## wpiea2021108-print-pdf

## Source details

**Canonical URL:** [wpiea2021108-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021108-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021108-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021108-print-pdf.pdf.json)

---

### I. Introduction — aggregate SDG spending need and context
- Pakistan committed to the 2030 Agenda for Sustainable Development in 2015 and adopted the SDGs as part of its national development agenda in 2016; Voluntary National Review (VNR) in 2019.
- Aggregate additional annual spending required in 2030 from public and private sectors combined: 16.1 percent of GDP.
- Performance highlights from the 2019 VNR and comparisons:
  - Progress: poverty, child stunting, transparency and accountability, gender equality.
  - Below-median performance (for EMDEs) in education, health, electricity, and water and sanitation.
  - Education and water and sanitation also below median for countries with GDP per capita below US$3,000.
- COVID-19 caveat:
  - Analysis completed before the global impact of COVID-19 became clear; does not incorporate pandemic effects on spending needs and GDP projections.
  - Pandemic likely reduces fiscal space and may increase SDG attainment costs.

### II. Education — outcomes, inputs, and required spending
- Current outcomes and inputs:
  - Primary gross enrollment rate: 73 percent in 2003 → 94 percent in 2018.
  - Estimated 23 million children aged 5 to 16 out of school (about 40 percent of students aged 5 to 16) (ADB, 2019).
  - Average years of schooling: 3.3 in 2000 → 5.2 in 2018.
  - Adult literacy rate: 55 percent in 2011 → 59 percent in 2017.
  - Total education spending (public and private): about 4 percent of GDP in 2018.
  - Student-teacher ratio: 22 (overall); primary >40; secondary and tertiary 13.
  - Teacher absenteeism: 24 percent in 2009 → 12 percent in 2016 (ADB, 2019).
  - Only 52 percent of public schools have all four essential infrastructure components; 11 percent have none (ADB, 2019).
- Required increases to meet SDG targets by 2030:
  - Annual total (public and private) education spending needs to increase by about 5.7 percent of GDP (from 3.9 percent in 2018 to 9.6 percent in 2030).
  - Assumed 2030 gross enrollment rates: pre-primary 50 percent, primary 100 percent, secondary 100 percent, tertiary 50 percent.
  - Note: maintaining current pre-primary gross enrollment rate of 83 percent would increase 2030 annual spending needs by 1 percentage point of GDP.
- Key benchmark and result figures (table values preserved):
  - Education spending (percent of GDP): High performers 7.2; 2018 Pakistan 3.9; 2030 Pakistan 9.6.
  - Public: High performers 5.7; 2018 Pakistan 2.4; 2030 Pakistan 7.7.
  - Private: High performers 1.5; 2018 Pakistan 1.5; 2030 Pakistan 1.9.
  - Spending per student (US$ 2018): High performers 6,742; 2018 Pakistan 2,445; 2030 Pakistan 4,570.
  - SDG4 index: High performers 86; 2018 Pakistan 47; 2030 Pakistan >80.
- Main drivers and targets:
  - Target student-teacher ratio: 15 (2018: 22).
  - Teacher wages (ratio to GDP per capita): high performers 2; Pakistan 3.
  - Share of other current and capital spending: high performers 55; Pakistan 30.
  - Private share of total spending: high performers 20; Pakistan 39.

### III. Health — outcomes, inputs, and required spending
- Progress and current gaps:
  - Infant mortality: 106 → 58 deaths per 1,000 live births between 1990 and 2018.
  - Under-5 mortality declined by about 50 percent from 1990 to 2018.
  - Infant and under-5 mortality about twice as high as EMDE and MENAP averages.
- Inputs and efficiency:
  - Pakistan spends 3 percent of GDP on healthcare.
  - Out-of-pocket share of healthcare spending: 65 percent (EMDE average: 35 percent).
  - Doctors per 1,000 population: 0.6 (high performers: 0.9).
  - Other medical personnel per 1,000 population: 0.7 (high performers: 4.3).
  - Spending efficiency: better than other countries with GDP per capita below US$3,000 but far from efficiency frontier.
- Required increases to meet SDG targets by 2030:
  - Health spending needs to increase from 2.8 percent of GDP in 2018 to 8.2 percent in 2030.
  - Public healthcare spending needs to increase from 0.9 percent in 2018 to 4.3 percent in 2030.
- Key benchmark and result figures (table values preserved):
  - Health spending (percent of GDP): High performers 7.6; 2018 Pakistan 2.8; 2030 Pakistan 8.2.
  - Public: High performers 3.9; 2018 Pakistan 0.9; 2030 Pakistan 4.3.
  - Private: High performers 3.7; 2018 Pakistan 1.9; 2030 Pakistan 4.0.
  - Per capita spending (US$ 2018): High performers 1654; 2018 Pakistan 115; 2030 Pakistan 1,151.
  - SDG3 index: High performers 75; 2018 Pakistan 50; 2030 Pakistan >70.
- Human resources and policy notes:
  - Target doctors per 1,000 inhabitants: increase from 0.6 to 0.9.
  - Other medical personnel: increase to almost six times current level.
  - Policy recommendations: increase public share of health spending to improve poor households’ access; consider conditional financial support for training other medical personnel; increase numbers while limiting wage growth relative to GDP per capita.

### IV. Water and Sanitation — coverage and investment needs
- Recent improvements and remaining gaps:
  - Open defecation: 40 percent in 2000 → 10 percent in 2018.
  - Access to basic sanitation facilities: 36 percent in 2000 → 73 percent in 2018.
  - Rural basic sanitation access: 16 percent in 2000 → 60 percent in 2018.
  - Access to non-contaminated drinking water: 44 percent of population.
  - Access to safely managed sanitation facilities on household premises: 46 percent of population.
  - Basic water sources and basic handwashing facilities similar to EMDE and MENAP averages.
- Required investments to achieve universal safely managed services by 2030:
  - Need to expand access to safely managed water to an additional 168 million people and safely managed sanitation to an additional 184 million people; basic handwashing facilities to 131 million people.
  - Aggregate investment required: US$55 billion by 2030.
  - Equivalent annual spending: 2 percent of GDP every year from 2020 to 2030.

### V. Electricity — access, capacity, technical assumptions, and investment needs
- Access and capacity:
  - 8.8 million households remain without grid access; these villages tend to be remote and sparsely populated.
  - Electricity access universal in Punjab and in Azad Jammu and Kashmir, Gilgit-Baltistan; limited in Balochistan (44 percent), Khyber Pakhtunkhwa (96 percent), Sindth (95 percent).
  - Per capita electricity consumption: 452 kWh in 2018 (about 10 percent of MENAP average; 19 percent of EMDE average).
  - Installed generation capacity: 23,663 MW in 2013 → 36,010 MW in 2018 (increase of 52 percent).
  - Government estimate of average capacity utilization factor: around 43 percent.
  - Transmission and distribution losses average about 20 percent.
  - Government plans: significantly increase generation capacity from hydro, nuclear, and renewable sources; plans to double generation capacity from renewable energy sources in the next five years; extend on-grid access to 97 percent of the population (remaining 3 percent off-grid).
- Projected demand and investment needs by 2030:
  - Total electricity consumption expected to increase by 78 percent between 2019 and 2030.
  - Aggregate investment required by 2030: US$24.5 billion.
  - Equivalent annual spending: 0.7 percent of GDP every year from 2020 to 2030.
- Detailed technical assumptions and estimates (preserved exactly):
  - Change in total consumption 2019-2030 (kWh): 109,183,957,844
  - Change in consumption 2019-2030 that cannot be met by current installed capacity (kWh): 61,423,117,980
  - Required change in installed capacity (incl. generation, transmission, and distribution) (kW): 16,255,365
  - Additional required change in transmission and distribution capacity (kW): 13,630,377
  - Assumptions:
    - Coefficient (electricity consumption to GDP per capita): 0.95
    - Plant capacity factor: 43%
    - Unit cost of installed capacity (incl. generation, transmission and distribution) (US$ per kW): 1,059
    - Unit cost - only transmission and distribution capacity (US$ per kW): 530
    - Depreciation per year (percent): 5.0
    - Investment required (US$ billion): 24.5
    - Annual spending 2020-2030 (percent of GDP): 0.7
  - Note: The unit cost per kW (including generation, transmission, and distribution) is estimated at US$ 1,059 based on data from the National Electric Power Regulatory Authority in Pakistan and Ministry of Power.

### VI. Roads — network, rural access, and investment needs
- Network and density:
  - Road length: 250,000 kilometers in 2000 → 269,000 in 2018 (added 19,000 kilometers since 2000).
  - Provincial roads: 124,442 kilometers; federal highways and motorways: 5 percent of total road length.
  - Road density: 66 percent of MENAP average; 58 percent of EMDE average.
  - Government plan to increase road density to 0.45 kilometers per square kilometer by 2025, implying a 38 percent expansion (Planning Commission, 2014).
- Rural connectivity and targets:
  - 61 percent of rural population within 2 km of an all-season road.
  - Road access target proxied by raising Rural Access Index (RAI) to at least 75 percent by 2030.
- Investment required to improve rural access:
  - Increasing rural access from 61 percent to 75 percent by 2030 requires about 94,146 additional kilometers of all-season roads (about a 36 percent increase).
  - Unit cost (US$/Km): 808,098
  - Depreciation per year (percent): 5.0
  - Investment required (US$ billion): 76.1
  - Annual spending 2020-2030 (percent of GDP): 2.3

### VII. Concluding remarks — fiscal constraints, financing, and governance priorities
- Aggregate additional spending needs:
  - Additional annual spending of about 16 percent of GDP in 2030 required for substantial SDG progress in critical sectors.
- Fiscal and financing challenges:
  - High fiscal deficit and need for fiscal consolidation constrain short-run resources.
  - Pakistan’s tax-to-GDP ratio: 11 percent.
  - Medium-term revenue strategy with structural tax reforms is crucial to create space for SDG spending.
  - Private financing is critical; development assistance may be needed to fill remaining gaps.
- Governance and institutional priorities:
  - Strong national ownership and governance frameworks needed: sectorial planning, public investment management, and spending reviews.
  - Structural reforms to boost growth and increase efficiency are needed to mobilize financing and ensure optimal use.
  - Institutional and capacity constraints must be addressed to increase spending efficiency (examples: more qualified teachers; more health workers and facilities; reducing out-of-pocket expenses; raising institutional and technical capacity in rural areas).
- Federal-provincial coordination:
  - The eighteenth amendment of 2010 devolved responsibilities to provinces for health, education, and water and sanitation.
  - Bulk of additional public spending would come from provinces for health, education, and water and sanitation; federal government for roads and electricity.
  - SDG support units have been established; further strengthening needed, notably in data collection and harmonization of tax regulations and procedures.

### VIII. Annex — IMF SDG costing methodology (overview and key formulas)
- General approach:
  - Development outcomes modeled as a function of inputs; annual cost of achieving high performance estimated for education, health, roads, electricity, and water and sanitation.
  - For education and health, key inputs (e.g., number of doctors and teachers) and unit costs set at median values observed in benchmark countries with similar GDP per capita and high SDG index scores.
  - For electricity, roads, and water and sanitation, estimates consider additional investment for growth-induced demand and coverage expansion.
  - Estimates assume higher spending and improved efficiency; if efficiency does not improve to high-performer levels, required spending would be higher.
  - Estimates account for country-specific demographics and GDP per capita projected in 2030.
  - Estimates presented as additional annual spending in 2030; for physical capital, additional spending is annualized to 2030, and after 2030 infrastructure spending assumed to decrease to about 60 percent to cover depreciation.
- Key formula summaries (preserved):
  - Education current spending = (TSR*ER*SAP*AWAGE)/(1-y-z)
    - TSR = teacher to student ratio
    - ER = enrollment rate
    - SAP = school age-population
    - AWAGE = teacher salaries
    - y = share of non-wage current spending
    - z = share of capital spending
  - Health current spending = (DPR*pop*(1+α/ρ)*DAWAGE)/(1-y-z)
    - DPR = doctor density (doctors per 1,000 habitants)
    - pop = total population
    - ρ = ratio of doctors to all other health staff
    - α = ratio of all non-doctor wages to doctor wages
    - DAWAGE = doctor salaries
    - y = share of non-wage current expenses
    - z = share of capital expenses
  - Water and Sanitation: based on WASH World Bank model (Hutton and Varughese, 2016) with three cost estimates (ending open defecation; basic WASH; “safely managed” services).
  - Electricity: cost = capacity needs × Pakistan’s unit cost per kilowatt; accounts for access expansion to 100 percent by 2030 and per capita consumption increases tied to real GDP per capita projections.
  - Roads: regression approach from Gaspar and others (2019) to derive additional kilometers to raise RAI to at least 75 percent; total cost based on Pakistan-specific unit cost per kilometer.

*Source: IMF staff calculations and estimates, pages and tables as reproduced in the supplied content.*

### References  _______________________________________________________________________________________  21

### References (Pages 4–12)

### I. Introduction — Pakistan and the SDGs
- Pakistan committed to the 2030 Agenda for Sustainable Development in 2015 and adopted the SDGs as part of its national development agenda in 2016.
- National measures: comprehensive national SDG framework, taskforces in national and provincial assemblies, Voluntary National Review (VNR) in 2019.
- Performance summary:
  - Progress noted in poverty, child stunting, transparency and accountability, and gender equality (2019 VNR).
  - Current performance in education, health, electricity, and water and sanitation is below the median for Emerging Market and Developing Economies (EMDEs).
  - Education and water and sanitation are also below the median for countries with GDP per capita below US$3,000.
- Aggregate spending need to achieve SDGs in critical sectors (methodology: Gaspar and others (2019); details in Annex):
  - Additional annual spending required in 2030 from public and private sectors combined: 16.1 percent of GDP.
- Sector breakdown of additional needs (high-level):
  - Education: additional spending needs about 5.7 percent of GDP.
  - Health: additional spending needs about 5.4 percent of GDP.
  - Electricity: aggregate investment needed of US$24.5 billion by 2030, equivalent to 0.7 percent of GDP per year from 2020 to 2030.
  - Roads: aggregate investment of US$76 billion by 2030, equivalent to 2.3 percent of GDP per year from 2020 to 2030; requires about 94,146 additional kilometers of all-weather roads (about a 36 percent increase).
  - Water and sanitation: aggregate investment of US$55 billion by 2030, equivalent to 2 percent of GDP per year from 2020 to 2030; bulk of burden from rural sanitation.
- COVID-19 caveat:
  - Analysis completed before global impact of COVID-19 became clear; does not incorporate pandemic effects on spending needs and GDP projections.
  - Pandemic likely reduces fiscal space and may increase SDG attainment costs, though could in some respects hasten progress (e.g., higher health spending, emphasis on handwashing).

### II. Spending Estimates — Education
- Current outcomes and context:
  - Primary gross enrollment rate increased from 73 percent in 2003 to 94 percent in 2018.
  - Estimated 23 million children aged 5 to 16 out of school (about 40 percent of students aged 5 to 16) (ADB, 2019).
  - Gross enrollment rates: pre-primary above peers; secondary and tertiary below peers.
  - Average years of schooling increased from 3.3 in 2000 to 5.2 in 2018.
  - Adult literacy rate improved from 55 percent in 2011 to 59 percent in 2017.
- Inputs and efficiency:
  - Total education spending (public and private) about 4 percent of GDP in 2018.
  - Student-teacher ratio: 22 (above high-performing peer average of 15).
  - Student-teacher ratio by level: over 40 for primary, 13 for secondary and tertiary.
  - Teacher absenteeism reduced from 24 percent in 2009 to 12 percent in 2016 (ADB, 2019).
  - Only 52 percent of public schools have all four essential infrastructure components (boundary walls, electricity, drinking water, toilets); 11 percent have none (ADB, 2019).
- Required increases to meet SDG targets:
  - Annual total (public and private) education spending needs to increase by about 5.7 percent of GDP (from 3.9 percent in 2018 to 9.6 percent in 2030).
  - Assumptions and targets by 2030:
    - Gross enrollment rates: pre-primary 50 percent, primary 100 percent, secondary 100 percent, tertiary 50 percent.
    - Note: maintaining current pre-primary gross enrollment rate of 83 percent would increase 2030 annual spending needs by 1 percentage point of GDP.
  - Main factors driving 2018 → 2030 changes (selected figures from table):
    - Students per teacher ratio: target 15 (2018: 22, high performers: 15).
    - Teacher wages (ratio to GDP per capita): high performers 2, Pakistan 3.
    - Other current and capital spending (percent of total spending): high performers 55; Pakistan 30.
    - Student-age population (percent of total population): high performers 39; Pakistan 42.
    - Enrollment rate (preprimary to tertiary): high performers 59; Pakistan 78 (this row reflects dataset structure in source).
    - Private share (percent of total spending): high performers 20; Pakistan 39.
  - Results (table values):
    - Education spending (percent of GDP): High performers 7.2; 2018 Pakistan 3.9; 2030 Pakistan 9.6.
    - Public: High performers 5.7; 2018 Pakistan 2.4; 2030 Pakistan 7.7.
    - Private: High performers 1.5; 2018 Pakistan 1.5; 2030 Pakistan 1.9.
    - Spending per student (US$ 2018): High performers 6,742; 2018 Pakistan 2,445; 2030 Pakistan 4,570.
    - SDG4 index: High performers 86; 2018 Pakistan 47; 2030 Pakistan >80.

### II. Spending Estimates — Health
- Progress and gaps:
  - Infant mortality decreased from 106 to 58 deaths per 1,000 live births between 1990 and 2018.
  - Under-5 mortality declined by about 50 percent from 1990 to 2018.
  - Infant and under-5 mortality about twice as high as EMDE and MENAP averages; life expectancy somewhat below peers.
- Inputs and efficiency:
  - Pakistan spends 3 percent of GDP on healthcare (lower than peer ranges of around 5.5 to 7.5 percent).
  - Out-of-pocket share of healthcare spending: 65 percent (EMDE average: 35 percent).
  - Doctors per 1,000 population: 0.6 (high performers: 0.9).
  - Other medical personnel per 1,000 population: 0.7 (high performers: 4.3).
  - Spending efficiency: better than other countries with GDP per capita below US$3,000 but still far from efficiency frontier.
- Required increases to meet SDG targets:
  - Health spending needs to increase from 2.8 percent of GDP in 2018 to 8.2 percent in 2030 (almost triple).
  - Public healthcare spending needs to increase from 0.9 percent in 2018 to 4.3 percent in 2030 (almost five-fold).
  - Results (table values):
    - Health spending (percent of GDP): High performers 7.6; 2018 Pakistan 2.8; 2030 Pakistan 8.2.
    - Public: High performers 3.9; 2018 Pakistan 0.9; 2030 Pakistan 4.3.
    - Private: High performers 3.7; 2018 Pakistan 1.9; 2030 Pakistan 4.0.
    - Per capita spending (US$ 2018): High performers 1654; 2018 Pakistan 115; 2030 Pakistan 1,151.
    - SDG3 index: High performers 75; 2018 Pakistan 50; 2030 Pakistan >70.
  - Human resources targets:
    - Doctors per 1,000 inhabitants: increase from 0.6 to 0.9.
    - Other medical personnel: increase to almost six times current level.
  - Policy notes:
    - Doctor wages slightly higher than high-performing peers; focus on increasing numbers while allowing salaries to increase by less than GDP per capita.
    - Increase public share of health spending to improve poor households’ access.
    - Consider financial support for training other medical personnel conditional on in-country service.

### II. Spending Estimates — Water and Sanitation
- Recent improvements and remaining gaps:
  - Open defecation share decreased from 40 percent in 2000 to 10 percent in 2018.
  - Access to basic sanitation facilities increased from 36 percent in 2000 to 73 percent in 2018.
  - Rural basic sanitation access increased from 16 percent in 2000 to 60 percent in 2018.
  - Access to non-contaminated drinking water: 44 percent of population.
  - Access to safely managed sanitation facilities on household premises: 46 percent of population.
  - Access to basic water sources and basic handwashing facilities similar to EMDE and MENAP averages.
- Required investments to achieve universal safely managed services by 2030:
  - Need to expand access to safely managed water to an additional 168 million people and safely managed sanitation to an additional 184 million people; basic handwashing facilities to 131 million people (Table 3 referenced).
  - Aggregate investment required: US$55 billion by 2030.
  - Equivalent annual spending: 2 percent of GDP every year from 2020 to 2030.

### II. Spending Estimates — Electricity (partial)
- Current status and demand projection:
  - Per capita electricity consumption: 452 kWh in 2018 (about 10 percent of MENAP average; 19 percent of EMDE average).
  - Access to on-grid electricity: 77 percent of population (slightly below EMDE average; above countries with GDP per capita below US$3,000).
  - Electricity consumption expected to increase by 78 percent between 2019 and 2030 (reflecting population growth and higher per capita consumption).
- Investment needs and transition:
  - Aggregate investment required by 2030 to meet higher consumption: US$24.5 billion.
  - Equivalent annual spending: 0.7 percent of GDP every year from 2020 to 2030.
  - Planned shift toward renewable energy recommended to bring substantial environmental advantages locally and globally.
- (Notes: figures on installed capacity by source, transmission and distribution investment, and international comparisons are referenced in figures and tables not fully reproduced in this extract.)

### II. Spending Estimates — Roads (partial)
- Current rural access and targets:
  - Current rural access: 61 percent.
  - Target rural access by 2030: 75 percent.
  - Required additional all-weather roads: about 94,146 kilometers (about a 36 percent increase in road length).
  - Aggregate investment required: US$76 billion by 2030.
  - Equivalent annual spending: 2.3 percent of GDP every year from 2020 to 2030.

*Source: IMF staff calculations and estimates, pages and tables as reproduced in the supplied content.*

### 8.8 million households—remain without grid access. These villages tend to be in remote and

### wpiea2021108-print-pdf - 8.8 million households—remain without grid access. These villages tend to be in remote and

### Electricity: access, capacity, and investment needs
- 8.8 million households—remain without grid access; these villages tend to be in remote and sparsely populated areas, making expanding the grid to reach them technically challenging and costly.
- Electricity access is universal in the provinces of Punjab and in the two autonomous territories (Azad Jammu and Kashmir, Gilgit-Baltistan). Access is limited in Balochistan (44 percent), Khyber Pakhtunkhwa (96 percent), and Sindth (95 percent).
- Installed generation capacity growth and utilization:
  - Pakistan increased generation capacity by 52 percent in the last 5 years, from 23,663 megawatts (MW) in 2013 to 36,010 MW in 2018.
  - Government estimate of average capacity utilization factor is around 43 percent.
  - Transmission and distribution losses in Pakistan average about 20 percent.
- Government plans and targets:
  - Plans to significantly increase generation capacity from hydro, nuclear, and renewable sources over the next ten years.
  - Plans to double generation capacity from renewable energy sources in the next five years.
  - The government plans to extend on-grid access to 97 percent of the population; the remaining 3 percent would have off-grid access.
- Projected demand and aggregate investment required by 2030:
  - Total electricity consumption is expected to increase by 78 percent between 2019 and 2030.
  - Aggregate investment required by 2030: US$24.5 billion.
  - Equivalent to annual spending of 0.7 percent of GDP every year from 2020 to 2030.
- Detailed technical assumptions and estimates:
  - Change in total consumption 2019-2030 (kWh): 109,183,957,844
  - Change in consumption 2019-2030 that cannot be met by current installed capacity (kWh): 61,423,117,980
  - Required change in installed capacity (incl. generation, transmission, and distribution) (kW): 16,255,365
  - Additional required change in transmission and distribution capacity (kW): 13,630,377
  - Assumptions:
    - Coefficient (electricity consumption to GDP per capita): 0.95
    - Plant capacity factor: 43%
    - Unit cost of installed capacity (incl. generation, transmission and distribution) (US$ per kW): 1,059
    - Unit cost - only transmission and distribution capacity (US$ per kW): 530
    - Depreciation per year (percent): 5.0
    - Investment required (US$ billion): 24.5
    - Annual spending 2020-2030 (percent of GDP): 0.7
  - Note: The unit cost of investment per kW (including generation, transmission, and distribution costs) is estimated at US$ 1,059 based on data from the National Electric Power Regulatory Authority in Pakistan and Ministry of Power.

### Roads: network, rural connectivity, and investment needs
- Road network and density:
  - Since 2000, Pakistan has added 19,000 kilometers to its road network (from 250,000 kilometers in 2000 to 269,000 in 2018).
  - Most of the road network (124,442 kilometers) is composed of provincial roads; federal highways and motorways represent 5 percent of total road length.
  - Road density in Pakistan is equivalent to 66 percent of the average for other countries in the MENAP region and 58 percent of the average for EMDEs.
  - Government plan to increase road density to 0.45 kilometers per square kilometer by 2025, which would imply expanding the existing road network by 38 percent (Planning Commission, 2014).
- Rural connectivity:
  - 61 percent of the rural population live within 2 km of an all-season road, similar to the EMDE and MENAP averages but above the average for low-income countries.
  - Road access for all is proxied by raising the Rural Access Index (RAI) to at least 75 percent.
- Investment required to improve rural access:
  - Increasing rural access from the current 61 percent to 75 percent by 2030 will require about 94,146 additional kilometers of all-season roads—a 36 percent increase in road length.
  - Unit cost (US$/Km): 808,098
  - Depreciation per year (percent): 5.0
  - Investment required (US$ billion): 76.1
  - Annual spending 2020-2030 (percent of GDP): 2.3

### Concluding remarks: aggregate SDG spending, fiscal constraints, and governance
- Aggregate additional spending needs:
  - Making substantial progress in critical SDG sectors will require additional annual spending of about 16 percent of GDP in 2030.
  - Overall, total (public and private) additional spending needs in Pakistan are substantial at around 16 percent of GDP in 2030.
- Sectoral pressures and COVID-19 caveat:
  - Additional spending needs for education and healthcare in Pakistan are especially large.
  - The analysis was completed before the global impact of COVID-19 became clear; the spending needs and GDP projections in this paper do not reflect the effects of the pandemic.
  - Health spending needs will be correspondingly higher and the resources to finance them will be commensurately scarcer as a result of the pandemic.
- Financing challenges and policy priorities:
  - High fiscal deficit and the need for fiscal consolidation limit availability of resources in the short run.
  - Pakistan’s tax-to-GDP ratio, at 11 percent, is low compared to its peers.
  - A medium-term revenue strategy with structural tax reforms are crucial to create space for SDG spending.
  - Private financing is critical; development assistance may be needed to fill remaining gaps.
  - Strong national ownership and governance frameworks are necessary, including sectorial planning, public investment management, and spending reviews.
  - Structural reforms to boost growth and increase efficiency are needed to mobilize financing and ensure optimal use.
  - Institutional and capacity constraints must be addressed to increase spending efficiency (examples: more qualified teachers to reduce class sizes; more health workers and facilities; reducing out-of-pocket expenses; raising institutional and technical capacity in rural areas).
- Federal-provincial coordination:
  - The eighteenth amendment of 2010 devolved responsibilities to provinces for health, education, and water and sanitation.
  - The bulk of additional public spending would come from the provinces for health, education, and water and sanitation and from the federal government for roads and electricity.
  - SDG support units have been established to support coordination, but further strengthening is needed, notably in data collection to monitor SDG progress and harmonization of tax regulations and procedures.

### Annex: IMF SDG costing methodology (overview and key formulas)
- General approach:
  - The methodology assumes development outcomes are a function of a mix of inputs and estimates the annual cost of achieving high performance across five SDG areas—education, health, roads, electricity, and water and sanitation.
  - For education and health, the methodology sets key inputs (such as number of doctors and teachers) and their unit costs at median values observed in benchmark countries with similar GDP per capita that perform well in the SDG index scores.
  - For electricity, roads, and water and sanitation, estimates consider additional investment required to support economic growth and expand coverage following SDG targets.
  - Estimates assume higher spending and “better” spending (higher efficiency). If efficiency does not improve to the level of high performers, required spending would be higher.
  - Estimates account for country-specific factors such as demographics and GDP per capita projected in 2030.
  - Estimates are presented as additional annual spending in 2030. For education and health, additional spending is the difference between required spending in 2030 and current spending. For physical capital (roads, electricity, water and sanitation), additional spending is the annualized spending required to achieve SDG performance; after 2030 infrastructure spending would decrease to about 60 percent to cover depreciation of the capital stock built through 2030.
- Education (formula and inputs):
  - Current education spending = (TSR*ER*SAP*AWAGE)/(1-y-z)
    - TSR = teacher to student ratio
    - ER = enrollment rate
    - SAP = school age-population
    - AWAGE = teacher salaries
    - y = share of non-wage current spending
    - z = share of capital spending
  - Total needed education spending derived by setting TSR, AWAGE, y, and z at median values from benchmark high-performing countries, accounting for economic growth and school-age demographics.
- Health (formula and inputs):
  - Current health spending = (DPR*pop*(1+α/ρ)*DAWAGE)/(1-y-z)
    - DPR = doctor density (doctors per 1,000 habitants)
    - pop = total population
    - ρ = ratio of doctors to all other health staff
    - α = ratio of all non-doctor wages to doctor wages
    - DAWAGE = doctor salaries
    - y = share of non-wage current expenses
    - z = share of capital expenses
  - Total needed health spending derived by setting DAWAGE, DPR, and ρ at median values from benchmark high-performing countries and using Pakistan-specific projections for economic and population growth.
- Water and Sanitation:
  - Based on the WASH World Bank model (Hutton and Varughese, 2016) with three cost estimates: ending open defecation via lower-cost latrines; achieving lower-level basic WASH services; and achieving “safely managed” services (on-plot water supply and safe fecal waste management).
- Electricity:
  - Cost based on estimating additional electricity network needed to provide access to 100 percent of the projected population by 2030 and accounting for per capita consumption increases based on real GDP per capita projections; total cost estimated by multiplying capacity needs by Pakistan’s unit cost per kilowatt.
- Roads:
  - Regression approach from Gaspar and others (2019) used to derive determinants of network needs; additional kilometers estimated to raise RAI to at least 75 percent given population and GDP per capita in 2030; total cost estimated using Pakistan-specific unit cost per kilometer.

*Source: IMF staff estimates and IMF paper text as provided.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021108-print-pdf.pdf_
