. PROBLEM: Market and Regulatory Failures 2.1 ‘DIVIDED’ MARKET STRUCTURE AND UNDERINVESTMENT IN RURAL INFRASTRUCTURE The broadband market in the Philippines is an providers (see Figure 6) acting as retailers, effective duopoly with two large telcos. Figure rely on wholesale infrastructure, either 6 shows the distribution of digital connectivity from the duopoly or market challengers market participants in the country. Globe and (such as Converge and InfiniVan). Such PLDT/Smart (PLDT) are vertically integrated. challengers, however, are still small. The They own international connectivity, backbone, third mobile operator, DITO Telecommunity, middle- and last-mile networks (see Box 1) and started operating in 2021 and has gained a have the majority subscriber share. Over 500 9-percent mobile subscriber share as of 2022. Fig 6 – Market participants Figure 6. Telecom and digital connectivity market participants in the Philippines Source: Author’s elaboration based on Serafica and Oren (2022) and company financial statements. Box 1: Broadband Value Chain The modern broadband internet is comprised of several segments under different operators, from international submarine cables to final consumers (Figure B1.1). Key segments of the broadband value chain include: International link. The international link connects a country to the global Internet. Most of the Philippines’ international connectivity goes through international submarine fiber optic cables, which land in one of several cable landing stations located mostly in Luzon and one in Davao in Mindanao. 5 National Backbone Network. From the cable landing stations, bandwidth is transported to telcos’ transmission centers, located mostly in the National Capital Region (NCR). The bandwidth then gets picked up and is transported throughout the country before distribution by service providers over the national backbone or core network. The Philippines has multiple private and public backbone networks. Middle mile. Also commonly referred to as backhaul, the middle mile network connects the domestic backbone to internet service providers (ISPs) in provinces and/or cities and municipalities through Points of Presence (PoP). The middle mile connects individual communities to the greater national, then to the international network. Growing the middle-mile segment beyond urban centers is the typical bottleneck in a geographically challenging environment. Internet exchange points (or IXPs, which allow exchange of local internet traffic), Content Delivery Networks (or CDNs, which are geographically distributed servers coordinated for fast delivery of internet content), and data centers (host servers containing digital content and applications, including by CSP: Cloud Service Providers) are typically located between the backbone and middle mile networks. IXPs and CDNs help facilitate services while limiting latencies. Last mile. From the middle mile, connectivity is brought to the Last Mile—also commonly referred to as ‘access network’—to distribute bandwidth to end users via computers, phones, mobile, and other devices. Last mile connections are provided through 3 types of technologies: wired, wireless, satellite. • Wired connection is most typical for fixed broadband internet and includes (i) digital subscriber line (DSL: utilizing analogue phone line for digital data transmission); (ii) coaxial cable (typically utilizing cable TV systems); and (iii) fiber optic cable (FTTH: fiber-to-the-home). These cables can be built either underground or aerially, and construction involves civil work (ducts for underground, or poles for aerial wiring). • Wireless connection is the primary modality for mobile broadband internet and a growing modality of fixed broadband nternet through Fixed Wireless Access (FWA). Fiber middle-mile ends at towers, from where radio i link connects antennas to user devices (mobile phones, or wireless home routers for FWA). • Satellite connection is used both for fixed broadband and mobile broadband, which can also be used for filling middle-mile gaps (often by connecting to the domestic backbone or, in some cases, providing international connectivity, while still requiring earth station infrastructure on the ground). As technology advances (e.g., 5G and LEO), roles of wireless and satellite grow in bridging connectivity gaps particularly in rural areas. For example, 5G FWA is estimated to be a cost-effective alternative to FTTH, with at least 100 Mbps download speed, under certain conditions (urban/rural, market share, presence of ducts/poles) saving up to 80 percent of 10-year total cost of ownership in Southeast Asia (GSMA Intelligence 2022). 6 Figure B1.1: Broadband infrastructure Value Chain, Technologies, and Reach Infographic Source: World Bank staff elaboration based on ITU (2020). The Philippine broadband provision is ‘divided’ one-third to more than one-half of revenue between urban and rural markets. Higher- per year to capital expenditure (CAPEX). income consumer segments and businesses in Small ISPs are estimated to have invested urban centers drive broadband demand growth less than 20 percent of revenue to CAPEX which is served by large telcos. Hundreds in the last 5 years, while cost of wholesale of small internet service providers (ISPs), bandwidth. Small ISPs are estimated to have including cable operators, push the internet invested less than 20 percent of revenue to rural areas where major telcos offer limited to CAPEX in the last 5 years, while cost of mobile broadband (4G) coverage and do not wholesale bandwidth increased, reaching provide fixed broadband services. While small one-third of revenue in 2022 (Table 1).20 ISPs are vibrant growing businesses (e.g., cable operators grew driven by internet service), Aggregated at the province level, PSA’s 2020 their investments have been limited. Large Census of Population and Housing shows and small players have different financials. On that the National Capital Region (NCR) was the one hand, Globe and PLDT together grew the only market where more than 30 percent revenue by roughly 30 percent from 2017 to of households had access to fixed wired 2022, earning roughly 50 percent of EBITDA broadband, while in the “less connected” (Earnings Before Interest, Taxes, Depreciation, provinces, accounting for two-thirds of all and Amortization) margin. On the other hand, provinces and 36 percent of the national small ISPs grew revenue by more than 80 population, fixed wired broadband reached percent but earned 16 percent of EBITDA less than 10 percent (Figure 7). margin. Large telcos are known to invest 7 Table 1: Market performance of Globe/PLDT and small ISPs Key metrics, 5 years (2017-22) Globe/PLDT Small ISPs Revenue growth, % 29 83 EBITDA, % of revenues 49 16 Average profit Net income, % of revenues 14 4 CAPEX (% in aggregate revenue) 47 17 Source: World Bank staff calculations based on financial statements from the Philippine Stock Exchange and Securities and Exchange Commission. Figure 7. Broadband internet penetration by connectivity types and provinces g 7 – Provincial Source: World Bank staff calculations based on data from the PSA 2020 Census of Population and Housing. Note: Bars are sorted by total access to fixed broadband. Bar width is weighted by population. Only NCR and the top 5 populated provinces by island group are labeled. 8 The broadband internet market in the proxy for accumulated last-mile investment Philippines is the most concentrated, most through fiber-to-the-home (FTTH) and profitable, while least invested, in the other fixed-wired technologies, was 8 in the region. For large telcos, the Philippines is Philippines, 12 in Malaysia, 18 in Thailand, an exceptionally lucrative market, with high and 22 in Viet Nam. prices and low competition. Among all ASEAN members and neighbors, the Philippines The Philippines has under-invested in alone has two operators dominating with 90+ broadband infrastructure compared to percent of mobile market share and earned regional and global peers. Over the last 15 more than 50 percent of EBITDA margin in years, around 100 developing countries have 2022 (Figure 8). Malaysia’s, Thailand’s, and invested at least 1 percent of GDP in telecom Viet Nam’s top two mobile broadband players infrastructure for at least 1 year (in many accounted for around 80 percent of mobile countries, for several consecutive years) while market share, with EBITDA margin around 30 the Philippines has not done so; investments to 40 percent. In the Philippines, accumulated slowed down from $2.2 billion (0.64 percent mobile last-mile investment measured by of GDP) in 2018 to $1.8 billion (0.44 percent tower density, was the lowest at 28 towers of GDP) in 2022. Regional and global peers’ per 100,000 population. In the ASEAN, trajectories show more investment and faster middle-income peers had 40 to 100+ towers coverage roll-out, across fixed and mobile per 100,000 population. Fixed broadband broadband (Figure 9). subscriptions per 100 inhabitants, as a Fig 8 – Mobile markets Figure 8: Top operators’ market share, profitability, and infrastructure, comparison in Asia and Pacific Mobile subscriber share of top operators Operators profitability Mobile tower density (%, 2022) (%, top 2 average EBITDA margin, 2022) (Towers per 100,000 population, 2022) #1 #2 #3 Others #4+ Japan 37 27 22 15 31 175 Myanmar 38 28 20 14 N/A 48 Singapore 46 24 23 7 26 N/A Indonesia 44 29 16 12 48 42 South Korea 43 32 26 27 58 Thailand 45 33 21 2 43 76 Viet Nam 53 26 18 4 31 92 Malaysia 57 22 17 4 42 130 Cambodia 42 38 18 3 N/A 66 Brunei 49 33 18 N/A 135 Australia 53 31 16 30 68 Laos 47 42 11 N/A 131 China 56 35 10 33 144 Philippines 52 39 9 51 28 Source: World Bank staff calculations based on companies’ financial statements and data from TowerExchange. 24 9 Fig 9 – Infra investments Figure 9: Infrastructure investment and broadband adoption, the Philippines and select peers Fixed broadband penetration Active mobile broadband subscribers (% of households; 2009-21; TeleGeography) (Per hundred inhabitants, 2013-22, ITU) CHN MYS THA 100 120 IDN CHN VNM KHM GEO 100 SEN 80 SRB MAR SUR PER BOL MDA 80 CRI 60 BRA SLV PHL TUN 60 MYS 40 MNG 40 PHL 20 20 0 0 0 2 4 6 8 10 12 14 0 2 4 6 8 10 Investment in telecommunications Investment by mobile operators (Cumulative % of GDP, 2009-21, ITU) (Cumulative % of GDP, 2013-22, GSMA) Source: World Bank staff calculations based on TeleGeography, ITU, and GSMA data. 2.2 INADEQUATE AND LAGGING POLICY FRAMEWORK Among ASEAN nations, the Philippines is Four countries with scores that decreased and the least favorable on policy environment were lower than the Philippines in 2021 were for affordable broadband, and among the Venezuela, Nicaragua, Haiti, and Sudan. ASEAN slowest in the world at promoting reforms peers made constant progress; for example, for affordable broadband, according to ADI Malaysia on competition, Indonesia on (Affordability Driver Index) policy score. Among infrastructure sharing, Viet Nam on spectrum, 57 developing countries assessed in the ADI and Cambodia on universal access. In the policy score from 2017 to 2021, the average Philippines, reforms stalled or even reversed in score of the 5 policy areas assessed (national all of these areas (Figure 10). broadband policy, competition, infrastructure sharing, spectrum policy, and universal access) declined in the Philippines and improved in 40 countries, including all other 6 ASEAN countries assessed, and countries with archipelagic or other challenging terrains. Scores in another 12 countries declined but were higher than the Philippines in 2021. 10 Fig 10 – ADI Figure 10: Affordability Driver Index, individual policy scores, and trends in select ASEAN countries Broadband Policy Competition Infrastructure Sharing Spectrum Universal Access 10 8 6 4 2 0 2015 2021 2015 2021 2015 2021 2015 2021 2015 2021 Malaysia Thailand Indonesia Viet Nam Cambodia Myanmar Philippines Source: A4AI Affordability Drivers Index (accessed December 12, 2023). Note: Data for 2015-21 (no data for 2016). See Annex 2 for more data and methodology. Box 2: Policy choices, sequencing, and trade-offs in addressing connectivity gaps Bridging the urban-rural digital divide to harness digitalization for inclusive growth, is a challenge many countries face. In pursuing the objective of universal broadband while adapting to changing technologies and business models, countries must weigh three policy choices and trade-offs (Oughton et al. 2021; World Bank, forthcoming): 1) Technology deployment. Although policies are generally technology neutral, governments need to consider a range of possible options (such as fiber, 4G or 5G wireless, satellite) to choose feasible coverage targets reflecting the level of capacity to be provided, the speed of roll-out, and the cost of network deployment. 2) Competition and cooperation. Competition is generally desirable to create a dynamic and efficient market. However, in remote geographic areas demand may not be high enough to support more than one infrastructure provider, and infrastructure sharing facilitated by supportive regulations may allow network expansion at significantly lower cost. 3) Fiscal policy options. These involve tensions between maximizing government revenues through policy for pricing spectrum and incentivizing or obliging population coverage to minimize the need for public subsidy to reach universal access goals. Global studies and experience sharing over the last decade—such as in ITU (2023), A4AI (2022), and Broadband Commission (2021)—have recommended policy approaches and good practices in pursuing and balancing these objectives. Sequencing of commonly recommended interventions (Muente- Kunigami et al. 2010) are as follows: (i) tackle supply side reforms for market to facilitate cost reduction (affordability for consumers) and infrastructure roll-out (network coverage) through competitive environment (efficiency and innovation); (ii) address demand side gaps to foster adoption (e.g., upskilling, relevant local content, device affordability); and (iii) channel public investment to narrow residual gaps (e.g., subsidy models such as viability gap funding, universal service funds) (Figure B2.1). 11 among the tools useful for assessing countries’ policy alignment with good practices and reform Box 2 Figure progress and identifying gaps against peers or benchmarks. Figure B2.1: Areas of intervention to promote affordable broadband Source: Adopted from Broadband Commission (2021). Market reforms to promote competition and operators’ conduct. Recent economic reforms, investment in broadband infrastructure should such as on removal of foreign ownership be an immediate focus. “Broadband policy” restriction in 2022 and introduction of as assessed by ADI (i.e., National Broadband the competition law in 2015, have not Plan) scores relatively well compared to other significantly altered the market dynamics in ADI areas, but the market environment is not the broadband sector under the prevailing conducive to broad private sector participation legal system, originating in 1930s and 1990’s for effective implementation of such plan (see Figure 11). Key regulatory weaknesses including in rural areas. “Universal access” include: policies, typically implemented through Barriers to market entry and investment: • obligations to operators and/or dedicated Requirements for legislative franchise funds, may not be achievable under the current for building and operating broadband licensing and spectrum assignment framework infrastructure; (see Box 3). Unlevel playing field: Licensing framework • favoring incumbent telcos, rooted in Binding constraints to competition and legislative franchise, and the absence of investment in the Philippine broadband ex-ante interconnection price regulation infrastructure are inter-related, with disadvantageous to small ISPs; cumulative effects of policy actions (or inactions) reinforcing the market structure and 12 • Ineffective infrastructure sharing policy Reforms for rural broadband infrastructure framework: Lack of infrastructure sharing in deployment could aim for ‘doubling’ practice exacerbated by weak incentives in investments. As done in many rapidly the market structure (vertically integrated progressing countries, 1 percent or more of duopoly), excessive administrative GDP deployed for telecom investment per burdens, and reinforced barriers to rural year can be seen as an indicative benchmark. infrastructure investment; and The investment gap in the Philippines is Rigid and non-transparent spectrum • around US$ 2 billion or PHP 110 billion per management: Limiting flexible and efficient year. The government’s direct investments use for better connectivity, also rooted in and fiscal measures—such as the 2024 franchise requirement and undermining DICT budget appropriation of around PHP 5 government revenue potential. billion—cannot be seen as the primary means to fill the. However, as will be discussed later, the government can find ways to complement pro-competition policy reforms with direct public investments in justifiable, efficient, and sustainable ways. Figure 11: Legislative and administrative landscape for Philippine broadband infrastructure Entry & Spectrum Competition, Shared infrastructure investments policy level playing field deployment Barriers Foreign ownership Legislative franchise Rigid radio frequency spectrum (re-) assignment Anti-competitive pricing in wholesale Uncontrolled mergers Weak (passive & active) Unclear rules in Permits, certificates, restriction requirement bandwidth competition infrastructure access to RoW for enforcement sharing govt infra infra assets construction Reforms 2022 PSA 1931 Radio Control Law 2015 2021 EO 2020 Common 2020-21 JMCs on amendment 1995 Public Telecom Act Philippine on satellite tower policy streamlining Compe- liberalization permits / 2023 EO tition Act Reforms benefitting... ... for whom? New or foreign entrants Smaller market participants Incumbents Source: World Bank elaboration based on Ian Mackay’s Swiss cheese metaphor of the respiratory virus pandemic defense system. Illustration by Azel Lorena. 13 3. POLICY AND REGULATORY GAPS: Industry-Wide Challenges 3.1 BARRIERS TO MARKET ENTRY AND INVESTMENT The legislative franchise requirement is a between a few months to a few years. Critics major barrier to market entry and expansion point out that the tedious and lengthy of Internet coverage. In the Philippines, process fails to put a check on operator Republic Act (RA) 7925 or the Public behavior (Espos 2003). Telecommunications Policy Act of 1995 requires a network operator to secure a The telco-centric licensing regime subsists legislative franchise before getting a license despite significant changes in Internet from the industry regulator, the National technologies, service providers, and Telecommunications Commission (NTC). A business models. Technology has progressed legislative franchise entails Congress passing faster than regulations, which constrain a Republic Act. Approval may typically take 2 the country’s capacity to benefit from years and is subject to the political process. technological advancements. When the All countries with similar requirements have Internet became available in the Philippines already introduced reforms to remove the in the 1990s, the local exchange or landline franchise. The Philippines remains the last telephone network was used to establish holdout. a dial-up Internet service. Other Internet technologies have since emerged that do The political and quasi-judicial nature of not depend on the traditional telephone, the franchising and licensing regime creates but the laws remain unchanged. RA 7925 uncertainty and increases the cost of investing dictates that connecting to the Internet must in digital infrastructure. RA 7925 liberalized go through a telco, otherwise an ISP needs the industry and broke the monopoly of to get a telco franchise and license before it PLDT in the early 1990s but required public can put up its own network—even if it does telecommunications entities (PTEs) to: i) not intend to offer the telecommunications obtain a Congressional franchise, ii) secure a services described in the RA. As a result, an Provisional Authority (PA) to operate, and later, ISP is limited to offering Internet service only iii) get a Certificate of Public Convenience in areas with available telco facilities. and Necessity (CPCN) from the NTC. The law identifies telecommunications services based In rural areas, ISPs face an adversarial on legacy categories of international carrier, investment environment due to a restrictive inter-exchange, local exchange, and mobile licensing regime. Apart from telcos, there radio services. Some observers assert that are small ISPs offering Internet services in obtaining a legislative franchise does not their communities. For cable TV operators guarantee that the franchisee is technically that provide wired broadband,33 extending and financially capable (Espos 2003). facilities to more towns around their service Meanwhile, the granting of a PA/CPCN is a area would help connect low-income, low- quasi-judicial process carried out by the NTC population communities. But to lay fiber to en banc, similar to a court hearing (Barcenas pick up bandwidth from a telco facility, which and Serafica 2018). This can take anywhere is usually located in the municipal center 14 several towns away, a cable operator is administratively or financially practical. This required to: (i) secure a PA, with an obligation imposes severe constraints on investments to operate in each town passed;34 (ii) secure a by ISPs and results in poor digital Congressional franchise to operate as a telco; infrastructure—also exacerbated by the or (iii) arrange with a telco that will build the limited sharing of passive infrastructure (as middle-mile infrastructure for an ISP. 35 In most will be discussed). rural markets, none of these options are 3.2 UNLEVEL PLAYING FIELD High barriers to entry have historically in investments by smaller providers who encouraged the sector to consolidate and could have otherwise injected fresh capital discouraged small players from growing into the broadband sector and expanded (Ortiz et al. 2017). The franchising and connectivity in unserved areas.36 licensing regime favor the business models of large, vertically integrated PTEs. Mergers There are no regulations preventing anti- and acquisitions, and the absence of a competitive price discrimination. While there comprehensive competition law prior to are NTC rules mandating interconnection 2015, led to two dominant telecom players between operators and across segments,37 accumulating the digital infrastructure assets it is still a commercial agreement. The NTC of other telecom companies, including radio approves the interconnect agreement and spectrum. In the absence of measures to may serve as arbiter if there is refusal by one prevent abuse of market power (such as anti- party to interconnect. However, the regulator competitive wholesale bandwidth pricing), the is only informed about the existence of duopoly has stifled competition in broadband an arrangement, as part of the value- value chain segments, especially in middle- added service (VAS) license application; and last-mile. pricing is not disclosed in the standard form submitted to the NTC. Small players The lack of a level playing field constrains find that connecting local communities to emerging and small players’ ability to deploy the nearest local telco facility costs more infrastructure and provide affordable internet than connecting to a data center overseas. services. Small ISPs rely on the telcos for Even if an interconnection agreement is wholesale bandwidth and facilities leased by signed, implementation and compliance large telcos; they also compete with the telcos are a different matter altogether. Dispute in the same retail market. As infrastructure settlement is usually the only time the NTC owners and wholesalers, the large telcos have intervenes by adjudicating the claims for an advantage on access and pricing, and small violations of an interconnect agreement. ISPs have little power to negotiate for a fair price. The restrictions on who can own and operate infrastructure have prevented growth 15 Asymmetric regulation,40 which can help level new players would otherwise find difficult to the playing field, is absent in the Philippines. duplicate at the same scale41 or enough to In other countries, a dominant operator must pose a market challenge. No such obligation provide access seekers open access to its is imposed on incumbent network operators network and other essential facilities, which in the Philippines. Box 3: What is Universal Service, and do Philippine operators have such obligations? The International Telecommunications Union (ITU) Information and Communications Technology (ICT) Regulatory Toolkit defines universal service as follows: “when every individual or household can have service, using it privately, either at home or increasingly carried with the individual through wireless devices.” For developing countries where full universal service may not be realistic, universal access goals may instead be defined in terms of the “proportion of the population that can afford private service.” To achieve universal service, many countries impose universal service obligations (USO) on operators as a condition for their registration or license, including for spectrum, or otherwise mandate the contribution of some amount—normally a proportion of revenues—to a universal service fund (USF). USOs may take the form of a requirement to cover a specific proportion of the population or population centers, or to provide access to designated unserved and underserved areas. A USF aims to meet the same goals either directly through infrastructure investments, or indirectly through grants and subsidies to operators. Among ASEAN middle-income countries, the Philippines alone does not impose USOs on operators as a condition of their license, nor does it mandate contributions to a USF. Indonesia, Malaysia, Thailand, and Viet Nam all operate USFs. Viet Nam, for example, requires a contribution of between 1 to 3 percent of annual revenues to the Viet Nam Public-Utility Telecommunications Service Fund (VTF) under the Ministry of Information and Communications. Investments by the VTF include, among others, connecting schools and hospitals, and providing connectivity for communities. During the landline telephone era, Philippine licensees were required to deploy a fixed number of local exchange connections. Telecom operators now no longer have service obligations, with the notable exception of DITO Telecommunity, which must meet service coverage targets for the first 5 years of its operation. 16 3.3 INEFFECTIVE INFRASTRUCTURE SHARING POLICY FRAMEWORK Lack of passive infrastructure sharing policy For underground cables, information and framework makes network deployment coordination are insufficient for shared costly and inefficient. In other countries, deployment and dig-once. The civil works policies on shared or common use of needed to dig up and restore roads for passive infrastructure (see Box 4) have laying fiber cables constitute as much as helped cut the cost of civil works, reduce 80 percent of the cost of deploying fixed- the number of required permits, and reduce broadband network. The number of permits the environmental and social impact of and time needed to secure right of way for network deployment. A comprehensive policy each operator, and little coordination across framework mandating various government infrastructure sectors, causes inefficiencies, agencies involved in infrastructure (i.e., which also increases cost and delays buildings, roads, bridges, railroads, power, water, service. Utilities, telcos, and cable operators etc.) to coordinate activities, share information, do roadworks on their own, one after the and issue rules and regulations to facilitate other. The lack of information on existing the shared use of infrastructure by multiple infrastructure constrains efficient public and operators—including data transmission service private sector planning. There needs to be providers—could outline the institutional a “dig-once” policy requiring coordination arrangement and provide the legal basis for among government agencies (including infrastructure sharing activities. LGUs) and providers, like in other countries (World Bank 2020). For mobile, common towers are still limited. The economic impact of tower sharing in the For aerial cables, harmonized policy is Philippines is estimated among the highest missing for common poles. The lack of in Asia (Roland Berger 2022), given that policy to harmonize the sharing of poles constructing and operating cell sites can take makes wired broadband deployment slow, up as much as 50 percent of a Mobile Network cumbersome, and expensive. There is little Operator (MNO)’s CAPEX and up to 60 percent coordination between the Energy Regulatory of operational expenditure (OPEX) (World Bank Commission (ERC) and the National 2020). The 2020 common tower policy has Electrification Authority (NEA), which encouraged the entry of independent tower regulate the electric distribution companies companies (ITCs), to which MNOs sell tower and cooperatives that own most electric assets or contract out the operation of some poles, and the NTC, which regulates the new towers. After 3 years, however, new builds PTEs and cable ISPs. For pole attachment, are limited to a little over 1,300, of which an agreement is signed between the pole only 1 percent are shared by multiple MNOs. owner (lessor) and the PTE or cable operator Enforcement can be improved by overcoming (lessee). However, the agreement can change legal challenges, mandating data-sharing of any time without the lessee being informed. planned cell sites, and promoting the use of It is also unclear which authority should ITCs for future installations. mediate in case of a dispute. Some wired 17 broadband providers opt to erect their own private associations. The Bayanihan 2 poles,47 resulting in slow and costly network Law, enacted in 2020 to support COVID-19 expansion and unsightly multiple poles in response, allowed (i) the temporary the one location. Harmonizing rules and suspension of requirements to secure permits regulations and the roles and responsibilities and clearances and (ii) the streamlining of of regulators across sectors could help set up relevant regulatory processes and procedures. a fair and reasonable arrangement for shared To implement these reforms, the Anti-Red pole use. Tape Authority (ARTA) led the issuance of joint memorandum circulars (JMCs) on Streamlining of permitting process has streamlining the permit approval process. started but is still insufficient. Across network After the Bayanihan Law expired, President segments, a major challenge for telcos and Marcos issued Executive Order 32 in 2023, ISPs is the long list of permits and clearances which continues the streamlining reforms required by national government agencies but with limited coverage. (NGA), local government units (LGU), and Box 4: Infrastructure Sharing – policy trade-offs and practices Infrastructure sharing is defined as “various kinds of arrangements by which an owner of [ICT] network facilities (including but not limited to, antennas, switches, access nodes, systems, ducts, poles, towers, premises and rights of way) agrees to share access and usage of those facilities with another legal entity, normally another network operator or service provider, subject to a commercial agreement between the parties.” The need for infrastructure sharing emerged as investments in fiber broadband and 5G networks grew, coupled with a shrinking and highly competitive market. Documented benefits of infrastructure sharing include: (i) reduction in OPEX and CAPEX and release of capital assets, (ii) lower barriers to entry, which increases service-based competition and (iii) lower asset duplication, which enables efficient use of scarce resources, such as real estate, and lessened environmental impact (Deloitte 2015). Types of infrastructure sharing and competition considerations Various arrangements support the benefits of infrastructure-sharing while considering their potential impact on competition and incentive to invest. The essential aspects of sharing infrastructure include: (i) who owns the infrastructure; (i) who is granted access to the owner’s infrastructure; (iii) what facilities are shared; and (iv) what form the commercial agreement will take.  assive infrastructure sharing refers to the sharing of non-electronic infrastructure, such as towers, •P poles, ducts and premise, and of physical space (e.g., buildings, sites, and masts), whereby active network components remain proprietary and separate, not requiring operational coordination between network operators. Active infrastructure sharing involves electronic infrastructure, such as switches and radio access nodes. In mobile networks, sharing may involve the active layers of a network such as antennas, the 18 • In one-way infrastructure sharing, an operator or infrastructure owner (access provider) leases space to another operator (access seeker). Where there are multiple access seekers, or where the access provider has significant market power (SMP), the regulator should ensure that the terms and conditions are cost-based, preferably through a published access price list. In reciprocal sharing, two operators with different geographic coverage agree to share access to their network in areas not covered by the other operator, which may not involve charges. In this case, the regulator should look out for the possibility of collusion or anti-competitive behavior between the two operators resulting in the exclusion of another service provider (Digital Regulation Platform 2022). Infrastructure can be owned and operated by a third party or a joint venture (JV), such as an ITC. The • regulator may need to investigate the financial arrangements of such arrangements to ensure that there is no undue advantage to the operator divesting and outsourcing operation of assets. In case of a JV, where parties sharing the infrastructure consolidate, own, and operate the network while retaining control of the assets and the transfer prices paid for their usage, the regulator should ensure that the arrangement is not exclusionary or otherwise anti-competitive (Digital Regulation Platform 2022). Regulatory Framework Infrastructure-sharing is usually done to increase coverage, improve quality, and lower prices for consumers. It can be mandatory or voluntary; the regulator may select which passive or active infrastructure is mandated for sharing. Infrastructure-based competition among operators can emerge given a regulatory framework that provides for two access remedies: (i) symmetric access remedies that apply to all market players, and (ii) asymmetric access remedies that apply only to operators with SMP (Centre on Regulation in Europe, 2020). The regulatory framework can also impose obligations on non- operators, such as water or electricity companies, operating passive infrastructure (e.g., ducts, masts, or poles) that data transmission networks can utilize. Access obligations are typically done via commercial negotiations. The regulator gets involved only in dispute resolution in case negotiations between the parties fail. However, the regulator can draw up model agreements and/or general principles of agreements, which can include: Fair, reasonable, and non-discriminatory (FRAND) access prices; as much as possible, cost-based 1.  pricing for operators with SMP. Guaranteed strategic independence of each party, to mitigate against the disincentive for the access 2.  seeker to invest in its own infrastructure and compete across a broader facilities base. Guaranteed wholesale access to third parties, allowing each party to offer wholesale access to its 3.  own infrastructure, and to its portion of joint infrastructure, as long as doing so does not adversely affect the quality of services to the other parties. Exclusivity, while necessary to make the agreement attractive to both parties, kept at a minimum. 4.  Protection of investor against opportunism from later co-investors, ensuring that there is greater rate 5.  of return for the original investor than for later co-investment, where sharing is mandated. Information exchange kept at minimum necessary, and termination rules detailed and explicit (Digital 6.  Regulation Platform 2022). There are generally accepted principles for infrastructure sharing regulations. 57 Some countries require operators to publish information on passive infrastructure sharing opportunities.58 Others require thecoordination of site planning among operators or a more targeted approach to incentivize new 19 players, such as requiring spectrum-license holders to provide roaming service for new entrants or to share spectrum with operators that have less spectrum holdings (BEREC 2018 survey in the European Union).59 4. POLICY AND REGULATORY GAPS: Spectrum Management Deep-Dive 4.1 OUTDATED LEGAL FRAMEWORKS Radio spectrum management in the radio station to those with a permit from an Philippines is based on analog-era law enacted executive authority. This has persisted for 92 in the 1930s.60 Act 3846 or the Radio Control years despite technological advancements Law of 1931 is the primary law governing in wireless technology, diverse wireless spectrum management, limiting access of digital services in various fields, and evolving licensed radio frequencies (see Box 5) to business models in the telecommunications entities with a legislative franchise from and broadband industries. Congress and the construction of a Box 5: What is radiofrequency spectrum? Wireless devices, from the earliest telegraph to the latest mobile phones, all depend on access to the radiofrequency spectrum, also known as radio spectrum or simply spectrum found in the atmosphere. Spectrum encompasses electromagnetic frequencies from 1Hz to 3000 GHz used to wirelessly transmit information. This range of radio frequencies are divided into frequency bands. Radio frequencies are finite—we cannot make more of them—albeit inexhaustible resources that can be used and reused without being depleted. Different technologies require different frequencies, although more than one technology can utilize the same band. For example, Wi-Fi routers and wireless headphones operate on the 2.4 Gigahertz (GHz) band. However, key mobile spectrum bands are necessarily assigned on an exclusive basis. Spectrum is considered the patrimony of the State. While entities may be assigned and pay for the right to hold or use frequencies, they do not own them. The use of the same frequencies by two or more entities can result in harmful interference, leading to a degradation or loss of performance for users. For this reason, the State ensures that interference is mitigated, and spectrum is used efficiently through what is called spectrum management. 20 Spectrum management involves the allocation, reallocation, assignment, reassignment, recall, sharing, and co-use of frequencies. As a member of the ITU, the Philippines allocates frequencies to technologies following international standards; spectrum management is carried out by the NTC, and spectrum allocation to a particular service is published in the National Radio Frequency Allocation Table (NRFAT). Telecommunications liberalization in But even as the technology changed and the 1990s focused on propagating fixed new players and business models emerged, telephones and gave less priority to spectrum the NTC’s rules and regulations on spectrum management. In 1995, RA 7925 or the Public remained anchored on outdated, restrictive Telecommunications Policy Act provided laws. the NTC general guidelines for spectrum management, through the following The current legal framework does not provisions: address key policy gaps in spectrum management. RA 7925: (i) does not define Radio frequency spectrum is considered 1.  “efficient and effective” use of radio spectrum, a scarce public resource that shall be leaving this to the NTC’s discretion; (ii) administered in the public interest and in has no provisions on spectrum recall, accordance with international agreements refarming, reassignment, and joint use; (iii) and conventions (Sec. 4.); fails to mandate transparency in spectrum Radio spectrum shall be granted to the best 2.  management; and (iv) does not hold qualified service providers who will (i) use spectrum holders accountable for their use radio spectrum efficiently and effectively to of a scarce and valuable resource. The law meet public demand and (ii) may avail of provides for periodic review of spectrum new and cost-effective technologies (Sec. 4.); allocation and assignment, but does not Radio frequency spectrum allocation and 3.  specify regularity. In practice, spectrum is assignment shall be subject to periodic assigned administratively, although the review (Sec. 15); law allows for “open tenders”62 and EO 307 The use of radio spectrum is subject to 4.  s. 2000 prescribes “auction” for spectrum reasonable spectrum user fees (Sec. 15); assignment. Absent an updated legal Where demand for specific frequencies 5.  framework, the licensing of radio spectrum exceeds availability, the NTC may hold open is unlikely to change, resulting in spectrum tenders and ensure wider access to radio management that is unable to promote spectrum (Sec. 15). [emphasis added] competition, respond to technological advancements, and adapt to the evolving The gaps in spectrum management only markets. became apparent as cellular mobile services began to overtake landlines in the 2000s. 21 4.2 UNCLEAR PROCEDURES IN SPECTRUM RECALL AND REASSIGNMENT Lack of clear spectrum policy framework The lack of fixed terms in spectrum has led to rigid, non-transparent spectrum assignment and practical absence of a recall/ management and fragmented assignment. review/reassignment process have made First, unlike in other countries, the Philippines frequency assignments in the Philippines allows spectrum to be held by an assignee almost de facto permanent. The absence indefinitely, without any service obligation of spectrum caps and a “spectrum screen” requirement. Spectrum assignments are valid process to review the impact of assignments for the duration of a PTE’s Congressional on market competition have allowed entities franchise which is up to 25 years. The spectrum to be assigned a huge chunk of spectrum for license is typically automatically renewed once what is tantamount to an indefinite period the franchise is renewed, without the State of time.70 Mergers and other consolidations recalling the spectrum. In most Asia Pacific have enabled the concentration of spectrum countries,66 spectrum licenses have a defined resources among a few players, reducing validity period, after which the regulator market competitiveness. This has resulted recalls the spectrum and may reassign it via in a scarcity of assignable spectrum, which competitive bidding. Second, the Philippines discourages the entry of new wireless does not reserve spectrum for potential new players.71 PLDT/Smart and Globe, who have entrants.67 Although the NTC administratively acquired smaller telcos holding spectrum assigns spectrum, recall and reassignment throughout the years, have had effective of the same frequency should be subject to control over much of the assignable mobile a quasi-judicial process, by law. However, access spectrum. The limited availability in practice, a telco has often resorted to of frequencies has been cited by the spectrum co-use with another telco/spectrum Philippine Competition Commission (PCC) holder it has acquired, which needs a simple as a challenge for new players. The third approval by the regulator. This arrangement telco, DITO Telecommunity, for example, has effectively superseded the recall and had to undergo a special process to acquire reassignment process, but has contributed to whatever was left of the mobile spectrum the concentration of spectrum in the hands of in 2018, which were mostly used for 4G a few, thus stifling competition.68 Third, co-use and 5G services.72 The current practice fails agreements have also resulted in piecemeal to maximize the revenue that could be and fragmented assignment of spectrum generated from subsequent auctions and across several entities. Licensing contiguous reassignment of frequencies after licenses lots of spectrum would result in more efficient expire. It also fails to create an opportunity spectrum use and better quality of service, as to impose concomitant obligations for in other countries.69 renewal of spectrum assignments or non-recall of spectrum holdings, which is common in other countries. 22 The NTC assigns spectrum on a nationwide costs and lower revenues of deploying basis, which contributes to the scarcity of and operating infrastructure. Because of spectrum. The exclusive assignment of this, as well as the absence of service- spectrum to one entity expected to use it rollout obligations attached to spectrum nationwide prevents other players from assignments, operators tend to invest more offering wireless Internet in areas where in upgrading existing infrastructure in highly the assignee does not operate a network. urbanized areas rather than in expanding In other countries, the regulator assigns rural coverage. Also disadvantageous to new certain spectrum bands (especially at higher players is the fee structure which imposes frequencies) on a smaller geographical scope, (i) an SUF representing a much higher ratio which could be regional, sub-regional, or of the revenue than for incumbents and metropolitan/local. The archipelagic nature of (ii) service coverage targets for the initial 5 the Philippines makes regional, sub-regional, years. or local-level spectrum assignment seem appropriate. The geographic coverage of any The lack of information about the spectrum particular spectrum assignment is not required management process leads to a non- by law. competitive industry. The NTC does not typically publish spectrum assignments (until Spectrum pricing does not incentivize recently, mainly in response to freedom of operators to roll-out rural infrastructure. information (FOI) requests).75 Neither does it Spectrum holders must pay an annual publish a spectrum management framework spectrum user fee (SUF) based on per or outlook76 which contains the government’s frequency and base station. This computation spectrum management approach, as well as actively disincentivizes service providers market and technological trends that will from rolling out and expanding wireless likely shape the demand for certain spectrum. infrastructure in areas that will not bring the The absence of critical spectrum information highest profit margins. While there are some makes it difficult to identify potential differential incentives to deployments in non- investment areas and discourages new urban areas (i.e., a 20 to 33 percent reduction), players from entering the wireless space. they are immaterial given the additional 23 Table 2: Radiofrequency spectrum management – Philippines and other countries Key weaknesses in the Philippines International Practices Spectrum assignment License for spectrum is valid Validity in New Zealand is 18- period effectively indefinitely (franchise 20 years, Malaysia 1-15 years, for 25 years, typically renewed). Singapore 15-20 years, Viet Nam 15 years, Myanmar 12-15 years, Indonesia 10 years. Regional assignment Spectrum license is awarded to In India, license is on per-circle telcos nationwide, according to basis; Indonesia national or the scope of their congressional regional; Brazil national and franchise. regional; Australia, Canada and US national, regional, and sub- regional/local. Spectrum recall Spectrum recall requires a In 2018, Korea gave 3 operators 3 quasi-judicial process with due years to meet rollout obligations process, which virtually has never for 5G licenses. Regulator recalled happened for major telcos. licenses of 2 of 3 operators on 28 GHz as they failed to meet obligations. Spectrum pricing and SUF Pricing, per frequency and base Colombia exchanges spectrum fees universal access station, disincentivizes operators for rural coverage; Mexico provides from rural roll-out and more base “social use” concession in spectrum station construction to improve assignments (e.g., for community); urban QoS. Canada integrates spectrum fee regime with universal access policy. Transparency There is limited availability of India publishes any changes to spectrum assignment data to the spectrum holding, e.g., with auction public and no set time frame for results, per frequency bands and spectrum review. geographic areas. Overall framework “Efficient and effective use” of Thailand, India, Australia, US, UK, spectrum is not defined. Monitoring among others, publish a spectrum of spectrum use is not reported to management strategy or roadmap. the public. Source: APT 2021; Gallitto and Camargos 2021; Kochhar 2023; ISED 2020, 2023; FCC; Ministry of Science and ICT Korea 2022; Contreras 2023; A4AI 2019. 24 4.3 SUBOPTIMAL FISCAL AND ECONOMIC OUTCOMES OF SPECTRUM POLICY In general, spectrum policy involves a trade- (0.91 percent), and Thailand (1.57 percent).77 off between short-term government revenue Governments are generally advised against and long-term sustainable growth of the using spectrum to maximize revenue so as industry—financially sustainable operation not to harm the sustainability of, and investor of wireless communications and its efficiency appetite in, the telecom sector. However, and innovation effects on sectors that use the Philippine government is estimated telecommunications and Internet. This is to be collecting less than half of the real because the government raises revenues cost of spectrum to operators, while lack of and the mobile industry incurs spectrum transparency makes accurate assessment cost—reflected in mobile broadband cost—for challenging.78 access to spectrum bands according to the pricing and assignment regime that a country More broadly, the Philippines directly adopts, either administratively or competitively or indirectly taxes the telecom sector through auction or other mechanisms. significantly less than the country’s regional peers. Governments raise revenue from While SUF revenue has significantly increased, the telecommunications sector through the Philippines is still likely under-collecting spectrum revenue, universal service revenue from spectrum. The NTC’s SUF obligation payment, corporate income tax, collection almost tripled in 5 years, from among several channels; these represent around PHP 2.4 billion in 2017 to more than policy choices fit for country and sector PHP 6.7 billion in 2022. This fact reflects circumstances. The sum of the operators’ industry growth. However, the significance of payment to the government through the spectrum revenue measured as a percentage of three main channels, as a percentage of total government revenue, has been miniscule: the top two operators’ revenue, was 5 to 6 from 0.09 percent in 2017 to 0.14 percent percent in the Philippines, compared to 9 in 2022. This is, by far, the lowest spectrum percent in Thailand and around 13 percent in revenue rate among regional peers, such as Indonesia and Malaysia (Figure 12). Malaysia (0.34 percent), Indonesia 25 Fig 11 – Operator payments Figure 12: Operators to government payment through main fiscal channels, select ASEAN countries Government SUF/ Operators’ payment to governments auction revenue Total spectrum revenue Spectrum cost: fee paid or Universal service Corporate income tax paid (% of government revenue, 3-year rolling average) amortization accounted obligation payment (% of top 2 operators service revenue, (% of top 2 operators mobile/wireless (Estimated % of revenue, as set in 2021-22 average) segment revenue, 2022) respective laws) 6% of weighted 3 4.7 2.9 net revenues 5.9 2.5% of net income, 5.4 1.4 2.2 plus 7% VAT 2 1.25% of 6.2 0 1.3 gross revenue 5.6 1 1.5% of telco Amount not disclosed 1.5 2.8 service revenue 0 2016 17 18 19 20 21 1.9 0 3.6 Malaysia Thailand Indonesia Viet Nam Philippines Sources: National regulators’ financial statements and other government sources; WPC 2023, company financial statements. There is growing awareness that adopting In the Philippines, an argument could fiscal and regulatory regimes conducive to be made in favor of revisiting the size, lowering 4G/5G infrastructure investment composition, and use of revenues from the costs can enable universal broadband, globally sector to address infrastructure weaknesses. (ITU 2020; Oughton et al. 2023; World Bank, Increased transparency and flexibility in forthcoming) as well as in East Asia (Brewer spectrum assignment would be an initial et al. 2022). Some countries are rethinking step toward such deliberation in the years to competition, licensing, infrastructure sharing, come. and universal access in the context of fixed/ mobile convergence and new sources of productivity gains and benefits to citizens enabled by new technology (5G, IoT, NaaS).82 26 5. WAYS FORWARD: Improving Philippine Broadband Infrastructure 5.1 URGENT BROADBAND INFRASTRUCTURE POLICY REFORMS Interrelated policy and regulatory barriers quality, and affordability of broadband underlying poor internet in the Philippines internet. require a comprehensive package of reforms. An outdated franchising and licensing regime Partial solutions are unlikely to yield and rigid and non-transparent spectrum desired entry, investment, and sector management have discouraged connectivity performance outcomes. There is rich infrastructure deployment and efficient literature citing global evidence on the resource re-allocation. Broader barriers to impact of needed reforms on broadband market entry and investment, unlevel playing infrastructure investment and performance field, and the lack of infrastructure-sharing (Table 3), including through liberalization policy framework have led to inadequate of foreign investments. However, given the infrastructure and weak competition. broad and significant legacy constraints Interventions addressing these weaknesses do in the Philippines (reflected in the overall not involve choices or trade-offs; rather, they low ADI policy scores as seen in Figure reinforce and complement each other. Figure 10), policymakers need to consider 13 outlines how a set of complementary identifying and integrating a viable reforms can address barriers to competition program for meaningful change, rather than and unlock investments from diverse market implementing discrete improvements. participants to eventually boost access, Fig 12 – Theory of change Figure 13: “Theory of Change” for the needed broadband infrastructure market reforms Source: World Bank elaboration on key provisions in OADT House Bill No. 6. 27 For example, RA 11659 amended the Barriers to entry and investment. The 1)  antiquated 1936 Public Service Act bill will lower barriers by removing and allowed greater market access for the Congressional franchise (political foreign investors. Yet, in the case of the process), allowing the NTC to broaden telecommunications sector, almost 2 years the qualification of broadband network after RA 11659 was enacted into law, no providers, simplifying and rationalizing the foreign company has entered the Philippine licensing process by doing away with the market. To complement the liberalization PA/CPCN (quasi-judicial process). Instead goal of the RA, key remaining reforms include of the latter, an administrative review removing the legislative franchise, completely process will determine an operator’s delegating licensing of network operators technical capability, financial capability, from Congress to the NTC, and strengthening and other qualifications that are deemed the latter’s mandate as the data transmission necessary. sector’s regulator. Reforms are also needed to Unlevel playing field. The bill will 2)  address the gaps of Act 3846 and RA 7925. allow PTEs and non-PTEs alike equal For spectrum management, the NTC needs a opportunity to invest in broadband legal framework that is responsive to the way infrastructure; promote open access people communicate and connect today. To to facilities on a fair, reasonable, and cater to the unprecedented demand for high- non-discriminatory basis; mandate speed Internet, infrastructure sharing also interconnection among network operators needs to be promoted to facilitate efficient at various segments and interfaces at cost- investments. based charging, such that entities of any size may freely enter and exit the market; The Open Access in Data Transmission (OADT) and prevent dominance by any single bill is a promising and viable entry point for player or group of players. These will help the needed reforms, among several proposals83 a more diverse set of players propagate in Congress that seek to lower the barriers Internet, especially in the countryside to entry and encourage competition. While where big telcos do not reach households. the title “open access”84 seems to indicate a 3)  Rigid, inefficient spectrum management. narrow scope, the bill contains provisions that The bill will make wireless broadband can address the key constraints hindering technologies more accessible to everyone, Philippine broadband development: and promote flexibility and transparency in spectrum management. More players will thus be able to access spectrum resources 28 Table 3: Global evidence on the impact of key reforms Analyses covering key policy levers Impact Competition, open access ITU (2021): opening ICT sector to 14% increase in capital investment, foreign investment (145 countries, network coverage by 13%, 8% price 2008-19) reduction ITU (2021): moving from service Increase of mobile investment by and network specific licenses to 10%, network coverage by 9%, price convergent, technology neutral reduction by 5% after two years licenses Gruber and Koutroumpis (2011): Unbundling to introduce intra- network unbundling (167 countries, platform, inter-firm retail competition 2000-10) and reduce incumbents’ market power to increase network diffusion. Spectrum efficiency ITU (2021): possibility of performing 18% increase in mobile investment, voluntary spectrum sharing network coverage by 17%, price agreements reduction by 10% Jung and Katz (2022): Secondary Three policies jointly adopted > 36% market, technological neutrality higher investment, network coverage and possibility of spectrum sharing by 10%, price reduction by 6%, after agreement (145 countries, 2008-19) 2 years; dynamic effects result in enhanced outcomes for an extended period. Bahia and Castells (2022): Spectrum Additional 20 MHz of 4G spectrum > assignment volume, auction price and enhanced download speed by 1-2.5 awards timing (64 countries, 2010-17) Mbps (increase up to 15%); high spectrum cost negatively impacts coverage and quality. Infra-structure sharing Koutroumpis et al. (2023): Mobile Average increases in population network sharing (140 mobile operators coverage of between 2-9 percentage in 29 European countries, 2000-19) points versus the baseline IBEREC (2018): passive and active Passive sharing > 16-35% CAPEX, 16- - infrastructure sharing 35% OPEX savings Active sharing (excl. spectrum) > 33- - 35% CAPEX, 25-33% OPEX savings Active sharing (incl. spectrum) > 33- - 45% CAPEX, 30-33% OPEX savings Houngbonon et al. (2023): common Significant drop in the price of tower sharing transactions, 2008-20 mobile connectivity as well as an increase in availability and uptake of mobile Internet, especially by rural households and women Note: See Annex 3 for a summary of literature reviewed. 29 to connect rural communities and GIDAs as  bill was included in the common legislative well as greater competition in fixed wireless agenda (CLA) of the Legislative-Executive and mobile Internet. Development Advisory Council (LEDAC).88 Ineffective infrastructure sharing policy 4)  framework. The bill will promote a policy Toward the development of a consolidated framework for co-location in passive bill for Senate deliberation, it is imperative infrastructure, including towers, poles, for policymakers to consider the breadth ducts, and dark fiber; institutionalize the of reforms needed to unlock the country’s streamlining of the permitting process digital potential and bridge the digital that was introduced by the current divide. Essential provisions in the bill, such administration; and mandate various as removing the Congressional franchise regulators and government agencies to for ISPs’ network rollout, are worth coordinate the deployment of infrastructure, keeping without excessive compromise. such as roads, bridges, and railways, Its infrastructure sharing provisions are together with broadband networks. an extension of the executive policy championed by the President. Provisions on The Open Access bill seems to be gaining spectrum management in some versions momentum, and this is seen by many industry of the bill may need strengthening, such participants85 as crucial in strengthening as on regulatory autonomy and spectrum broadband infrastructure. While the House of sunset clauses. Improved connectivity, which Representatives has consistently approved86 can be facilitated by improved spectrum the OADT bill since the 17th Congress in 2016, management, is proven to bring significant the bill has not hurdled the Senate Committee benefits to the Philippines’ GIDAs (see Box approval for the past two Congresses.87 In 6). In the longer term, there are strong the current 19th Congress, the leadership of arguments for new radiocommunications the House of Representatives and the Senate legislation to replace the antiquated law have both filed their own versions of the Open from the colonial era and introduce a new Access bill. The past two PDPs (2016-22; 2023- scheme for regulating radiocommunications 28) mentioned the bill; in September 2023, the in the Philippines fit for the 21st century. Box 6: Connectivity in unserved remote area is proven to generate significant welfare impact In the Philippines, the population covered by a mobile-cellular network was 99 percent (of which 96 percent was at least 3G) in 2021, while household penetration of cell phones was lower at 93 percent, especially in low-income households (77 percent in the bottom income quintile). This represents the adoption gap, driven by lack of affordability, relevant content, and/or awareness, as seen in many countries. In filling this gap, the Philippines was the pilot country for the world’s first experiment, which rigorously evaluated the impact of first-time access to low-cost community cellular networks (CCNs). The experiment provided powerful evidence on the benefits of connectivity to rural household welfare, while also revealing the limitations in the current licensing and spectrum regime in scaling such interventions. 30 The Connecting Communities through Mobile Networks (CoCoMoNets) project deployed CCNs using 2G GSM cellular technology in 7 barangays (villages) in the province of Aurora. The aim of this research was to test, through a randomized control trial (RCT), the viability of low-cost networks (one-tenth of the cost of traditional mobile towers) that allowed voice calls and text messaging in remote areas. The 7 treatment barangays where the towers were deployed) were randomly chosen from a pool of 14 rural and geographically isolated barangays based on observable similarities that complied with preset criteria. The remaining 7 control barangays served as comparison groups for estimating project impact. Blumenstock, et al. (2020) found that, first, villages that were assigned to receive a new cell phone tower were significantly more likely to access and use communications technology. This was driven mainly by an increase of 42 percentage points in households’ reported ability to place a call from their house. Prior to the project, 67 percent of households already owned at least one mobile phone but had to travel far to access a mobile network. Second, the study found increased social connectedness in treatment barangays. Lastly, and more importantly, the study found a 17-percent increase in annual income among households in treatment barangays, a 10-percent increase in household expenditure over the last week, and a 13-percent increase in reported adequate food in the last month (a proxy measure of food security). Despite the successful results, the project team encountered barriers to continued operation, which led to shutting down sites or pivoting to Wi-Fi Internet access networks. Barela, et (2023) documents the implementation challenges, including 2 years negotiating for spectrum co-use with existing telco. As the authors noted, “The Philippines does not have a dedicated spectrum policy for last-mile service delivery, and the current regulatory framework forces small operators to adapt to the model used for national telecoms and other large organizations.” In March 2020, the 2G networks were taken offline as the telco and project support ended. However, two barangays decided to adopt the management of Wi- Fi networks and continue providing connectivity to the community (Barela et al. 2023). 31 5.2 ROLE OF THE GOVERNMENT IN BRIDGING DIGITAL DIVIDE Bridging the digital divide is an urgent call  lways leverage private sector with a for the government’s role beyond policy financial capacity, execution capability, and reforms to channel public resources to meet knowledge of markets and technologies, the needs of the poor. Many countries, from across network deployment and service those with low- to high-income economies, delivery, especially for middle-/last-mile. have pursued universal service policies for •D ynamic, inclusive, and well-informed broadband internet. The range of approaches engagements enable learning and course and instruments include and oftentimes corrections. Overbuilt, under-utilized, combine (i) service obligations (geographic stranded infrastructure assets and failed or population coverage) attached to license commercial endeavors can sometimes to operate, including with spectrum; (ii) be repurposed or brought back into market-oriented allocation of capital subsidy productive use. Government can harness (e.g., reverse auction mechanism for viability capabilities and voices of large and gap funding); (iii) dedicated fund with area small market participants, broader public targeting or site selection mechanisms utilities, civil society, academia, and (universal access fund); and (iv) government- other national and local stakeholders to driven direct investments in rural connectivity, adopt modalities and (re-)target smart including through Wi-Fi access in public interventions. places. Government spectrum revenue typically funds these instruments except for service The Philippines faces unique challenges in obligations, which effectively oblige operators bridging the digital divide given the inter- to internally cross-subsidize, from more to less related policy and regulatory gaps not seen profitable customer bases. in most peer countries. Key constraints include the congressional franchise and Global and regional experiences in filling rigid spectrum assignment, a divided market broadband access gaps provide three key structure with weak competition, lack of lessons:92 government opportunities to reallocate •A government’s attempt to narrow the access resources and oblige operators to contribute gap alone, without well-functioning markets, to public purposes, and suboptimal revenue does not work. A well-sequenced approach from the growing digital sector. always starts with policy reforms to address market and regulatory failures and promote The government can complement pro- competition and investment. competition policy reforms with direct public •E ven with ongoing or advanced market investments in justifiable, efficient, and reforms, experiences of USF and other sustainable ways, through initiatives such as fiscal measures are mixed, typically facing the National Fiber Backbone, “Broadband ng challenges in design (sustainability, Masa” (Free Wi-Fi for All program), satellite transparency, scale) and implementation connectivity, and Public-Private Partnerships (procurement, maintenance, service quality, to leverage private sector expertise and upgrading). As much as feasible, capabilities. The expansion of the National 32 Fiber Backbone93 can benefit from the new Rural areas (around 30 percent of the 3.  features of RA 11966 or the PPP Code, signed population): Connectivity is largely limited into law in December 2023. The Free Wi-Fi to mobile, fixed wireless, and satellite. program can benefit from better procurement Complementary public investments and procedures and improved absorptive capacity regulatory reforms are needed to facilitate of the DICT. By simplifying and rationalizing and expand commercial sustainability guidelines for the registration of satellite frontier for affordable infrastructure providers, Executive Order No. 12794 can be deployment. leveraged to connect more GIDA and remote Far-flung areas (last 10 percent of the 4.  rural areas. The DICT may also subject ongoing population): Commercial viability gap infrastructure projects to PPP, which has is the largest among all areas and clearer guidelines for national and local PPP government interventions in missionary projects, and new procedures for unsolicited areas are necessary. projects.95 Government competency, including Approaches under the current market maturity institutional frameworks, capacities, and and constraints that include geographic, resources of DICT and NTC, are vital in socioeconomic, and other local challenges, designing and implementing the policy can be differentiated across four geographic reforms and complementary public segments as summarized in Figure 14. investments. 1. Urban centers (around 30 percent of the These require improvements, such as national population): Broadband demands updating regulatory knowhow, investing in are relatively well-served and competition new skills, restructuring or creating new should continue to improve infrastructure offices or positions as necessary, adapting access and performance. regulatory issuances to reflect the immediate 2. Peri-urban areas (around 30 percent of the reforms, and collaborating across relevant population): Fixed broadband is still nascent government agencies to implement the new and regulatory reforms need to diversify policy framework. and facilitate more market entrants and investments. 33 Figure 14: Fig 13 – Differentiated Differentiated investment interventions fit for market constraints Source: World Bank staff calculations based on PSA CPH Form 3 data aggregated at municipality level. Note: Distance in the x-axis is weighted by land area. The primary focus of the government should affordable broadband connectivity, plausible be to adopt and implement a coherent only through significant increase in volume set of market reforms, to lower barriers to and improvements in efficiency and quality entry, promote investment, and encourage of private infrastructure investments. This competition via efficient use of spectrum makes it urgent for the government to put resources and shared infrastructure in place necessary policy reforms to reduce deployment. In the PDP 2023-28, the national residual gaps that have to be addressed target is 60 percent broadband penetration by through public investment. 2028. This requires “doubling” the reach of 34 6. Conclusion and Recommendations Weak internet infrastructure in the Philippines, bills, policymakers are advised to focus on with more expensive service, and lower maintaining and strengthening the breadth network coverage than its regional peers, • Remove the legislative franchise constrain the digital potentials of citizens requirement for internet infrastructure and firms by slowing inclusive recovery and deployment. This is a requirement no growth and worsening the digital divide. An other country in the world imposes and inadequate and antiquated policy environment, is closely related to other constraints on as evidenced by the Philippines’ exceptionally internet infrastructure, such as licensing, low ranking across regional and global digital spectrum assignment, and foreign indices, have created and reinforced a market investment. structure with weak competition and limited • Ensure spectrum reassignment flexibility investments in rural connectivity infrastructure. through reforms in the OADT bill, such Unaddressed, weak internet might derail the as introducing a time-bound spectrum country’s achieving upper middle-income assignment period and stronger regulatory status in the coming years and its aspiration autonomy for the NTC. for a prosperous middle-class society by 2040. • Explore mid-/longer-term solutions to digital connectivity gaps to enable rural Key policy and regulatory gaps that call for infrastructure deployment through fiscal immediate reforms include (i) barriers to policy options for spectrum revenue, market entry and investment; (ii) unlevel suitable universal service policies, and playing field; (iii) rigid and non-transparent efficient public investment approaches. spectrum management; and (iv) ineffective infrastructure sharing policy framework. Once enacted into law, the OADT’s immediate These are closely interrelated and require reforms can be an entry point to broader, comprehensive reforms for meaningful medium- and longer-term digital connectivity changes in market participants’ incentives and agenda. The next phase of spectrum reforms, conduct. In the short-term, the passage of the such as mobile spectrum restacking, regional proposed Open Access for Data Transmission spectrum licensing, piloting spectrum (OADT) Act in the 19th Congress is a auction, and revising SUF pricing, could viable, promising opportunity to reform the ensure adequate fiscal policy options for Philippines’ decades-old policy frameworks, longer-term agenda to lower end-user price thereby adapting to technology and business for broadband connectivity and expand model changes in the broadband value chain. rural coverage while maximizing a scarce spectrum resource. The government’s To compromise on the breadth and depth of infrastructure investments (through the Internet policy reforms would be a missed National Fiber Backbone and Broadband ng opportunity to resolve the problems created Masa, among others) can complement pro- by regulatory weaknesses. In the deliberations competition reforms in a justifiable, efficient, over reforms in the remaining year of the and sustainable manner. Government 19th Congress, including the Open Access competency will need strengthening in bills, policymakers are advised to focus on designing and implementing policy reforms maintaining and strengthening the breadth and complementary initiatives. 35 The cost of inaction—loss of growth opportunity, people remaining unequipped for future jobs, and widening of the digital divide—is too high for the Philippines. Outdated policy and regulations have long stunted the growth of the country’s broadband industry and expansion of digital infrastructure. As the government prioritizes digitalization with a view toward benefiting all Filipinos, updating policy to promote competition, encourage investment, and upgrade broadband infrastructure is urgent and necessary.  36 References A4AI. 2019. “Planning spectrum for all communities.” Alliance for Affordable Internet. https://a4ai. org/research/good-practices/planning-spectrum-for-all-communities/ A4AI. 2021. “Affordability Report 2021.” Alliance for Affordable Internet. https://a4ai. org/report/2021-affordability-report/ A4AI. 2022. “Meaningful Connectivity for Rural Communities: Geographic Barriers & Policy Strategies for Digital Inclusion.” Alliance for Affordable Internet. https://webfoundation.org/ docs/2022/04/MC-Rural-Report-English.pdf Amadasun, Kennedy, Michael Short, and Tracey Crosbie. 2020. “Telecommunication infrastructure sharing a remedy for the reduction of network operator cost and environmental pollution.” 2020 IEEE International Conference on Environment and Electrical Engineering and 2020 IEEE Industrial and Commercial Power Systems Europe (EEEIC/I&CPS Europe). Madrid, Spain, 2020, pp. 1-6. https://doi.org/10.1109/EEEIC/ICPSEurope49358.2020.9160661. APT Wireless Group. 2021. “APT Report on Information of Mobile Operators’ Frequencies, Technologies and License Durations in Asia Pacific Countries.” Asia-Pacific Telecommunity. https:// www.apt.int/sites/default/files/2021/04/APT-AWG-REP-15Rev.7_-_APT_Report_on_Information_of_ Mobile_Operators_Frequencies_Technologies_and_License_Durations_in_Asia_Pacific_Countries. docx Bahia, Kalvin and Paul Castells. 2022. “The impact of spectrum assignment policies on consumer welfare.” Telecommunications Policy, 46(1):102228. https://doi.org/10.1016/j.telpol.2021.102228 Barela, Mary Claire, Maria Therese Cunanan, Cedric Angelo Festin, Kurtis Heimerl, Esther Jang, Matthew William Johonson, Philip Martinez, and Ronel Vincent Vistal. 2023. “Connecting Communities Through Mobile Networks: A Case Study of Implementing Community Cellular Networks in the Philippines.” In Introduction to Development Engineering, edited by Madon, Temina and A.J. Gadgil. Cham: Springer. https://doi.org/10.1007/978-3-030-86065-3_21 Barnes, Sebastian, Romain Bouis, Philippe Briard, Sean Dougherty, and Mehmet Eris. 2013. “The GDP Impact of Reform: A Simple Simulation Framework.” OECD Economics Department Working Papers No. 834. Paris: OECD Publishing. https://doi.org/10.1787/5kgk9qjnhkmt-en Barcenas, Lai-Lynn and Ramonette Serafica. 2018. “Regulatory Measures Affecting Service Trade and Investment: Information and Communications Technology Services.” Philippine Institute for Development Studies, Discussion Paper Series No. 2018-48. https://pidswebs.pids.gov.ph/CDN/ PUBLICATIONS/pidsdps1848_rev.pdf Bernabe, Johannes Benjamin. 2019. “Meeting the challenge of a new telco player.” Business Mirror, March 6, 2019. https://businessmirror.com.ph/2019/03/06/meeting-the-challenge-of-a-new-telco- player/ 37 BEREC. 2018. “BEREC Report on Infrastructure Sharing.” Body of European Regulators for Electronic Communications. https://www.berec.europa.eu/en/document-categories/berec/reports/berec- report-on-infrastructure-sharing Blackman, Colin and Lara Srivastava. 2011. Telecommunications Regulation Handbook – 10th anniversary edition. Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/ en/297991468154471456/Telecommunications-Regulation-Handbook-10th-anniversary-edition Blumenstock, Joshua, Niall Keleher, Arman Rezaee, and Erin Troland. 2020. “The Impact of Mobile Phones: Experimental Evidence from the Random Assignment of New Cell Towers.” Working Paper. https://www.jblumenstock.com/files/papers/jblumenstock_2020_ccn.pdf Bohlin, Anders, Harold Gruber, and Pantelis Koutroumpis. 2010. “Diffusion of New Technology Generations in Mobile Communications.” Information Economics and Policy, 22(1):51-60. https:// doi.org/10.1016/j.infoecopol.2009.11.001 Bourreau, Marc, Pinar Dogan, and Matthieu Manant. 2010. “A Critical Review of the ‘Ladder of Investment’ Approach.” Telecommunications Policy, 34(11):683-696. https://dash.harvard.edu/ handle/1/4777447 Bourreau, Marc, Steffen Hoernig, and Winston Maxwell. 2020. “Implementing Co-investment and Network Sharing.” Center on Regulation in Europe. Brussels: CERRE. https://cerre.eu/publications/ telecom-co-investment-network-sharing-study/ Bouis, Romain and Romain Duval. 2011 “Raising Potential Growth after the Crisis: A Quantitative Assessment of the Potential Gains from Various Structural Reforms in the OECD Area and Beyond.” OECD Economics Department Working Papers, No. 835. Paris: OECD. https://doi. org/10.1787/5kgk9qj18s8n-en Brewer, Jonathan, Yoonee Jeong, and Arndt Husar. 2022. “Last Mile Connectivity: Addressing the Affordability Frontier.” ADB Sustainable Development Working Paper Series No. 83. Manila: Asian Development Bank. https://www.adb.org/sites/default/files/publication/847626/sdwp-083-last- mile-connectivity-affordability-frontier.pdf Broadband Commission. 2021. “21st Century Financing Models for Bridging Broadband Connectivity Gaps.” Research Report. Broadband Commission for Sustainable Development, ITU, UNESCO. https:// www.broadbandcommission.org/publication/21st-century-financing-models/ Cave, Martin. 2006. “Encouraging infrastructure competition via the ladder of investment.” Telecommunications Policy, 30(3-4):223-237. https://www.sciencedirect.com/science/article/abs/ pii/S0308596106000164 38 Contreras, Alexandra. 2023. “Spectrum prices and regulatory framework.” Cullen International. https://ctu.int/wp-content/uploads/2022/12/WS_3.-Spectrum-prices-Jamaica-Feb-2023.pdf Deloitte. 2015. “Unlocking Broadband for All: Infrastructure Sharing Policies and Strategies in Emerging Markets. https://www.apc.org/sites/default/files/Unlocking%20broadband%20for%20 all%20Full%20report_1.pdf Digital Regulation Platform. 2022. “The Infrastructure Sharing Initiative” Digital Regulation Platform.” International Telecommunication Union and the World Bank. https://digitalregulation. org/the-infrastructure-sharing-imperative/ DTI. 2021. “Philippine National AI Strategy Roadmap.” Department of Trade and Industry. Egert, Balazs and Peter Gal. 2016. “The Quantification of Structural Reforms in OECD Countries: A New Framework.” Economics Department Working Paper No. 1354. Paris: OECD. https://www. oecd.org/competition/reform/The-quantification-of-structural-reforms-in-OECD-countries-a-new- framework.pdf Espos, Edna. 2003. “Institutions, Regulations, and Performance: The Caw of Philippine Telecommunications (Regulation of the Philippine Telecommunication Industry).” Philippine Journal of Public Administration, 47(1-4)38-61. https://www.pssc.org.ph/wp-content/pssc- archives/Philippine%20Journal%20of%20Public%20Administration/2003/07_Institutions,%20 Regulation%20and%20Performance.pdf FCC. n.d. “Auction Maps - Geographic Licensing Schemes.” Federal Communications Commission. https://www.fcc.gov/economics-analytics/auctions-division/auctions/auction- maps#Geographic%20Licensing%20Schemes Forge, Simon and Khuong Vu. 2020. “Forming a 5G strategy for developing countries: A note for policy makers.” Telecommunications Policy, 44(7):101975. https://doi.org/10.1016/j. telpol.2020.101975 Gallitto, Lucas and Luciana Camargos. 2021. “Brazil multi-band auction: one of the largest in mobile history.” GSM Association. https://www.gsma.com/spectrum/brazil-multi-band-auction-one- of-the-largest-in-mobile-history/ Garcia, Jose Marino. 2016. “The economics and policy implications of infrastructure sharing and mutualization in Africa.” World Development Report 2016 background papers. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/480181468185331506/World- development-report-2016-digital-dividends-the-economics-and-policy-implications-of- infrastructure-sharing-and-mutualization-in-Africa 39 Garity, John and Arndt Husar. 2021. “Digital Connectivity and Low Earth Orbit Satellite Constellations: Opportunities for Asia and the Pacific.” ADB Sustainable Development Working Paper Series No. 76. Manila: Asian Development Bank. https://www.adb.org/sites/default/files/ publication/696521/sdwp-076-digital-connectivity-low-earth-orbit-satellite.pdf Geerdts, Christopher. 2016. “Developing Smart Public Wi-Fi in South Africa.” Research ICT Africa. Ottawa: International Development Research Centre. https://www.researchictafrica.net/ publications/Other_publications/2016_Public_Wi-Fi_Policy_Paper_-_Developing_Smart_Public_Wi- Fi_in_South_Africa.pdf Gruber, Harald and Pantelis Koutroumpis. 2011. “Procompetitive infrastructure sector regulation and diffusion of innovation: The case of broadband networks.” 22nd European Regional ITS Conference, Budapest 2011: Innovative ICT Applications – Emerging Regulatory, Economic and Policy Issues 52161, International Telecommunications Society (ITS). Gruber, Herald and Frank Verboven. 2001. “The evolution of markets under entry and standards regulation—the case of global mobile telecommunications.” International Journal of Industrial Organization, 19(7):1189-1212. https://doi.org/10.1016/S0167-7187(01)00069-8 GSMA. 2012. “Mobile Infrastructure Sharing.” London: GSM Association. https://www.gsma.com/ publicpolicy/wp-content/uploads/2012/09/Mobile-Infrastructure-sharing.pdf GSMA Intelligence. 2015. “Spectrum For New Entrants, Lessons Learned.” London: GSMA Intelligence. https://data.gsmaintelligence.com/api-web/v2/research-file-download?id=18809282 &file=spectrum-for-new-entrants,-lessons-learned-1482139870208.pdf GSMA Intelligence. 2022. “The 5G FWA Opportunity: A Scenario for Southeast Asia.” GSMA Intelligence. https://data.gsmaintelligence.com/api-web/v2/research-file- download?id=72941596&file=220622-FWA-Opportunity-Southeast-Asia.pdf Houngbonon, Marc Ivaldi, Emil Palikot, and Davide Strusani. 2023. “The Impact of Shared Telecom Infrastructure on Digitial Connectivity and Inclusion.” Working Paper No. 1472. Toulouse School of Economics. ISED. 2020. “Service areas for competitive licensing.” Innovation, Science, and Economic Development Canada. https://ised-isde.canada.ca/site/spectrum-management- telecommunications/en/spectrum-allocation/service-areas-competitive-licensing#tierMap ISED. 2023. “Spectrum Outlook 2023 to 2027.” Government of Canada. https://ised-isde.canada. ca/site/spectrum-management-telecommunications/en/spectrum-allocation/spectrum- outlook-2023-2027 40 ITU. 2008. “Mobile Sharing.” 8th Global Symposium for Regulators Discussion Paper. Geneva: International Telecommunication Union. https://www.itu.int/ITU-D/treg/Events/Seminars/GSR/ GSR08/discussion_papers/Camila_session4.pdf ITU. 2010. “Best Practice Guidelines for Enabling Open Access.” 10th Global Symposium for Regulators. Geneva: International Telecommunication Union. https://www.itu.int/ITU-D/treg/ Events/Seminars/GSR/GSR10/consultation/guidelines/GSR10_guidelines_V3-en.pdf ITU. 2015. “IMT Vision – framework and overall objectives of the future development of IMT for 2020 and beyond.” Geneva: International Telecommunication Union. https://www.itu.int/dms_ pubrec/itu-r/rec/m/R-REC-M.2083-0-201509-I!!PDF-E.pdf ITU. 2018. “Setting the Scene for 5G: Opportunities & Challenges.” Geneva: International Telecommunication Union. https://www.itu.int/en/ITU-D/Documents/ITU_5G_REPORT-2018.pdf ITU. 2020. “The Last-Mile Internet Connectivity Solutions Guide: Sustainable Connectivity Options for Unconnected Sites.” Geneva: International Telecommunication Union. https://www.itu.int/en/ ITU-D/Technology/Pages/LMC/LMC-Home.aspx ITU. 2021. The Impact of Policies, Regulation, and Institutions on ICT Sector Performance. Geneva: International Telecommunication Union. https://handle.itu.int/11.1002/pub/817bc896-en ITU. 2023. “Best Practice Guidelines: Regulatory and Economic Incentives for an Inclusive and Sustainable Digital Future.” Global Symposium for Regulators 2023. Geneva: International Telecommunication Union. https://www.itu.int/itu-d/meetings/gsr-23/wp-content/uploads/ sites/20/2023/06/GSR-23_Best-Practice-Guidelines-E.pdf Johansson, Asa, Yvan Guillemette, Fabrice Murtin, David Turner, Giuseppe Nicoletti, Christine de la Maisonneuve, Philip Bagnoli, Guillaume Bousquet, and Francesca Spinelli. 2013. “Long-term Growth Scenarios.” OECD Economics Department Working Papers No. 1000. Paris: OECD. https://doi. org/10.1787/5k4ddxpr2fmr-en Jung, Juan and Raul Katz. 2022. “Spectrum flexibility and mobile telecommunications development.” Utilities Policy, 75(April):101351. https://doi.org/10.1016/j.jup.2022.101351 Kochhar, S.P. 2023. “Importance of spectrum and its efficient management in telecom.” The Economic Times. https://telecom.economictimes.indiatimes.com/news/industry/importance-of- spectrum-and-its-efficient-management-in-telecom/101443056 Kochhar, S.P. 2023. “Importance of spectrum and its efficient management in telecom.” The Economic Times. https://telecom.economictimes.indiatimes.com/news/industry/importance-of- spectrum-and-its-efficient-management-in-telecom/101443056 41 Koutroumpis, Pantelis, Pau Castells, and Kalvin Bahia. 2023. “To share or not to share? The impact of mobile network sharing for consumers and operators.” Information Economics and Policy, 65(December): 101061. https://doi.org/10.1016/j.infoecopol.2023.101061 Levi-Faur, David. 1999. “The Governance of International Telecommunications Competition: Cross- International Study of International Policy Regimes.” Competition and Change, 4(1):93-120. https:// doi.org/10.1177/102452949900400104 Licsi, Ayie. 2022. “Filipinos spend the most time online and on social media at almost 11 hours per day—study.” The Philippine Star, February 2, 2022. https://philstarlife.com/geeky/844848-sortlist- social-media-filipinos-study Lusinyan, Lusine. 2018. “Assessing the Impact of Structural Reforms through a Supply-side Framework: The Case of Argentina.” IMF Working Paper 18/183, International Monetary Fund, Washington, DC. https://www.imf.org/en/Publications/WP/Issues/2018/08/03/Assessing-the- Impact-of-Structural-Reforms-Through-a-Supply-side-Framework-The-Case-of-45937 Mamushiane, Lusani, Albert A. Lysko, and Sabelo Dlamini. 2018. “SDN-enabled Infrastructure Sharing in Emerging Markets: CapEx/OpEx Savings Overview and Quantification.” 2018 IST-Africa Week Conference (IST-Africa), Gaborone, Botswana, 2018. Matheson, Thornton and Patrick Petit. 2017. “Taxing Telecommunications in Developing Countries.” Working Paper No. 2017/247. Washington, DC: International Monetary Fund. Mejes, Dan and Cristina Batalla. 2022. “Common Towers in the Philippines: An Update.” Better Internet PH. https://medium.com/@betterinternetph/common-towers-in-the-philippines-an- update-68f3128e4348 Ministry of Science and ICT Korea. 2022. “As for 5G 28GHz Spectrum, license duration reduced on SKT, and assignment cancelled on LG Uplus and KT.” Ministry of Science and ICT Korea. https://www.msit.go.kr/eng/bbs/view.do?sCode=eng&mId=4&mPid 2&pageIndex=&bbsSeqNo=42&nttSeqNo=753&searchOpt=ALL&searchTxt= Mirandilla-Santos, Mary Grace. 2016. “Philippines Broadband: A Policy Brief.” Arangkada Philippines. Mirandilla-Santos, Mary Grace, Jonathan Brewer, and Jaime Faustino. 2018. “From Analog to Digital: Philippine Policy and Emerging Internet Technologies.” Manila: The Asia Foundation. Motta, Massimo and Emanuele Tarantino. 2021. “The effect of horizontal mergers, when firms compete in prices and investments.” International Journal of Industrial Organization, 78(September):102774. 42 Muente-Kunigami, Arturo and Juan Navas-Sabater. 2010. “Options to Increase Access to Telecommunications Services in Rural and Low-Income Areas.” World Bank Working Paper No. 178. Washington, DC: World Bank. http://hdl.handle.net/10986/5927 NEDA. 2023. “Philippine Development Plan 2023-2028.” National Economic and Development Authority. https://pdp.neda.gov.ph/philippine-development-plan-2023-2028/ OECD. 2013. “Broadband Networks and Open Access.” OECD Digital Economy Papers No. 218. OECD Publishing, Paris. https://doi.org/10.1787/20716826 OECD. 2023. Extending Broadband Connectivity in Southeast Asia. OECD Publishing. Paris:OECD. https://doi.org/10.1787/b8920f6d-en Ortiz, Ma. Kristina, Ramonette Serafica, and Jose Carlos Alexis Bairan. 2017. “Rebooting Philippine Telecommunications Through Structural Reform.” Philippine Institute for Development Studies Discussion Paper No. 2017-19. https://pidswebs.pids.gov.ph/CDN/PUBLICATIONS/pidsdps1719.pdf Oughton, Edward, Nicolo Comini, Vivien Foster, Jim W. Hall. 2021. “Background Paper: Policy Choices Can Help Keep 4G and 5G Universal Broadband Affordable.” Policy Research Working Paper No. 9563. Washington, DC: World Bank. Oughton, Edward, David Amaglobeli, and Mariano Moszoro. 2023. “Estimating Digital Infrastructure Investment Needs to Achieve Universal Broadband.” IMF Working Paper No. 2023/27. Washington, DC: International Monetary Fund. PSA. 2023. “Digital Economy Contributes 9.4 Percent to the Gross Domestic Product in 2022. Philippine Digital Economy Satellite Accounts.” Press Release, April 25, 2023, accessed December 12, 2023. Philippine Statistics Authority. https://psa.gov.ph/statistics/digital-economy Riaz, Aamir. 2016. “Spectrum Access and Auctions. Workshop on Spectrum Auctions: Economic Aspects.” Presented in ITU Workshop on Spectrum Managment: Economic Aspects, November 21- 23, 2016, Tehran, Iran. https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/Documents/ Events/2016/Nov-SM-Economics/Presentations/Day%202-Session5_Aamir%20(Auction%20 Design).pdf Roland Berger. 2022. “Towering Above: Building Tomorrow’s Digital Infrastructure in Asia.” White Paper. Munich: Roland Berger. https://www.rolandberger.com/en/Insights/Publications/Towering- Above-Building-Tomorrow-s-Digital-Infrastructure-in-Asia.html Romero, Alexis. 2023. “Lawmakers vow to pass Marcos’ priority bills.” The Philippine Star, September 21, 2023. https://www.philstar.com/headlines/2023/09/21/2297846/lawmakers-vow-pass-marcos- priority-bills 43 Serafica, Ramonette and Queen Cel Oren. 2022. “The Philippines Digital Sector and Internet Connectivity: An Overview of the Value Chain and Barriers to Competition.” Philippine Institute for Development Studies Discussion Paper No. 2022-47. https://www.pids.gov.ph/publication/ discussion-papers/the-philippine-digital-sector-and-internet-connectivity-an-overview-of-the- value-chain-and-barriers-to-competition TRAI. 2019. “Enabling 5G in India.” New Delhi: Telecom Regulatory Authority of India. https://trai. gov.in/sites/default/files/White_Paper_22022019.pdf TRAI. 2023. “Consultation Paper on Telecommunication Infrastructure Sharing, Spectrum Sharing, and Spectrum Leasing. New Delhi: Telecom Regulatory Authority of India.” https://www.trai.gov.in/ sites/default/files/CP_INF_13012023.pdf UN. 2022. “UN e-Government Survey 2022.” United Nations. https://publicadministration.un.org/ egovkb/en-us/Reports/UN-E-Government-Survey-2022 UN ESCAP. 2017. “The Impact of Universal Service Funds on Fixed-Broadband Deployment and Internet Adoption in Asia and the Pacific.” Asia-Pacific Information Superhighway Working Paper Series. https://hdl.handle.net/20.500.12870/4126 USAID. 2018. “Next Generation Universal Service toward Ubiquitous Broadband Ecosystems in ASEAN.” White Paper. United States Agency for International Development. https://pdf.usaid.gov/ pdf_docs/PA00TD3J.pdf Uy, Krystal Lyn. 2022. “Adapting Telecommunications Regulation to Competition: A Selection of Key Issues for Reform in the Philippines.” Asian Journal of Comparative Law, 17(2):344 – 368. https:// doi.org/10.1017/asjcl.2022.21 Wang, Liang and Qiming Sun. 2022. “Market Competition, Infrastructure Sharing, and Network Investment in China’s Mobile Telecommunications Industry.” Sustainability, 14(6):3348. https://doi. org/10.3390/su14063348 World Bank. 2016. “World Development Report 2016: Digital Dividend.” Washington, DC: World Bank. https://www.worldbank.org/en/publication/wdr2016 World Bank. 2018. “Innovative Business Models for Expanding Fiber-Optic Networks and Closing the Access Gaps.” Washington, DC: World Bank. https://www.worldbank.org/en/topic/ digitaldevelopment/publication/innovative-business-models-for-closing-the-internet-access-gaps- lessons-for-policy-makers World Bank. 2020. “Philippines Digital Economy Report 2020: A Better Normal Under COVID-19: Digitalizing the Philippines Economy Now.” Washington, DC: World Bank. 44 World Bank. 2022. “Philippines Economic Update June 2022 Edition: Strengthening the Digital Economy to Boost Domestic Recovery.” Washington, DC: World Bank. World Bank. 2023. “East Asia and the Pacific Economic Update October 2023: Services for Development.” Washington, DC: World Bank. World Bank. Forthcoming. The Path to 5G in Developing World: Planning Ahead for a Smooth Transition. Washington, DC: World Bank. 45 Annexes ANNEX 1: WORLD BANK CEM3.0 GROWTH AMBITION AND STRUCTURAL REFORMS TOOL The World Bank Country Economic It assumes a standard Cobb-Douglas Memorandum (CEM) is a comprehensive production function Y=F(K, L, θ): analysis of a country’s economic developments, prospects, and policy agenda, and identifies policy reforms for key economic sectors. The Growth Ambition and Structural Reforms Tool, The impacts on the capital-output ratio and developed within the CEM 3.0 framework, employment rate are estimated in panel is designed to identify key structural reform regressions where random effects models policies associated with GDP growth by with robust standard errors are estimated to looking at their respective impacts on capital capture both between- and within-country accumulation, labor utilization, and total factor information. Technical efficiency estimates productivity or efficiency. come from stochastic frontier analysis following Lusinyan (2018): The analysis behind the Growth Ambition and Structural Reforms Tool builds on the supply- side approach introduced by Bouis and Duval (2011), Barnes et al. (2013), Johannson et al. The structural reform areas covered in the (2013), and Egert and Gal (2016). It assumes model are listed in the table below, including a standard Cobb-Douglas production function the indicators used and data sources. Y=F(K, L, θ): Structural reform area Indicator Data Source Business regulations/ Regulation Fraser/WGI regulatory quality Labor market regulations/female Labor market Fraser/WGI labor participation Top marginal income and payroll Taxation Fraser tax rate Financial development Credit to private sector WDI Trade openness Freedom to trade internationally Fraser Governance Government effectiveness WGI Telecommunication infrastructure Telecommunication infrastructure UN index Note: Fraser – Fraser Institute; WGI – World Bank Worldwide Governance Indicators; WDI – World Bank World Development Indicators; UN – United Nations. 46 For each structural reform area, the tool The UN’s Telecommunication Infrastructure calculates the change in real GDP growth if Index is a composite of four indicators based a certain percentage of a structural gap with mainly on ITU data (UN 2022): (i) internet an aspirational peer is closed over a specified users per 100 inhabitants; (ii) number of period. The analysis should be viewed to show mobile subscribers per 100 inhabitants; (iii) potential growth benefits from an ambitious number of wireless broadband subscriptions reform agenda that is well-implemented. per 100 inhabitants; and (iv) number of In the case of the Philippines (World Bank fixed broadband subscriptions per 100 PH CEM, forthcoming), the tool was used to inhabitants. For each of these indicators, estimate the annual GDP growth that would standardized Z-scores were calculated. The be added if it were to close the structural gaps telecommunication infrastructure index, with Malaysia over 25 years. It shows that ranging from 0 to 1, is the normalized fully catching up with Malaysia on business arithmetic mean of the four standardized regulations and telecoms infrastructure will indicators. contribute over 85 percent or 2.75 percentage points of the 3.21 percent change in GDP growth. 47 ANNEX 2: AFFORDABILITY DRIVER INDEX (ADI ) POLICY SCORE DATA AND METHODOLOGY The Affordability Driver Index (ADI) is Policy and regulation for competition (4) a composite measure that summarizes •A1: To what extent are ICT licensing an assessment of the drivers of internet frameworks flexible, simple, and affordability in various countries, and is technology and service neutral? (4) published by Alliance for Affordable Internet •A2: To what extent does the government (A4AI), a public-private consortium hosted ICT regulator perform its functions at World Wide Web Foundation. In the according to published and transparent Philippines, the Congressional Policy and rules, with the ICT regulatory decisions Budget Research Department (CPBRD), influenced by public consultations? (4) House of Representatives has monitored and •A3: To what extent does the regulator and/ published on the Philippines’ position in the or the competition commission enforce ranking. the country’s ICT licensing requirements and regulations? (4) Calculation of the ADI score involves 14 •A4: To what extent is ICT regulatory indicators as primary data from A4AI’s own decision-making informed and influenced research, complemented by secondary data by adequate evidence? (4) from other providers such as ITU, GSMA Intelligence, World Bank, and Packet Clearing Broadband policy (5.75) House. Primary data is collected through a •A5: To what extent does the national multi-country expert survey, with questions broadband Internet plan (or in some cases scored on a scale of 0–10 by country experts the national ICT policy and plans) set clear, based on predetermined criteria, and verified time-bound targets and interventions for by peer reviewers. Survey issues include reducing broadband cost and increasing policy, regulation, and other aspects around penetration? (6.5) broadband and affordable access to internet. •A8: To what extent has the government From 2015 to 2021, the number of countries defined specific, limited and well- the index fully covered was 49 (2015), 57 justified guidelines for public (2017), 59 (2018-19), and 70 (2020-21). No infrastructure funding or subsidies in data was collected for 2016. For the 2021 telecommunications? (5) report, A4AI (2021) conducted policy surveys on 72 countries between April and June, 2020. Universal access (3.33) A11: To what extent have Universal • The 14 questions, based on and aligned Access/Service Funds (USAF) prioritized with the A4AI Policy and Regulatory Good infrastructure investments that will reduce Practices, attempt to assess the extent to costs and increase access for underserved which countries have achieved a policy and communities and market segments? (1) regulatory environment that reflects the A12: To what extent does the government • best practice outcomes. Thirteen of the 14 ICT regulator perform its functions indicators constitute 5 areas as listed below according to published and transparent (with the Philippines 2021 score in brackets): rules, with the ICT regulatory decisions influenced by public consultations? (1) 48 A13: Are there specific policies to promote • A10: To what extent are the government’s • free or low-cost public internet access, such plans for implementing more spectrum as budget allocations for internet access availability for broadband (both licensed in public libraries, schools and community and unlicensed) transparent, and are done centers, or provisions for spectrum use by through a competitive process via public community wi-fi options? (8) auctions? (3) Infrastructure sharing (3.5) The 14th indicator on gender equity does not •A6: To what extent are national-level belong to above categories: policies or rules in place to facilitate •A14: To what extent do the country’s efficient access to public rights of way and broadband policies include strategies tower zoning permission? (3) and programs to improve access and use •A7: To what extent does the government among women and girls? (5) facilitate resource sharing across telecommunications operators? (4) A4AI publishes the Expert Survey Codebook and Interview Guidelines, aligning scoring Spectrum policy (3) criteria and guidelines among the responding A9: To what extent has government • domestic experts and international experts established an implementation plan with as peer reviewers for data verification. As a time-bound target for making sufficient an illustrative example, guidance related to spectrum available for broadband within spectrum availability (A9) is as follows: a reasonable period of time to meet the growing demand for high-speed data services? (3) Scoring Criteria 10 5 0 There is a detailed plan There are general principles around increasing the No plans to improve for the timely release of amount of spectrum available for broadband, but the amount of sufficient spectrum to meet those plans may lack ambition (not enough spectrum spectrum available projected increases in is being targeted to meet likely increases in demand) for broadband demand. The government has or specificity around time-bound benchmarks. Plans exist, or such plans implemented and executed may lack details around the specific interventions are so vague and significant and important the government plans to take to achieve such lacking in details portions of its plans to make targets. The government has made some progress as to render additional spectrum available towards implementing portions of its plans to make them useless and for broadband according to its additional spectrum available for broadband, but impossible being timeline for implementation. important gaps remain. The government may have monitored and/or Such changes may include prioritized less important elements of the plan implemented. both meaningful regulatory, over more important but potentially costly and/or policy and/or legal reforms. politically sensitive elements. 49 For more information, see: •CPBRD Facts in Figures, Apr 2022 (No. 29): https://cpbrd.congress.gov.ph/images/ PDF%20Attachments/Facts%20in%20 Figures/FF2022-29_2021_Boardband_ Affordability_Drivers_Index_of_the_Phil.pdf •A4AI Affordability Report 2021: https://a4ai. org/report/2021-affordability-report/ •A4AI ADI Data: https://adi.a4ai. org/affordability-report/data/?_ year=2021&indicator=INDEX •A4AI Policy and Regulatory Good Practices: https://a4ai.org/good-practices/ •A4AI Expert Survey Codebook and Interview Guidelines: https://docs.google.com/ document/d/1OG12mI2E3SAjqTk6-cF_ dFG8qfH7fO-_smquf4Nv6Mg/ (Accessed December 12, 2023) 50 Annex Table: Reforms for Affordable Internet by Country, 2017-21 2017 ADI Scores by Policy Area 2021 ADI Scores by Policy Area Broad Infrastr Spectr Broad Infrastr Spectr 17-21 Comp band Universal ucture um Comp band Universal ucture um delta etition policy access sharing policy Avg. etition policy access sharing policy Avg. Thailand 3.5 4.0 5.3 5.0 5.5 4.7 4.5 7.5 8.0 5.5 8.5 6.8 2.1 Burkina Faso 5.6 3.0 1.7 1.5 3.5 3.1 5.8 6.0 2.3 5.0 5.5 4.9 1.9 Senegal 4.8 3.8 4.3 5.5 3.5 4.4 5.5 8.0 5.0 5.5 5.0 5.8 1.4 Nepal 5.3 4.8 5.0 4.8 4.0 4.8 5.4 7.5 6.5 5.5 6.0 6.2 1.4 Cambodia 3.8 6.0 2.2 5.0 3.3 4.0 4.5 6.3 6.3 4.5 5.5 5.4 1.4 Tanzania 6.1 4.3 4.2 4.0 3.5 4.4 5.6 5.0 5.8 6.0 6.5 5.8 1.4 Malawi 5.3 2.5 3.2 2.5 4.0 3.5 5.8 5.0 3.8 4.5 4.0 4.6 1.1 Jordan 6.3 5.0 3.0 4.8 4.5 4.7 7.0 6.5 2.7 6.3 6.5 5.8 1.1 Argentina 5.8 7.0 6.3 5.5 5.8 6.1 6.4 7.5 7.3 6.5 7.8 7.1 1.0 Jamaica 5.6 4.5 6.8 3.5 5.5 5.2 6.5 6.0 7.3 4.0 7.0 6.2 1.0 Ethiopia 1.0 3.5 1.3 1.0 1.5 1.7 3.0 4.0 1.0 2.5 2.5 2.6 0.9 Botswana 6.5 6.8 5.2 5.0 5.0 5.7 7.0 8.0 7.0 5.0 6.0 6.6 0.9 Malaysia 5.6 7.8 7.3 7.0 6.3 6.8 7.0 9.0 8.0 5.5 9.0 7.7 0.9 Indonesia 6.0 6.0 6.3 3.3 5.5 5.4 4.3 5.8 8.5 6.0 7.0 6.3 0.9 Tunisia 5.3 4.0 2.7 5.3 4.0 4.2 5.8 6.5 3.8 5.3 4.3 5.1 0.9 Zimbabwe 3.6 3.0 3.3 2.5 3.5 3.2 4.1 4.0 4.5 4.3 3.0 4.0 0.8 Uganda 6.3 6.0 6.0 5.5 3.8 5.5 5.8 6.3 6.3 7.0 6.0 6.3 0.8 Myanmar 3.8 7.0 3.3 4.5 4.5 4.6 4.5 7.3 4.0 5.5 5.5 5.4 0.7 India 5.8 5.5 5.7 5.0 6.5 5.7 6.8 6.0 6.3 7.0 6.0 6.4 0.7 Dominican Republic 5.5 5.5 6.7 3.5 6.0 5.4 5.6 8.0 6.3 5.5 5.0 6.1 0.7 Mali 5.9 6.0 3.7 6.5 3.8 5.2 6.0 6.5 5.0 6.0 5.5 5.8 0.6 Pakistan 7.0 4.8 5.3 5.0 6.5 5.7 7.0 7.0 7.0 4.5 6.0 6.3 0.6 Rwanda 4.9 6.5 7.2 5.5 3.8 5.6 5.5 6.8 7.7 6.0 4.5 6.1 0.5 Viet Nam 4.3 5.5 7.2 5.5 3.0 5.1 4.5 6.0 5.0 6.5 6.0 5.6 0.5 Kazakhstan 4.0 5.0 2.8 3.5 3.5 3.8 4.4 6.0 2.7 4.5 3.8 4.3 0.5 China 4.0 6.5 3.0 5.5 2.5 4.3 4.5 6.0 1.3 7.0 5.0 4.8 0.5 Benin 5.5 6.5 5.3 6.0 5.5 5.8 6.0 7.0 6.0 6.5 5.5 6.2 0.4 Morocco 6.1 6.0 7.2 6.0 5.8 6.2 6.5 6.8 7.0 6.0 6.8 6.6 0.4 Costa Rica 6.5 7.5 8.0 5.5 7.0 6.9 7.0 8.5 7.3 5.5 8.0 7.3 0.4 Ghana 5.5 4.5 5.3 5.3 5.5 5.2 4.9 5.0 5.8 6.0 6.0 5.5 0.3 Bangladesh 4.1 5.0 2.7 6.0 3.5 4.3 4.5 5.5 4.3 4.5 4.0 4.6 0.3 Namibia 7.4 2.5 1.5 5.0 3.5 4.0 6.1 4.0 0.8 5.5 4.5 4.2 0.2 Gambia 6.0 5.3 2.7 4.5 4.8 4.6 6.9 6.5 2.3 4.0 4.5 4.8 0.2 Kenya 7.3 5.0 3.3 4.3 4.0 4.8 5.8 5.0 6.3 3.5 4.3 5.0 0.2 Mauritius 6.8 6.5 6.2 4.8 5.0 5.8 6.8 7.5 6.3 4.5 5.0 6.0 0.2 Turkey 6.6 5.8 6.0 6.8 5.8 6.2 6.3 6.5 5.7 6.5 6.8 6.3 0.2 Peru 8.0 8.5 6.2 7.8 7.0 7.5 7.5 8.0 6.0 8.3 8.0 7.6 0.1 Sierra Leone 4.5 3.5 2.2 3.0 1.5 2.9 4.0 3.0 1.0 4.5 2.5 3.0 0.1 Zambia 5.9 4.0 4.5 5.0 4.5 4.8 5.8 4.0 4.2 5.8 4.5 4.8 0.1 Colombia 6.9 8.5 8.7 6.5 8.0 7.7 7.4 7.5 8.0 8.3 7.5 7.7 0.0 Cameroon 4.0 4.5 4.7 6.0 2.5 4.3 3.8 4.5 4.7 5.5 3.3 4.3 0.0 Nigeria 7.0 6.2 6.3 5.8 6.5 6.3 6.1 7.5 6.0 6.0 6.0 6.3 0.0 Mozambique 5.5 4.5 5.0 5.5 4.5 5.0 3.8 5.3 5.0 5.5 5.3 5.0 0.0 Côte d'Ivoire 6.8 5.5 6.3 5.5 4.5 5.7 6.8 4.0 6.3 5.5 5.5 5.6 -0.1 Philippines 5.8 5.0 2.0 4.0 4.5 4.3 4.0 5.8 3.3 3.5 3.0 3.9 -0.3 Egypt 4.3 6.0 4.7 3.0 5.0 4.6 3.8 5.0 5.8 2.8 3.8 4.2 -0.4 Bolivia 5.0 4.0 6.3 5.0 6.0 5.3 3.9 5.5 4.7 6.0 4.0 4.8 -0.5 Honduras 4.8 5.0 4.3 6.0 6.0 5.2 3.8 3.5 4.3 6.5 5.5 4.7 -0.5 Sudan 3.4 4.5 3.7 3.0 2.0 3.3 3.0 2.5 3.3 2.5 2.5 2.8 -0.5 Ecuador 5.8 7.0 7.3 7.0 5.0 6.4 5.4 6.8 4.0 6.5 6.0 5.7 -0.7 Haiti 2.1 0.3 1.8 2.0 3.8 2.0 2.9 1.0 0.3 1.3 1.0 1.3 -0.7 Mexico 7.5 7.5 6.7 7.0 6.5 7.0 8.0 6.0 1.7 8.0 8.0 6.3 -0.7 Nicaragua 3.0 3.0 4.0 3.5 4.0 3.5 2.4 2.5 3.3 3.5 2.3 2.8 -0.7 South Africa 5.6 5.3 5.7 5.0 3.8 5.1 5.6 4.8 2.7 5.0 3.5 4.3 -0.8 Brazil 6.1 6.0 4.7 7.5 7.3 6.3 7.3 3.0 3.3 5.5 6.5 5.1 -1.2 Sri Lanka 3.3 5.0 6.0 7.5 6.5 5.7 2.6 4.0 5.3 4.5 5.0 4.3 -1.4 Venezuela 4.3 4.5 6.7 4.0 5.0 4.9 3.0 3.0 3.5 4.0 2.5 3.2 -1.7 51 ANNEX 3: EVIDENCE OF REGULATORY REFORMS IMPACT ON INTERNET PERFORMANCE AND GROWTH Competition / Open Access countries during the period 2008-19 to test the model, finds that when the market is Gruber and Koutroumpis (2011) evaluate how open to foreign players, capital investment competition-oriented infrastructure regulation is stimulated and increases by 14 percent, can promote the spread of innovation using network coverage by over 13 percent, price broadband service adoption data from 167 reduction by close to 8 percent, and mobile countries. The authors note that broadband penetration by close to 3 percent after competition can manifest in two distinct two years. The report notes that foreign forms: (i) as intra-platform competition, investment facilitates the growth and where various service providers compete over development of the ICT sector, enabling the same network infrastructure, and such access to capital for network development competition is facilitated by open access at and modernization, and allowing for the different network levels; (ii) as inter-platform transfer of technology and know-how. These competition, where competition arises through lead to increased productivity, innovation, the use of diverse technological platforms and competitiveness. The report also finds for facility-based connectivity. The authors that moving from service and network find that inter-platform competition does not specific to convergent licensing is associated significantly speed up broadband diffusion. with an increase in mobile investment by However, intra-platform competition, through 10 percent, network coverage by over 9 retail competition and full unbundling, percent, price reduction by over 5 percent, leads to faster diffusion. The effect of retail and mobile penetration by approximately 2 competition on diffusion is about twice as percent after 2 years. The report estimated strong and is most pronounced in the first a structural model for the mobile sector, 3-4 years after implementation, likely due to focusing on a set of specific policy reforms, intensified price competition. The authors note namely the adoption of a national broadband that full unbundling is opposed by incumbent plan, the introduction of convergent licenses, firms as this measure is considered intrusive the possibility for mobile operators to with respect to network architecture, requires conduct infrastructure and spectrum sharing additional investments, and may reduce the agreements, the requirement of mobile incentives to invest in new infrastructure. The number portability, the market openness authors employ a technology diffusion model to foreign operators and the creation of a in the vein of Gruber and Verboven (2001), and competition authority. All policies were found Bohlin et al. (2010), modeled in the way of to yield increased mobile capital investment disease spread in biology. levels, which in turn translated to coverage gains, price reductions, higher adoption ITU (2021) employed a structural causal model levels, and, ultimately, GDP per capita growth. consisting of four equations, one each on Allowing spectrum sharing agreements, coverage, demand, price, and investment to promoting a national broadband plan, and study the impact of policies, regulations, and opening the industry to foreign operators institutions on ICT performance. The study, stood out as the policies that exhibit the which used a database composed of 145 largest payback. 52 Spectrum Efficiency by 1–2.5 Mbps (equivalent to an increase of up to 15 percent). ITU (2021) found that voluntary spectrum sharing agreements are associated with an Jung and Katz (2022) study the effects on increase after 2 years in mobile investment mobile market outcomes from three specific by 18 percent, network coverage by over 17 attributes linked to a flexible spectrum percent, price reduction by close to 10 percent, policy: the presence of a secondary market, and mobile penetration by over 3 percent. The a technological neutrality approach, and the authors posit that spectrum sharing allows the possibility of conducting sharing agreements maximization of opportunities for operators among operators. The authors find that to make investments profitable, creating when the three policies are jointly adopted, incentives for network deployment. Network- mobile capital investment is 35.9 percent sharing agreements can optimize the use of larger than when that is not the case. In infrastructure, generally reducing costs, thus addition, promoting these policies can being beneficial for both service providers and increase network coverage by 9.8 percent consumers. after 2 years, bring down mobile broadband prices by 5.8 percent, and increase mobile Bahia and Castells (2022) investigate the broadband unique subscribers’ penetration by impact of spectrum assignment policies 0.9 percent. The authors built an unbalanced on consumer welfare, focusing on network panel consisting of 145 countries for 2008– coverage and quality. Using data from 19 and employed a system-GMM (generalized 229 mobile operators across 64 countries method of moments) methodology to from 2010 to 2017, the authors employ an estimate their model. econometric approach with instrumental variables to address potential endogeneity Infrastructure Sharing in spectrum prices. They find that spectrum assignment policies, including the amount, Network infrastructure is a strategic timing, and cost of spectrum, significantly asset vital in competitive downstream influence network coverage and quality. markets, impacting service costs and The results show that earlier and increased quality. Owners of this infrastructure spectrum allocation enhances network can potentially discriminate against coverage, while higher spectrum costs competitors by increasing access costs negatively affect network quality and coverage, or limiting service quality of competitors particularly in developing countries. They seeking network access under bilateral find that, ceteris paribus, if an operator was agreements. Access regulation is meant to assigned 4G spectrum at least 2 years earlier mitigate such negative impacts. From an than another operator, 4G network population economic theory perspective, the impacts coverage would average between 11 and of such arrangements are ambiguous. 16 percentage points higher. They find that Consumers might gain advantages such as an additional 1 MHz of spectrum allocated better coverage, enhanced network quality, drove a 0.1–0.2 percentage point increase in and reduced prices if cost savings from 4G coverage and an additional 20 MHz of 4G infrastructure sharing are transferred to spectrum increased average download speeds them. However, a decrease in 53 infrastructure-based competition could lead driven by CAPEX reductions, higher returns to less variety in services, reduced innovation, on investment—providing operators with and may also diminish the motivation to both the ability and incentive to invest—and invest. The debate on innovation in business increased competition. In some cases, smaller encompasses two main theories: the operators benefited in terms of cost savings “Schumpeterian Effect” where large firms that allowed them to reduce prices and in concentrated markets are more driven to improve and expand their networks. CAPEX innovate, and the “Escape-Competition Effect” savings, profit margin improvements, and where increased competition spurs innovation. price reductions were particularly associated These theories converge in an inverted with passive sharing. Passive sharing was U-shaped relationship, suggesting that also linked to increases in 3G coverage, moderate competition enhances innovation to as the majority of passive agreements in outperform rivals, while extreme competition Europe were established during the phase diminishes innovation by reducing monopoly of 3G deployments. By contrast, the positive profits. The actual scale of trade-offs will be effects of network sharing on 4G coverage dependent on the specificities of a particular and download speeds were mostly driven by country and regional context. active sharing, which became more prevalent when operators rolled out 4G networks after Empirical evidence on the benefits of shared 2010. mobile telecom infrastructure remains limited at this stage, especially in developing Houngbonon et al. (2023) investigates the countries. Among the growing crop of recent impact of shared telecom infrastructure studies, Koutroumpis et al. (2023) assess on digital connectivity and inclusion the impact of mobile network sharing in using a dataset on mobile tower sharing Europe during the 2000-19 period, looking transactions between 2008 and 2020, i.e., at 140 mobile operators in 29 countries. acquisitions of towers by independent They implement a difference-in-difference companies from mobile network operators (DID) model at the operator-level, to to be rented back to all operators. Estimates test the framework set out in Motta and based on difference-in-differences with Tarantino (2021) with Bertrand competition, different timing of treatment find that these differentiated goods, and multiple firms. They transactions resulted in a significant drop in find that network sharing generated significant the price of mobile connectivity as well as an benefits for operators and consumers, increase in availability and uptake of mobile including lower prices and improved network Internet, especially by rural households coverage and quality, and average increases and women. They find that tower sharing in population coverage of between 2-9 deals have a statistically significant impact percentage points versus the baseline. This on 3G and 4G coverage in the year of the was driven by cost reductions, higher returns transaction as well as in the following years. on investment, and increased competition. This effect amounts to an 8.5 percentage Operators that entered into network sharing points increase in 3G coverage 2 years of the agreements were able to reduce prices transaction, which corresponds to 13 percent, (proxied by ARPUs) and increase network and an increase of 7.8 percentage points in coverage and quality. This was the case of 4G (13.1 percent). 54 They find an impact on price of mobile cellular The BEREC (2018) report on infrastructure connections (20 percent decrease) and an sharing (2018) presents a provisional analysis increase in access to internet by women-led of infrastructure sharing arrangements, which households and by rural households in the are currently in place in many European year of the transaction (approximately 18 countries. The report indicates that as per the percent). figures provided by some National Regulatory Authorities (NRAs): Cost Saving channel CAPEX OPEX Passive sharing 16 - 35% 16 - 35% Active sharing (excl. spectrum) 33 - 35% 25 - 33% Active sharing (incl. spectrum) 33 - 45% 30 - 33% The report notes that not all network cost results show that both market competition savings qualify as an efficiency that is likely and infrastructure sharing individually have to be passed on to and benefit consumers. a significant positive impact on the industry’s Certain fixed cost savings (that is, cost savings network investment, while their interaction that do not relate to incremental costs) are has a significant negative impact on network unlikely to be passed on to consumers in the investment. One possible explanation is form of lower prices or increased incentives to that when the higher level of infrastructure invest in a foreseeable timeframe. In addition, sharing and more fierce market competition the extent to which savings in incremental appear at the same time, the competitive costs are passed on to consumers depends on strategy of rational operators changes the market structure and, more generally, the and begins to shift from Facility-based level of competition in the market. competition to Service-based competition. Although in-depth infrastructure sharing can Wang and Sun (2022) use fixed effects and further save network investment and improve dynamic panel data models to empirically network quality, it may also damage market examine the relationship among the competition and service differentiation, as market competition, infrastructure sharing, well as reduce operators’ willingness to and network investment in China’s mobile invest in the advanced telecommunications telecommunications industry. Compared with network and develop new products and operators in developed countries, preferring services. service-based competition, China’s operators in the factor/investment-driven growth phase Amadasun et al. (2020) estimated that prefer facility-based competition. According to sharing of sites and antennas could reduce a Deloitte report, the number of base stations CAPEX costs by 20-30 percent. Also sharing per 10,000 people in China is 14.1, which is of the radio network can reduce CAPEX by significantly higher than the 8.7 in Europe and 25-45 percent. Sharing of all the assets could 4.7 in the United States. Their estimation reduce CAPEX by another 10 percent. 55 could reduce CAPEX by another 10 percent. Arakpogun et al. (2020) investigated barriers to infrastructure sharing in seven SSA countries based on interviews with different stakeholders and found that: (i) incumbents might be reluctant to share networks as this would erode their competitive advantage, and (ii) many markets that have been dominated by one or two MNOs for years, lack institutional incentives to engage in infrastructure sharing. Mamushiane et al. (2018) quantified the economic impact of infrastructure sharing using Software Defined Networking (SDN) and estimated that in the case of South Africa, full implementation of sharing can reduce the time to recover CAPEX investments and reach profitability from 5.4 months to less than 1.3 months in rural areas, suggesting profitable operation in even less economically attractive zones. 56 Endnotes In this Policy Note, the term access is used for both infrastructure coverage and affordability, resulting in service 1 subscription, in line with the “household with internet access” as the national target set through the Philippines Development Plan 2023-28. According to the World Bank Findex 2021-22 database, ownership of a mobile money account (among population age 2 15+) was 22 percent in the Philippines, 16 percent in Viet Nam, and 9 percent in Indonesia. The ratio of citizens who received government payment (among population age 15+) was 37 percent in the Philippines, 29 percent in Indonesia, and 19 percent in Viet Nam. See BusinessWorld, August 17, 2023, “WB: Digital competitiveness needs liberal regulation of cross-border data” https:// 3 www.bworldonline.com/economy/2023/08/17/540317/wb-digital-competitiveness-needs-liberal-regulation-of-cross- border-data/ Among the 37 extra-regional submarine cable systems (connecting to other regions of the world) in Southeast Asia, 4 the Philippines has landing points for 17, the second largest number in the region following Singapore with 28 landing points. Many of the 12 ongoing submarine cable construction projects that are expected to be in service in 2023-24 involve the Philippines (OECD 2023). For example, Zoom recommends 1.5 Mbps (up/down) for each additional participant in a group call using HD (High- 5 definition, 720p) video, meaning 15 Mbps dedicated bandwidth for a conference call with 10 participants. This can be accommodated either by fixed broadband via fiber or coaxial cable, and mobile if 5G, but typically not 4G (LTE). Absolute prices were estimated based on the ITU cost of a basket of fixed and mobile broadband in 2022 as a percent 6 of GNI (see Figure 1) and World Development Indicators GNI per capita data at 2022 current prices. The cost of mobile broadband in the Philippines in absolute USD is higher than in Indonesia, Malaysia, and Vietnam (economies with higher GNI per capita than the Philippines) and Lao, Cambodia, and Myanmar (with lower GNI per capita). Indicators published by multiple data sources show fixed and mobile broadband access per country, and this Policy 7 Note makes choices. For fixed broadband, as shown in Figure 1 and Figure 3 (first graph), ‘household penetration’ available from TeleGeography (Country Statistics, Fixed Broadband) is used because of its cross-country comparability and time-series availability, although it has limitations. TeleGeography aggregates reporting from major operators and nation-wide ISPs, omitting smaller players such as local cable operators and satellite service providers, both of which play important roles in rural connectivity in the Philippines. However, no alternative cross-country data sets address this data gap. ITU publishes an indicator ‘households with Internet access at home’, which this Policy Note does not use, due to limited time series (for the Philippines, 2010 and 2019 only) and a comparability issue (ITU relies on reporting from responsible country authorities, namely, DICT in case of the Philippines. Eighteen percent of households have Internet access according to ITU using the National ICT Household Survey 2019. Though the 18 percent is indicated to include ‘wired broadband, wireless broadband, satellite broadband, and mobile broadband’ in DICT’s analysis, as in https://dict. gov.ph/ictstatistics/wp-content/uploads/2022/06/National-NICTHS-Results_Updated-June-2022.pdf slide 8 and 11, it is likely that potential inconsistencies—definitional, methodological, interpretive, and/or through reporting to ITU— have caused the Philippines’ reporting on fixed broadband only and differently from other countries reporting on all connection types. For example, for the same 2019 indicator, internet access was 71 percent in Viet Nam, 75 percent in Thailand, 79 percent in Indonesia, and 90 percent in Malaysia. For mobile broadband, as shown in Figure 1 and Figure 3 (second graph), ‘active mobile-broadband subscription per 100 inhabitants’ available from ITU is used, following ITU’s definition of mobile broadband (1.5M bps or above), because alternative data sources such as TeleGeography and GSMA report on subscribers either per 2G, 3G, 4G, and 5G services, or mobile internet, without isolating mobile broadband (including text-only 2G services). The Philippines’ National AI Strategy Roadmap (DTI 2021) states that “The need for reliable and robust network and 8 internet connection is paramount to the transformation of the Philippines into a knowledge-based economy” (p. 10). The PSA has been collecting detailed and representative technology data for both manufacturing and services firms 9 that can be matched to their firm-level productivity surveys to gain novel insights on technology diffusion and the implications for productivity, wages, and jobs. The average Philippine services firm has around a third more data and software assets per worker than their manufacturing counterpart; the average technology services firm has seven times more data and software assets per worker than the average manufacturer. Moreover, the use of data is uneven across services firms in the Philippines, with foreign firms five times more data-intensive than those domestically owned and firms with fiber more than twice as data-intensive as those without broadband (World Bank 2023). 57 According to the BPO Expansion survey conducted by the Contact Center Association of the Philippines (CCAP) in 2023, 10 28 percent of industry players planned to expand their business in Cavite City until 2028; followed by 23 percent in San Fernando City and 21 percent in Batangas, Puerto Princesa, and Metro Rizal. According to TeleGeography Cloud Region Data Center Locations database, Amazon Web Service (AWS) started to 11 operate in Singapore in 2010, Meta in Malaysia in 2012, Google Cloud in Indonesia in 2020. In 2022, All of AWS, Google Cloud, Microsoft (Azure), and Meta operated in Singapore; Google, Microsoft, and Meta operated in Malaysia; AWS, Google, and Microsoft operated in Indonesia; and none of these major hyper-scalers were operating in the Philippines. In 2023, AWS announced its plan to operate in the Philippines. Another indicator under SGD 4.4.1 is “Writing a computer program using a specialized programming language,” a proxy 12 to advanced ICT skills, which less than 1 percent in the Philippines are able to do in 2019. Malaysia showed significant increase from 8.2 percent to 16 percent between 2019 and 2021. Data for all countries are 2019 values (the only data available for the Philippines), except for Cambodia and Indonesia, which are 2017 values. No data is available for Lao and Myanmar. The analysis is based on Lusinyan (2018) methodology and data from the Fraser Institute, the UN E-Government 15 Knowledgebase, World Development Indicators (WDI), and World Governance Indicators (WGI). PLDT 2021 (large CAPEX, and temporal net income declined to PHP 10.7 billion in 2022); Globe, Dito, Converge, and 16 Now 2022. InfiniVan represents segments revenue of holding company IPS excluding Japan business. Main backbone networks in the Philippines include PLDT’s Domestic Fiber Optic Network (DFON), Globe Telecom’s two 17 Fiber Optic Backbone Networks (FOBNs), Converge ICT’s Domestic Submarine Cable Network (CDSCN), and the planned NGCP/TransCo Backbone Network based on which the government, through DICT, plans to build a National Fiber Backbone. Third generation (3G) mobile technology, introduced around 2000, brought about mobile broadband. Fourth generation 18 (4G), since 2010, significantly improved data speed and enriched mobile broadband applications beyond the static web to social networks and data-intensive, productivity-enhancing applications. Fifth generation (5G), introduced from 2019-20 and once mature, will substantially enhance data rate, support high connection loads suitable to massive machine-type communications, and provide ultra-reliable and low-latency communications suitable for mission- critical applications. With a conducive regulatory and competitive environment and business models for cost-efficient deployment models, 5G technology is expected to be a viable alternative to fiber-to-the-home to provide connectivity in rural areas, in particular. Satellite communications coverage is composed of geostationary orbit (GEO) satellites (also known as geostationary 19 Earth orbit), medium Earth orbit (MEO), and low Earth orbit (LEO) constellations. Due to their lower altitude, LEO satellites provide low latency, and high bandwidth transmission and are suited for broadband internet applications. While LEO satellites have been in deployment for decades, they again require intensive investment as new large constellations are in the early stages of deployment. These new LEO constellations, such as those being deployed by Starlink (SpaceX), already provide nation-wide coverage in the Philippines. They may prove to be transformational to the connectivity landscape because of their suitability for areas not served by fiber optic cable networks, global coverage, scalability, and potential cost advantages (Garity and Husar 2021). ISPs and Cable Operators as included in disclosed lists through NTC’s response to Freedom of Information Requests 20 from 2019 to 2023, supplemented by a list of Cable Operators offering internet service provided by the Philippine Cable and Telecommunications Association (PCTA), for which financial statements for the recent 5 years were registered for SEC. Figures in Table 1 are based on the World Bank’s preliminary analysis of the 30 sample ISPs headquartered outside NCR or Region IV-A, whose financials were obtained from the SEC. The financial statements of the subset of these sample ISPs that disclose the cost for wholesale bandwidth indicate that the cost of bandwidth as a share in revenue increased from 8 percent in 2017 to 33 percent in 2022 in aggregate, while the cost of cable TV programming decreased from 51 percent of revenue in 2017 to 29 percent of revenue in 2022. The analysis is currently expanded to more granular regional comparisons. 58 In aggregate, 15 percent of households had access to fixed wired internet; 4 percent had fixed wireless; 2.4 percent 21 had satellite broadband network; and 42.1 percent had mobile broadband network. See: PSA “More than 50 million have Access to the Internet (2020 CPH)”, https://psa.gov.ph/content/more-50-million-have-access-internet-2020- census-population-and-housing. Analysis of microdata to account for households with more than one type of internet connection shows that incremental fixed wireless access (households without fixed wired and with fixed wireless) was 3.4 percent, and incremental satellite access (households without fixed wired or fixed wireless and with satellite internet) was 1.8 percent; and ratio of households with mobile broadband only (without any types of fixed broadband) was 35.8 percent, leaving 44 percent of households without any type of their own access to broadband internet. EBITDA (Earnings before Interest, Taxes, Depreciation, and Amortization), divided by revenue, is a commonly used metric 22 to evaluate firms in capital-intensive sectors, where net income after accounting and financing decisions can disguise baseline profitability. Mobile tower density as supplementary analysis indicates cumulative capital formation in mobile telecom. ITU data for 2022. This indicator is distinct from access by or coverage of population or households (one subscription 23 typically covers multiple inhabitants in a household; subscriptions by businesses are also included). Malaysia data for operators’ share and top two operators’ profitability reflects the merger of Celcom and Digi in 2022, 24 therefore treating Maxis Mobile (#3 until mid-2022) as #2. In Thailand, True and DTAC (#2 and #3 in 2022) completed their merger in early 2023 (Thailand is now another country with a dominant duopoly); operators profitability data from the post-merger True Corp (new #1) and AIS (new #2). For Viet Nam, VNPT only (no Viettel disclosure). The analysis of profitability may not represent comparable mobile network operations, as many of the top 2 operators in LIC/MIC are vertically integrated, and their published financial reporting consolidates backbone and fixed access (net of transfer price for wholesale network). According to the national source on Gross Capital Formation (GCF) in Durable Equipment – Telecommunications & 25 sound equipment, at current prices, from PHP 116.5 billion (2018), 106.6 (2019), 87.4 (2020), 96.1 (2021) to 99.9 (2022). Note that industry classification for GCF accounting may not include ISPs or cable operators providing broadband internet services. In 2021, among the 70 countries with ADI scores, the Philippines ranked 33rd on broadband policy; 48th on universal 26 access; 51st on competition; 58th on spectrum policy; and 59th on infrastructure sharing. Aside from these 5 areas, the Philippines ranked 8th on gender equity in broadband access and use. See Annex 2 for more data and methodology. Countries with ADI scores improving from 2017 to 2021 and higher than the Philippines in 2021 included: archipelago 27 (Indonesia, Jamaica, Dominican Republic, Mauritius), landlocked (Nepal, Botswana, Uganda, Zambia, Bolivia, Kazakhstan), and fragile and conflict-affected (Myanmar, Mali, Cameroon, Burkina Faso). Broadband Commission (2021) focused on regulatory interventions largely coinciding with the scope of the discussion 28 in this note. A4AI (2022) and ITU (2023) elaborated more on public investments and fiscal measures to address the residual broadband access gaps – namely, A4AI (2022) policy and regulatory good practices include (i) nurture a healthy and competitive market; (ii) promote policy and regulatory leadership, supported with strong capacity; (iii) streamline broadband’s national and international value chains; (iv) manage spectrum effectively and efficiently; (v) strengthen the taxation system and establish affordability targets; (vi) promote public infrastructure investment mechanisms focused on market failures; (vii) promote efforts to collect, disaggregate and systematize data; and (viii) promote broadband policies for underserved rural areas. ITU (2023) best practice guidelines, through its infrastructure pillar (incentives towards achieving meaningful connectivity), include market access, universal access and service, universal access funding, balancing fiscal policies, innovative regulatory last mile connectivity solutions, R&D, and spectrum reform. 59 ITU Regulatory Tracker is another index assessing countries’ telecommunication policy and regulatory framework. 29 It covers a larger number of countries (193 countries in 2022), and includes 50 indicators under its four pillars of (i) regulatory authority, (ii) regulatory mandate, (iii) regulatory regime, and (iv) competition framework. However, the index does not emphasize affordable broadband infrastructure deployment (e.g., no indicator on the content of universal access policy). The Philippines does not perform as well as its ASEAN middle-income peers in the ITU Regulatory tracker (2022 score 69.2, compared to Thailand 85.5, Malaysia 85, Viet Nam 70.8 and Indonesia 62). The UN Telecommunication Infrastructure Index, another source of international benchmarking, is a composite index calculated from a set of sector outcome indicators (e.g., internet users, mobile subscribers, fixed line subscribers), rather than from a qualitative assessment of policy and regulatory framework. In the UN Telecommunication Infrastructure Index, the Philippines does not perform as well as its ASEAN middle-income peers (2022 score 0.5683, compared to Malaysia 0.7945, Thailand 0.7338, Vietnam 0.6973, and Indonesia 0.6397). As of 2023, the Regulatory Watch Initiative (RWI), produced by the World Bank for data-driven benchmarking around five regulatory themes (licensing and authorization, fair markets, international access, spectrum management, and regulatory governance), covers 66 countries in Africa, the Middle East, Central and South Asia, and Latin America but not East Asia and the Pacific as of 2023. In the initial wave of liberalization, in most countries, the incumbent private telecommunications firm replaced the 30 state telco and operated as a monopoly franchise. In the second wave, governments authorized the entry of new market entrants, which typically involved 1) modifying the licensing framework to allow the entry of new players, and 2) introducing complementary rules and regulations to allow the new operators to participate in the marketplace. Read more about the evolution of regulatory reforms in Blackman and Srivastava (2011, pp. 9-12). For a discussion of the step-by-step approval process for the legislative franchise and CPCN, see Barcenas and Serafica 31 (2018). According to the 2006 Rules of Practice and Procedures Before the NTC, “decisions, judgments, … the authorizations 32 granted to telecommunications/broadcast/CATV operators… shall be decided by the Commission en banc.” https://ntc. gov.ph/wp-content/uploads/2020/07/11-1-THE-2006-RULES-OF-PRACTICE-AND-PROCEDURES-BEFORE-NTC.pdf, pp. 10-11. EO 436 s. 1997 allows cable TV operators to offer value-added services using their excess capacity. 33 Communication with Joel Dabao of KCAT, a cable ISP company, and member of the Philippine Cable and 34 Telecommunications Association (PCTA), November 30, 2023. It can be 40-60 percent cheaper for a cable operator to lay its own infrastructure instead of going through a telco. Even 35 if a telco provides the middle-mile infrastructure for an ISP, it may unilaterally decide to charge a high rate for the middle-mile link. Outside the legal framework, some ISPs opt to lease an idle telco franchise to build the middle-mile infrastructure. However, this kind of arrangement adds to the ISP’s costs and to industry inefficiencies. Smaller and/or non-telco providers deploying middle-/last-mile broadband infrastructure to rural areas in Indonesia, 36 India, Georgia, Serbia, Kenya, South Africa, among others, include electric utilities, cable operators, rural ISPs, and community networks. For experiences and lessons, see World Bank (2018). For example, interconnection is mandatory between cable landing stations and backhaul operators (NTC MC 06-10- 37 2008). See Article V, Section 14 of NTC MC 14-07-2000 on the Implementing Rules and Regulations for the Interconnection of 38 Authorized Public Telecommunications Entities https://ntc.gov.ph/wp-content/uploads/2015/10/LawsRulesRegulations/ MemoCirculars/MC2000/MC-14-07-2000.pdf. See, for example, the third telco, DITO Telecommunity, complaining that interconnection issues have slowed down 39 subscriber growth. https://www.cnnphilippines.com/business/2023/8/7/dito-subscriber-base-interconnection-issue. html In response, Globe and PLDT ask NTC to penalize DITO for allegedly violating interconnect rules and contractual obligations. https://news.abs-cbn.com/business/08/09/22/globe-dito-trade-barbs-over-interconnection; https://www. bworldonline.com/corporate/2022/10/19/481573/pldt-says-dito-open-to-settle-p430-m-contractual-obligation/; https://www.pna.gov.ph/articles/1180922 60 Asymmetric regulation establishes different procedures and standards for different actors (Levi-Faur 1999; Uy 2022). 40 In the “ladder of investment” theory (Cave 2006), for example, the new entrant is successively given different levels of access (“rungs of the investment ladder) to the incumbent’s network, while inducing it to climb up the ladder of investment, such that the new entrant will gradually ramp up its own infrastructure investment. See Bourreau et al. (2010). Another example of asymmetric regulation is when UK’s OFCOM found it necessary to require an SMP operator, BT Openreach, to share its infrastructure with new entrants, while implementing wider infrastructure access regulations that applied to all infrastructure owners, to provide sufficient competitive constraint on BT, the dominant provider of telecoms infrastructure (Digital Regulation Platform 2022). Uy (2022) argues that asymmetric regulation and fair, reasonable, and non-discriminatory pricing regulation, together 41 with the appropriate balance in an access regime, are necessary to level the playing field, curtail the dominant firm’s significant market power, and encourage competition in the telecommunications sector. Examples of countries that implement some form of infrastructure sharing include Australia and France (duct sharing), 42 Japan (ducts, manholes, and poles), and the United States (poles, including any attachment to a pole, duct, conduit, or right-of-way (Mirandilla-Santos 2016). DICT DC 008 s. 2020 or the policy guidelines on the co-location and sharing of passive telecommunications towers and 43 infrastructure (PTTI). As of September 2022, the DICT has accredited 23 independent tower companies (Mejes and Batalla 2022). 44 DICT presentation at the “Fostering Digitalization through the National Competition Policy,” a PCC forum. November 6, 45 2023. For example, in 2018, the NEA issued a Memorandum Circular (MC), which increased the pole rental charge two to four 46 times that of the previous year, without consulting or informing cable operators. When a request to attach (RTA) takes too long for the lessor to approve and the provider risks losing subscribers. 47 Republic Act 11494, or the “Bayanihan to Recover as One Act” Law, was signed on September 11, 2020. 48 The EO takes elements from JMC No. 1, s. 2020 or the Streamlined Guidelines for the Issuance of Permits, Licenses, 49 and Certificates for the Construction of Shared Passive Telecommunications Tower Infrastructures (PTTIs). The JMC was expanded and revised, resulting in JMC No. 1, s. 2021. See https://arta.gov.ph/wp-content/uploads/2021/07/Revised- Telco-JMC-.pdf. EO 32 limits and identifies the national and local permits and clearances that can be required for the construction, 50 operation, and maintenance of telecommunications and Internet infrastructure. An EO is a directive to all government agencies under the Executive branch, based on the President’s authority over 51 all executive departments, bureaus, and offices as the Chief Executive. It may be issued by the President to effectively discharge the enforcement and administration of laws. However, an EO does not limit the powers of local government units, for example, to introduce requirements which they deem necessary, given LGU-specific contexts. Constitutional commissions and other autonomous government agencies are also outside the purview of an EO. The Digital Regulation Platform, a collaboration between the International Telecommunications Union (ITU) and the 52 World Bank (Digital Regulation Platform 2022). For example, Orange and Vodafone have an active mobile network sharing agreement in Spain, while managing their 53 own traffic independently and remaining competitors at the wholesale and retail level. This is expected to result in improved quality of service due to a more extensive network, reduced churn, and the ability to invest in content and app services that can drive competitive differentiation. https://www.vodafone.com/news/technology-news/vodafone- announces-expanded-network-sharing-agreement-with-orange-in-spain; https://www.fitchratings.com/research/ corporate-finance/vodafone-network-infrastructure-sharing-smart-move-ahead-of-5g-29-04-2019 Evidence indicates that as the number of towers held by independent tower companies increases (a measure of 54 infrastructure sharing), concentration in the market decreases, benefitting competition levels (Deloitte 2015). 61 CERRE asserts that the regulatory framework in the EU starts with the premise that infrastructure-based competition 56 should be fostered wherever possible. Aside from competition, access obligations must also be imposed for rural coverage, city and environmental planning (e.g., coordinating civil works or limiting the number of masts in cities), or spectrum authorizations. The infrastructure sharing guidelines by the Communications Regulatory Authorities of Southern Africa (CRASA) are 57 generally considered a good starting point. Some of the key principles point to permitting all types of sharing, as long as competition is not adversely affected, obligating all participants to negotiate the sharing of infrastructure mandated for sharing, and requiring operators designated as having SMP to publish a reference access offer (RAO) approved by the national regulatory authority. The United Kingdom also regulates the SMP operator differently because requiring access to its infrastructure by other players offers the most competitive constraint on the dominant player. See Digital Regulation Reform (2022). Information regarding cell site location, for example, can be made available either through privately organized 58 databases or portals managed by the regulator. Most regulators gather information on infrastructure sharing agreements in specific circumstances only, such as in case of disputes. In Austria, the regulator grants a new entrant national roaming access to incumbents’ mobile networks for a limited 59 term. Last amended in 1950 by Republic Act 584, https://elibrary.judiciary.gov.ph/thebookshelf/showdocs/2/14129. Act 3846 60 mandated the NTC to (i) classify radio stations (equipment) and prescribe the nature of service to be rendered by each class and by each station; (ii) assign frequencies and license radio stations (equipment); and (iii) make rules and regulations to prevent and eliminate interference between stations. See Mirandilla-Santos et al. (2018). The Radio Control Law required a permit from the Secretary of Commerce and Communication to operate a radio 61 station. EO No. 546 s. 1979 created the NTC and delegated it the power and function of spectrum management. Open tender is the most common (and in some jurisdictions, the only) method for procurement in many Latin American 62 countries, and is also used in the Philippines for some projects. Open tender can be a ‘one-stage tender process,’ where the request for proposal (RFP) is published at the same time as the contract, and launching implies the invitation to propose, with the tender open for bidding to any potential bidder. The proposal requirements also include the qualification requirements. It may be structured as one single document (less desirable) or as two separate documents (the RFP and the contract). The two-stage open tender is a variant of the one-stage tender. The only difference is the timing of the issuance of documents, separating the RFQ and the RFP. Therefore, there is an initial stage in which potential bidders are invited to pre-qualify (under an open basis) before the issuance of the RFP and contract (which will be directed to those bidders that have prequalified), but there is no short listing. Issuance of the RFP implies invitation to propose, and usually there is only one round of bidding with no negotiations. See https://ppp-certification. com/ppp-certification-guide/2-main-types-ppp-tender-processes# A spectrum auction is a method where a license to use specific bands is awarded to an entity offering the greatest 63 monetary sum for their use. The most compelling reason for an auction is for the state to generate substantial revenue and for the spectrum to be assigned to the more efficient operator. However, this must be balanced with the objective of extending network coverage and without adding complexity to the method that might increase its risk of failure. For a more detailed discussion of spectrum auctions, see Riaz (2016). According to RA 7925 and related NTC rules, mobile operators are supposed to deploy 400,000 local exchange carriers 64 (LEC) or landline telephones as a service obligation to help fulfill the government’s universal service goals. However, by the early 2000s, mobile service had overtaken landline subscription, and the deployment of landlines was eventually abandoned. The law has not been amended. In November 2015, PLDT and Globe appealed that San Miguel Corporation‘s subsidiaries had under-utilized the 700- 65 MHz frequency band and that NTC should recall and ”equitably distribute it in a transparent manner among existing operators and new entrants.” However, the NTC deemed there was no good reason to recall and reassign the spectrum band, as SMC had obtained permits to purchase equipment to build a network. 62 For example, Malaysia licenses for 1-15 years; Singapore for 15-20; Viet Nam for 15; South Korea for 5-10; and Japan 66 for 5. APT Report on Information of Mobile Operators’ Frequencies, Technologies and License Durations in Asia Pacific Countries. No. APT/AWG/REP-15(Rev.7) Edition: April 2021 https://www.apt.int/sites/default/files/2021/04/APT-AWG- REP-15Rev.7_-_APT_Report_on_Information_of_Mobile_Operators_Frequencies_Technologies_and_License_Durations_in_ Asia_Pacific_Countries.docx In countries like Singapore (2015), Austria (2013), and Mexico (2010), regulators have been observed to set aside 67 spectrum only for new entrants when new frequencies are made available (GSMA Intelligence 2015). The only exception was in 2018 when the NTC identified and reserved 2 x 70 to 90 MHz of paired spectrum and 140 MHz of spectrum in the 700 MHz, and 2.1, 2.6 and 3.3/3.4 GHz bands for assignment to a third player, through a special selection process, upon the directive of then President Rodrigo Duterte. While the frequency assignment stays with the entity that was acquired and no re-assignment of spectrum occurs, the 68 acquiring entity, by default, gets to use and hold the frequency assets. This way, the quasi-judicial process, which would have required a stakeholder consultation/hearing, is avoided. An internal arrangement is also made among the parties on payment of the spectrum user fee. However, the approval of the spectrum co-use is done behind closed doors and no review of its impact on competition or public interest is published or conducted, if at all. In May 2016, PLDT and Globe jointly acquired SMC’s telco assets for USD1.13 billion and applied for co-use of the 700 MHz band with SMC, which the NTC approved in a matter of days. https://www.gmanetwork.com/news/money/companies/568217/globe- smart-get-ntc-nod-to-use-frequencies-in-700-mhz-spectrum/story/. The PCC announced a move to open a review of the acquisition, but was prevented by a temporary restraining order (TRO) issued by the Court of Appeals, upon Globe/ PLDT’s petition against such review to the court. https://business.inquirer.net/224820/antitrust-body-gets-gag-order- tro-on-review-of-p70-b-smc-telco-sale-stays In the Philippines, fragmentation in spectrum assignment is noticeable in 900 MHz band (5 assignments only for Smart 69 and Globe), 1800 MHz band (11 assignments only for Smart and Globe), 2300 MHz band (5 assignments only for Smart and Globe, and 3 guard bands, which are unutilized frequency that act dividers to avoid interferences), and 2600 MHz band (7 assignments for Smart, Globe, and DITO, and 3 guard bands). According to Windsor Place Consulting in Asia and the Pacific, Australia, Indonesia, and Malaysia recently undertook a spectrum restacking/refarming exercise to de-fragment spectrum assignment and provide more contiguous bands per operators, resulting in significant network performance improvements. For example, peak download speed improved by 33 percent, downlink throughput and data volume increased by 10-15 percent, and uplink spectral efficiency (affecting upload speed) improved by 50 percent, in certain cities in Australia. In the Philippines, the regulator’s efforts for similar initiative supported by international partners have not materialized to date. For example, at one point, 90 percent of the 700 MHz was assigned to only one entity. 70 With other entities holding smaller shares of the bands, 3.48 percent of the mobile access spectrum remains available 71 for a potential fourth player, unless more spectrum is made available via the second ”digital dividend” in the 600 MHz band (i.e., post the analog to digital television switchover) and the 3.5 and 6 GHz bands post refarming. The digital dividend refers to the portions of radio spectrum in the upper part of 470-862 MHz to be refarmed from broadcasting to mobile, made available by the transition of terrestrial television broadcasting from analogue to digital transmission. In 2018, DITO was assigned 4G and 5G frequencies and practically no 2G frequencies, which, at the time, many people 72 were still using for basic voice calls and text messaging. DITO initially offered 4G mobile services and began rolling out 5G only in 2022Q1. This lack of sufficient frequencies in certain bands discouraged the entry of new players (Bernabe 2019). This entails making spectrum recall and reassignment part of the process once a license expires. The current practice 73 of PLDT and Globe is to renew their franchise and license before they expire, so they are able to keep their current spectrum assignments. In other countries, renewing a license is an opportunity for the regulator to scrutinize an operator’s performance and the market impact of assignments. According to a USAID estimate, payment for Spectrum User Fee accounted for 3 percent of PLDT’s and Globe’s revenue 74 and 20 percent of DITO’s revenue, in 2022. 63 The Philippines’ Freedom of Information policy is contained in EO No. 2, s. 2016 https://www.officialgazette.gov. 75 ph/2016/07/23/executive-order-no-02-s-2016/. On various occasions, the NTC has held that spectrum assignment information is confidential. In response to one FOI request, the NTC said it ”maintains its commitment to protect the information of licensed/authorized entities transacting in this Commission.” The NTC first published mobile access spectrum in 2017, under a new President and newly appointed DICT secretary. Its most recent published spectrum assignment and use, including the bands under ‘co-use’ arrangement, was in response to an FOI request in 2022. See: https://www.foi.gov.ph/requests/aglzfmVmb2ktcGhyHQsSB0NvbnRlbnQiEE5UQy0zOTkwOTQ4MDQ1MTAM See, for example, the Australian Communications and Media Authority’s five-year spectrum outlook. 76 In Thailand, overheated auction prices made operators’ financials unsustainable, motivating a major merger between 77 DTAC and Truemove. Going forward, there will unlikely be spectrum auctions in Thailand as there are only two MNOs unless the NBTC artificially restricts the mobile spectrum it offers by auction. Estimate by USAID BEACON, assuming 10-15 years amortization period for spectrum acquisition costs. However, 78 complete information of all acquisitions is not available, and published financial statements of the two major telcos are limited in terms of the accounting of the spectrum assets (either as goodwill or other intangible assets), their amortization schedule, and spectrum-related expenses including those for bands under ‘co-use’ agreements with the acquired entities with franchise and spectrum. For example, PLDT stopped disclosing “taxes and licenses – Wireless Related” in 2017; Globe changed spectrum EUL from 10 years until 2021 to “10-59 years” in 2022. In many countries, taxes on the telecoms sector have proliferated, such as through elevated customs duties on 79 capital equipment and handsets, diverse regulatory fees, elevated CIT and VAT rates, and telephone call excises. Such taxes have become a source of considerable uncertainty and compliance costs to the industry. Heavy taxation of the telecom sector has arisen from several factors, including perceived high profit levels, the large size and formality of the sector, the difficulties of spectrum valuation and MNE (multinational enterprises) income taxation, and “claw-back” of initially over-generous concessions. Where rents from limited competition cannot be addressed through regulatory measures, they may justify a higher level of taxation. However, many of the fiscal instruments applied to telecoms do not target rents well and moreover distort product and output markets, thus impeding efficiency, affordability, and growth. Auctions, which have the capacity to capture rents without distorting investment and prices, are a favored option. However, because they are based on expected rents, they may either leave the operator with a windfall or strain its financial position. Alternative options for targeting rents are a rent tax or CIT surcharge. Like auctions, these instruments capture a portion of telecom rents with minimal impact on communications prices. They are both more equitable and flexible than auctions in that they are based on actual rather than expected profits. However, they share a common weakness in being subject to earnings manipulation and transfer pricing. For more discussions on direct and indirect taxes to the telecom sector in developing countries (see Matheson and Petit 2017). Annual fees and acquisition price (e.g., auction) for spectrum where applicable. PH collects annual fees only. For 80 Indonesia, data is available for 2017-21 (2017-18 shown for annual data, not three-year rolling average). Spectrum cost (fee paid or amortization accounted) as available through operators’ financial statements, except for Viet 81 Nam where the top two operators (Viettel and VNPT, both state-owned enterprises) do not disclose financial details. For the Philippines, where PLDT’s and Globe’s financial statements embed spectrum user fees in a broader expense category (taxes and licenses), available breakdown of acquisition and amortization of assets is limited and accounting practices for the effective useful life of spectrum assets have not been consistent, therefore the estimate of spectrum cost relied on the operators’ communication to media of around PHP 2.4 billion each in 2021 (BusinessWorld, 2021, Major PHL telcos: Spectrum fees ‘excessive’: https://www.bworldonline.com/corporate/2021/11/30/414041/major-phl- telcos-spectrum-fees-excessive/), assumed at a similar level in 2022 given the only marginal increase in total spectrum revenue at NTC. Denominator is mobile/wireless segment revenue for spectrum cost; and gross revenue for USO payment and corporate income tax. USO payment is estimated based on top two operators’ 2022 disclosure multiplied by percentage levy as studied by Windsor Place Consulting, October 2023, with respective definitions cited in brackets. 64 ITU (2015, 2018) defines 3 classes of 5G use case: enhanced mobile broadband (eMBB), ultra-reliable low-latency 82 communications (uRLLC), and massive machine-type communication (mMTC). eMBB is a natural extension of 4G (LTE) mobile broadband, supported by higher upload/download speed (e.g., gigabit per second) and associated with rich content and human-centric applications (such as mobile ultra-high definition videoconferencing, virtual/augmented reality). uRLLC supports high reliability and low latency (e.g., millisecond response time), suited for mission-critical applications (autonomous vehicles, remote surgery, public safety, and industrial applications). mMTC supports a large number of parallel connections exchanging small data (e.g., million devices per km2), suited for Internet of Things (IoT) applications such as smart city, transport and logistics, intelligent agriculture, and disaster response systems. While viable business models in these use cases are yet to evolve among operators and respective vertical ecosystem participants, governments face regulatory challenges, in particular for spectrum assignment, as the frequency bands available to operators essentially determine technical capacity to deploy different applications (e.g., <1 GHz low band suited for high coverage, 24-47 GHz high band suited for high data rate). In addition, one key component of 5G network design is the potential to use software defined networks (SDNs) to run different types of networks simultaneously over the same physical infrastructure (referred to as ‘network slicing’). These are implemented by ‘virtualising’ the network functions in software and delivered as packaged service beyond data transmission (NaaS – Network as a Service) at low cost, and are configured as separate and dedicated virtual networks for specific applications and user groups. Suppliers claim that such capabilities will extend the basic functionality of mobile services with 5G from entertainment video streaming into running factories, cities, farms, and autonomous vehicles with IoT applications. The possibility of accessing spectrum, including shared or unlicensed spectrum, will be the key determinant of viable offerings. Many governments are evaluating suitable regulatory frameworks—for example, India, through TRAI white papers (TRAI 2019; 2023), examines technical configurations (e.g., network slicing, beamforming, RAN sharing, network elements on cloud), spectrum considerations, and other regulatory issues including active infrastructure sharing, and measures to reach remote and far-flung areas, for stakeholder consultations. In the Philippines, as of December 2023, no mobile operator has invested in NB-IoT technology, meaning that for the most part, any discussion concerning the Internet of Things in 5G networks in the country (e.g. PNA, March 2023 “Building smart cities through 5G technologies, PPPs”, https://www. pna.gov.ph/articles/1198551) is theoretical. For more discussions on 5G strategy for developing countries, see Forge and Vu (2020), and World Bank (forthcoming). Other legislative initiatives include: the Rural Wired Internet Development (RWID) bill, which incentivizes market 83 players to expand wired Internet services in rural areas through tax credits (3-5 years); the Better Internet Act/ Faster Internet Service Act (FISA) bill, which requires service providers to expand Internet coverage and comply with Internet quality of service standards, including minimum download speeds for certain areas and delivery of at least 80 percent of advertised speed. The RWID bill is filed as HB 5469 (Rep. Dionisio) and HB 6922 (Rep. Salceda) and referred to the House ICT Committee; it is undergoing TWG meeting in the Senate facilitated by the Science and Technology Committee, as of December 2023. The Better Internet Act/FISA bill is awaiting a Committee Report by the ICT Committee, as of November 2023; it is pending a Senate hearing by the Public Services Committee, as of November 2023; and the proposed Philippine Spectrum Management Act, amending the Radio Control Law, seeks to ensure the proper, efficient, and equitable distribution of spectrum by mandating the NTC to make spectrum management transparent. It is filed as HB 0896 and is pending at the House ICT Committee, as of July 2022. There is no counterpart Senate bill. The definitions of Open Access approach have evolved over the last decade through global experience sharing and 84 normative deliberations among regulators and policy practitioners, from “the possibility for third parties to use an existing network infrastructure” (ITU 2010) to “wholesale access to network infrastructure or services that is provided effectively on fair and reasonable terms, for which there is some degree of transparency and non-discrimination” (OECD 2013). Such as the Joint Foreign Chambers (JFC) and AmCham, June 2021; the Chamber of Commerce and Industry (PCCI), 85 May 2022; Philippine Exporters Confederation, Inc. (PHILEXPORT), Employers Confederation of the Philippines (ECOP), National ICT Confederation of the Philippines (NICP), Fintech Alliance.PH, Philippine Cable Television Association, Inc. (PCTA), Internet Society - Philippines Chapter, among others, July 2023. 86 On December 12, 2022, the bill was approved on Third reading by the House of Representatives. The bill underwent a TWG meeting facilitated by the Senate Science and Technology Committee, as of December 1, 87 2023. 65 The Legislative-Executive Development Advisory Council (LEDAC) was created by law in 1992. It serves as a 88 consultative and advisory body to the President—Chair of the National Economic and Development Authority (NEDA) Board—on certain programs and policies essential to achieving national economic development. It also serves as a venue to facilitate high-level policy discussions on vital issues and concerns affecting national development. The Common Legislative Agenda (CLA) is a list of priority legislative measures that the LEDAC has agreed to submit to Congress to pass into law. See GOV.PH. [n.d.] The LEDAC in Brief. http://ledac.neda.gov.ph/the-ledac-in-brief/ For the latest CLA list, see Romero (2023). Population covered by a mobile-cellular network, and the population covered by at least a 3G mobile network, both 89 according to ITU. World Bank staff calculation of household penetration of cell phone devices per income quintile based on the PSA’s Annual Poverty Indicator Survey Annual (APIS), 2022. Demand-side interventions to fill the adoption gaps, such as for device affordability, local content including with local languages, and awareness campaigns, have been studied through A4AI (2022), Broadband Commission (2021), among others. Implemented from 2015 to 2019 by researchers from Innovations for Poverty Action, University of the Philippines, and 90 researchers from the University of Washington. Results from the balance tests using data collected prior to tower deployment provide confidence in the randomized 91 assignment of barangays. A more detailed description of coverage, project design, randomized control trial design and results, and implementation is found in Barela, et al. (2023) and Blumenstock, et al. (2020). See A4AI (2022); Broadband Commission (2021); World Bank (2018); USAID (2018); UN ESCAP (2017); and Geerdts 92 (2016). In December 2023, the DICT had a soft launch of the first phase of the National Fiber Backbone (NFB), which according 93 to Undersecretary Jeffrey Dy, will make internet available for government entities and other DICT last-mile projects, such as the Free Public Internet Access Program. DICT. 2023. National Fiber Backbone Phase 1 Soft Launch. https:// www.facebook.com/DICTgovph/videos/2028828200825250 EO 127 s. 2021 on “Expanding the Provision of Internet Services through Inclusive Access to Satellite Services,” 94 amending EO 467 s. 1998, allows telcos and non-telco ISPs and Value-Added Service (VAS) providers direct access to international satellite providers to offer broadband services to end users. See https://www.officialgazette.gov. ph/2021/03/10/executive-order-no-127-s-2021/ Apart from Build, Operate, Transfer (B.O.T.) scheme under the BOT Law, RA 11966 or the PPP Code of the Philippines 95 includes joint ventures; lease agreements for the rehabilitation, operation, and/or maintenance of existing government- owned facilities by at least one (1) year; lease agreements when such lease is a component of a PPP project; and all other contractual agreements which possess characteristics or elements of a PPP. See RA 11966 https://www. officialgazette.gov.ph/downloads/2023/12dec/20231205-RA-11966-FRM.pdf; PwC https://www.pwc.com/ph/en/ publications/2023/pwc-ph-salient-provisions-ppp-code.pdf. For example, on institutional frameworks, observers (e.g., Serafica and Oren 2022) have pointed out strengthening 96 regulatory independence to effectively implement and improve policies. On capacity and resources, the unmet targets in the implementation of the Free Wi-Fi for All program—10.7 percent of the 2026 target 5 years since the project was launched in 2018—reflect, among others, the limitations of implementing agencies. A deeper analysis of these challenges and potential solutions are beyond the scope of this Policy Note. The 19th Congress will end on June 30, 2025. 97 66