## _wp16187

## Source details

**Canonical URL:** [_wp16187](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16187.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16187.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16187.pdf.json)

---

### Major themes and scope
- Reviews the history of Local Government Financing Vehicles (LGFVs) and Public-Private Partnerships (PPPs) in China, analyzes recently introduced PPP regulatory framework, and makes recommendations for further reform.
- Organization:
  - Section I: background on LGFVs in China, links to PPPs and history.
  - Section II: government policies responding to fiscal risks from LGFVs and issues with recent measures.
  - Section III: proposes a four-pillar PPP regulatory framework based on international experience.
  - Section IV: translates the framework into a three-stage reform strategy.
  - Section V: conclusion.

### Background: LGFVs and PPPs
- Definitions and mechanics:
  - LGFVs: companies set up and owned by subnational governments (SNGs) to finance and implement public infrastructure projects.
  - Typical LGFV repayment mechanism: repayment financed by proceeds from the sale of government-owned land near infrastructure projects.
  - Land-sale proceeds: only proceeds net of resettlement payments to displaced households are available for use; these payments are estimated to be around 40 percent of land sale proceeds.
- Institutional context:
  - Prior to the budget law revision in August 2014 (effective in 2015), SNGs could not borrow "on budget" without central government approval, incentivizing off-budget LGFV borrowing since the late-1990s.
  - LGFVs share structural features with PPPs (special-purpose vehicles, long-term project responsibilities); the paper argues LGFVs were a specific type of unregulated PPPs.
- Roles LGFVs take:
  - As project contractor: SNG is project originator/project owner and signs a contract with the LGFV to build and operate the infrastructure.
  - As project originator/project owner: LGFV signs a contract with another enterprise (private or public) acting as project contractor; often another special purpose vehicle (project company) is created.

### LGFV scale, trajectory, and fiscal exposure (exact figures)
- According to a 2013 NAO report: total LGFV debt stood at 7.0 trillion RMB (13.1 percent of 2012 GDP) as of end-June 2013.
- LGFV debt measured at 10.9 trillion RMB (20.4 percent of 2012 GDP) at end-June 2013, which increased to 15.4 trillion RMB at end-2014 (24.2 percent of 2014 GDP).
- Total local government contingent liability increased from 7.0 trillion RMB (12.1 percent of GDP) to 8.6 trillion RMB (13.5 percent of GDP) over the same period.
- Pressure to repay maturing debt is particularly high for highways built by LGFVs.

### Policy response since 2013 (measures and objectives)
- Three policy dimensions:
  1. Overall fiscal management reform.
  2. Relaxing SNG fiscal rules and developing domestic municipal bond market.
  3. Introducing a new PPP regulatory framework and promoting a new generation of PPPs.
- Key legal and institutional changes:
  - 2014 budget law revision:
    - Lifts the prohibition on local government borrowing.
    - Provincial governments allowed to borrow up to a central-government-set ceiling for capital spending only.
    - Expanded information disclosure requirements and fiscal responsibility oversight.
    - Requires inclusion of government-management funds, state-owned assets, and social security funds in the budget document.
    - Prohibits LGFVs from financing local governments going forward.
    - Introduces a three-year medium-term fiscal framework.
- Implementation activity (exact reported quantities and dates):
  - Since 2013, over 40 PPP regulatory documents issued by State Council, NDRC, MOF, and others.
  - MOF introduced 233 pilot PPP projects in 2014-2015, worth about 800 billion RMB.
  - NDRC and MOF databases (by April 2016):
    - NDRC: over 2,000 projects worth about 3.5 trillion RMB.
    - MOF: about 7,700 projects worth about 8.8 trillion RMB.
  - MOF and various local governments created PPP units to centralize PPP regulation.

### Issues and implementation challenges with new measures
- Potential bias toward PPPs:
  - PPPs change timing of cash flows rather than necessarily reduce net present value (NPV) of government spending.
  - Budget and medium-term frameworks are shorter than PPP lifetimes, creating bias toward PPPs in cash-based systems.
  - Recommendation: apply Value for Money (VfM) tests to determine suitability of PPPs versus traditional procurement.
- Limited private-sector participation:
  - “Social capital” often includes SOEs; many private investors remain hesitant.
  - SOEs enjoy advantages: experience, better bank access, and stronger dispute-position given government connections; this undermines expansion of truly private participation.
- Coordination and regulatory fragmentation:
  - State Council, NDRC, and MOF have issued differing framework documents; roles of NDRC and MOF over PPP lifecycle remain unclear.
  - Fragmentation risks mirror past LGFV-related fiscal risks when projects are approved without fiscal checks.
- Weak post-contract oversight:
  - Current focus on pre-contract appraisal; post-contract regulation (renegotiations, ex-post audits) lacks detail.
  - Recommendation: PPP unit approval for renegotiations and strengthened audit-office ex-post audits.

### International best-practice framework — four pillars (policy recommendations)
- 1. Good project selection:
  - Integrate PPPs and traditional investments into capital budget cycle, medium-term fiscal framework, and public investment strategy.
  - Three-step decision process:
    - Step 1: economic and social appraisal (cost-benefit) and inclusion in investment planning and medium-term fiscal framework.
    - Step 2: choose procurement form (PPP vs traditional) based on VfM.
    - Step 3: if PPP chosen, apply a gateway process (due-diligence gates where projects can be stopped if fiscally unaffordable or too risky).
  - Clarify NDRC and MOF roles; MOF should ensure VfM and fiscal-affordability checks at each gateway.
  - Potential national PPP pipeline for projects above thresholds consistent with national priorities.
- 2. Good institutional framework:
  - A dedicated PPP unit (or public investment unit) preferably located in the MOF to assess fiscal affordability and control the gateway process.
  - China has established a PPP center in MOF in 2014; recommendation to coordinate closely with NDRC and consider joint MOF/NDRC arrangements or a “one-stop shop.”
  - PPP unit coordination with PBoC and CBRC is necessary because major state-owned banks are key creditors of LGFVs/PPPs.
  - Strengthen post-contractual regulation and audits; PPP unit should approve renegotiations and audit offices should perform ex-post audits.
- 3. Good laws:
  - Consolidate into a PPP framework law (or consolidated high-level regulatory document) covering scope, budget integration, role assignment (MOF as gatekeeper), competitive bidding, renegotiation/termination guidelines, aggregate exposure limits, and transparent accounting/reporting.
  - Recommendation for China: consolidate existing framework documents into a single high-level PPP regulatory document; pilot before elevating to PPP law.
  - Clarify whether disputes are governed by civil or administrative law; civil law generally places partners on equal commercial footing and may encourage private participation.
- 4. Good accounting and reporting:
  - International standards: IPSAS 32 (accounting), IMF GFSM 2014 (reporting), and PSDS 2011 (debt statistics); all accrual-based.
  - Adoption of IPSAS 32, GFSM 2014, and PSDS 2011 would, in practice, lead to most PPPs being treated on-budget.
  - Under IPSAS 32, projects where government controls/regulates services, prices, or recipients should be considered public and affect fiscal aggregates.
  - Transition to accrual standards should be gradual and sequenced; initial steps could include disclosure in budget or complementary budget documents of contract value and long-term implications.
  - IMF–World Bank PPP Fiscal Risk Assessment Model (PFRAM) can assess fiscal risks from individual PPPs based on IPSAS 32.

### PPP ceilings and disclosure
- Rationale:
  - Fiscal rules can be circumvented by PPPs, especially under cash-based accounting and limited headline fiscal coverage.
  - SNGs used LGFVs to circumvent past borrowing prohibitions and retain incentives to use PPPs to circumvent recently imposed debt ceilings.
- Potential PPP ceilings for SNGs could include:
  - PPP contract value over current revenue ratio (excluding land sale proceeds) or over GDP ratio.
  - PPP debt over current revenue or GDP ratio.
  - Government commitments in PPPs over current revenue or GDP ratio.
- Disclosure and partial balance sheet:
  - Appendix 4 proposes disclosure requirements for PPPs and guarantees.
  - With enhanced disclosure, a partial government balance sheet could be gradually compiled to capture PPPs and related contingent liabilities.
  - MOF classification of SNG debt from the 2013 NAO report into three types: (1) general obligations, (2) obligations arising from specific revenue-generating projects, and (3) debt converted to company debt through PPPs.
  - MOF documents require disclosure of government commitments in PPPs in the government’s comprehensive fiscal report, when such a reporting system is ready.

### Three-stage strategy to establish a PPP regulatory framework in China (exact timelines)
- Near term (1-2 years) — central government aims to:
  - Classify and disclose existing LGFV projects and new PPPs as part of budget documents.
  - Ensure conversion of LGFVs to PPPs is not mere “relabeling” or “reclassification”.
  - Evaluate introduction of ceilings/limits for PPPs at central and subnational levels.
  - Gradually introduce PPP regulations in line with seven key areas; urgently consolidate existing framework documents (over 40 issued since 2013).
  - Improve coordination between MOF, NDRC, PBoC, CBRC over LGFVs and PPP issues.
- Medium term (2–5 years) — elements to take shape:
  - Issue a consolidated PPP regulatory document by the State Council; integrate central and SNG PPP projects into the normal budgetary process and people’s congress scrutiny.
  - Expand MOF PPP unit into a national PPP/public investment unit as a “one-stop shop”; include NDRC and line ministries and coordinate with PBoC and CBRC.
  - Mandate unit oversight of fiscal risks for all major PPP projects above a threshold; require monitoring submissions for projects below threshold.
  - Consolidate PPP contract guidelines into standardized contracts and procedures.
  - Allow SNGs to outsource evaluation to the national unit while building subnational capacity.
  - Enforce PPP ceilings at central and subnational levels, supervised by central government and subnational people’s congresses.
  - Gradually improve PPP reporting in line with international standards and the plan of statistical improvement.
- Longer term (5-10 years) — framework aligned with international best practice:
  - After 5-10 years of experience, upgrade consolidated PPP regulatory document into a PPP framework law passed by the NPC.
  - Make monitoring of PPP projects regular in annual budgets and medium-term fiscal frameworks at all levels.
  - Enable national PPP/public investment unit to oversee fiscal risks of major projects, with provincial units overseeing smaller projects under a consistent framework.
  - Regularly disclose detailed information about PPP projects and their fiscal impact in near-, medium-, and long-term budget documents.

### Findings on risks, biases, and empirical examples
- General findings:
  - PPPs are long-term contracts; potential efficiency gains can be offset by higher private borrowing costs, company profits, and transaction costs.
  - Cash-based budgeting creates a bias in favor of PPPs because PPPs may appear “off-budget” short-term even if NPV is unchanged ceteris paribus.
  - Accrual-based accounting (e.g., IPSAS 32) helps capture true fiscal implications and reduces bias.
- Empirical examples of fiscal risk materialization:
  - Mexico in mid 1990s: government eventually assumed about US$7.7 billion in debt after taking over private toll road concessions.
  - Hungary in late 1990s: M1 highway traffic forecasts proved too optimistic and the government assumed traffic risk.
  - Portugal in early 2000s: highways built using PPPs were reclassified as central government liabilities after the 2009 crisis, adding substantially to deficit and debt.
  - Spain in early 2000s: local airports and railways built by local governments using PPPs faced traffic shortfalls, closures, and persistent long-term commitments financed by transfers.

### Appendix 3 — IPSAS 32 (Service Concession Arrangements: Grantor) — key accounting rules and example
- Overview:
  - Formal name: “Service Concession Arrangements: Grantor”.
  - Issued by the IPSASB in October 2011; compatible with IFRIC 12.
  - Scope: covers both government-funded and user-funded PPP contracts; accrual-based.
- Recognition criteria (assets and liabilities recorded on grantor’s balance sheet if both met):
  - (a) Grantor controls or regulates what services the operator must provide, to whom, and at what price.
  - (b) Grantor controls any significant residual interest in the asset at the end of the term.
  - For whole-of-life assets, only (a) needs to be met.
- Accounting treatment — government-funded projects (financial liability model):
  - Initial recognition: grantor records same amount of asset and liability at fair value.
  - Subsequent accounting: expense items include asset depreciation, finance charge (interest), charges for services paid to operator; principal repayment reduces liability.
- Accounting treatment — user-funded projects (grant of a right to the operator model):
  - Initial recognition: grantor records same amount of asset and liability at fair value.
  - Subsequent accounting: expense recorded is asset depreciation only; grantor records imputed revenue accrued during contract period; liability reduction equivalent to accrued revenue.
  - Cash flow implication: no cash flows for grantor in user-funded PPPs; government-funded PPPs require cash payments.
- Mixed funding: divide project into government-funded and user-funded parts and apply rules respectively.
- Illustrative BOT road example (exact figures from example):
  - Construction costs: road base cost $940; road surface cost $110.
  - Usable life: base 25 years; surface 6 years (surface re-constructed in year 8).
  - Financing approaches: (1) government-funded: government pays operator $200 per year during years 3-10; (2) user-funded: operator charges users $200 per year during years 3-10.
  - Illustration outcomes:
    - Government nonfinancial assets recorded identical in both user-funded and government-funded PPPs.
    - Government liabilities broadly similar; user-funded liability slightly lower because it does not incur interest cost.
    - Government cash flows: no cash flows in user-funded PPP; government-funded PPP requires cash payments.
    - Government net lending/borrowing (accrual surplus/deficit): higher in user-funded PPP due to imputed accrued revenue.
- Policy implication:
  - IPSAS 32 treats PPPs similarly to traditional procurement, reducing government bias in favor of PPPs and placing decisions primarily on expected efficiency gains.

*Source: _wp16187 (IMF working paper content excerpt)*

### References .............................................................................................................

### _wp16187 - References

### Major themes and scope
- The paper reviews the history of Local Government Financing Vehicles (LGFVs) and Public-Private Partnerships (PPPs) in China, analyzes recently introduced PPP regulatory framework, and makes recommendations for further reform.
- Organization of the paper:
  - Section I: background on LGFVs in China, links to PPPs and history.
  - Section II: government policies responding to fiscal risks from LGFVs and issues with recent measures.
  - Section III: proposes a four-pillar PPP regulatory framework based on international experience.
  - Section IV: translates the framework into a three-stage reform strategy.
  - Section V: conclusion.

### Background: LGFVs and PPPs
- LGFVs are companies set up and owned by subnational governments (SNGs) to finance and implement public infrastructure projects.
- Prior to the budget law revision in August 2014 (effective in 2015), SNGs could not borrow "on budget" without central government approval, leading to widespread use of LGFVs as an "off-budget" borrowing mechanism since the late-1990s.
- Typical LGFV repayment mechanism:
  - Repayment of loans taken by LGFVs was typically financed by proceeds from the sale of government-owned land near infrastructure projects.
  - Once infrastructure projects were completed, nearby land typically appreciated sharply (during the 2000s economic boom) and was sold to repay LGFV debt.
- Relationship to PPPs:
  - By comparing LGFVs and typical PPPs, the paper argues that LGFVs were a specific type of unregulated PPPs.
  - The Chinese authorities are promoting PPPs with a new regulatory framework as a new model for infrastructure development.

### Fiscal risks and developments
- Expansion and risks:
  - The number and size of LGFVs expanded sharply in the 2008 stimulus package and posed significant fiscal risks.
  - As growth of land sale proceeds slowed and did not keep up with LGFV debt increases, concerns about the sustainability of the LGFV financing model grew.
- Precise reported figure:
  - According to a 2013 report issued by the National Audit Office (NAO), total LGFV debt stood at 7.0 trillion RMB (13.1 percent of 2012 GDP) as of end-June 2013.

### Policy response and assessment
- Chinese policy responses include:
  - Relaxation of on-budget SNG borrowing constraints.
  - Active promotion of PPPs and a corresponding regulatory framework as a new model for infrastructure financing.
- Assessment:
  - The shift toward regulated PPPs is characterized as a positive step forward.
  - The paper notes that there remains "much room to improve the regulatory framework."

### Tables, figures, and appendixes (as listed in the source)
- Tables:
  - 1. Selected PPP Regulatory Documents Issued by the Chinese Government Since 2013
  - 2. Comparison of Three Framework Regulatory Documents Issued by the Chinese Government on Public-Private Partnerships with International Best Practice
  - 3. A Potential Gateway Process for China
- Figures:
  - 1. Government Managed-fund Income from Land Usage Right Transfer in China
  - 2. Local Government Financing Vehicle as the Project Contractor
  - 3. Local Government Financing Vehicle as the Project Owner
- Appendixes:
  - 1. Examples of Local Government Financing Vehicles in China
  - 2. A Primer on PPPs
  - 3. International Public Sector Accounting Standard 32 (IPSAS 32)
  - 4. IMF Proposal of Disclosure Requirements for PPPs and Guarantees
- Appendix Figures:
  - 1. Traditional Public Procurement versus Public-Private Partnership
  - 2. NPV: Government Procurement versus PPP
  - 3. IPSAS 32 Illustrative Examples

*Source: _wp16187 - References*

### 10.9 trillion RMB (20.4 percent of 2012 GDP) at end-June 2013, which increased to 15.4 trillion

### _wp16187 - 10.9 trillion RMB (20.4 percent of 2012 GDP) at end-June 2013, which increased to 15.4 trillion

### LGFV scale, trajectory, and fiscal exposure
- LGFV debt measured at 10.9 trillion RMB (20.4 percent of 2012 GDP) at end-June 2013, which increased to 15.4 trillion RMB at end-2014 (24.2 percent of 2014 GDP).
- Total local government contingent liability increased from 7.0 trillion RMB (12.1 percent of GDP) to 8.6 trillion RMB (13.5 percent of GDP) over the same period.
- Pressure to repay maturing debt is particularly high for highways built by LGFVs.
- Note on land-sale proceeds: only proceeds net of resettlement payments to displaced households are available for use; these payments are estimated to be around 40 percent of land sale proceeds (i.e. income from land usage right transfer).

### How LGFVs operate (organizational features and common practices)
- Typical structure:
  - LGFV established as a special purpose vehicle controlled by a subnational government (SNG).
  - SNG transfers “high-quality assets” to the LGFV to improve creditworthiness; such assets may include public land and shares of public utilities (e.g., water, sewage, public transportation).
  - Land and cash flows from public utilities are used as collateral when borrowing for infrastructure development.
- Roles LGFVs can take:
  - LGFV as project contractor: SNG is project originator/project owner and signs a contract with the LGFV to build and operate the infrastructure.
  - LGFV as project originator/project owner: LGFV signs a contract with another enterprise (private or public) acting as project contractor; often another special purpose vehicle (project company) is created.
- Consequences:
  - A web of special purpose vehicles is frequently created to implement infrastructure on behalf of SNGs.
  - Prior to 2010, none of these entities were classified as budgetary units in China; their transactions were not recorded on budget.

### Linkages between LGFVs and PPPs
- PPP definition: long-term contracts between the government and a private contractor to build public infrastructure and provide services; usually implemented via a special-purpose vehicle.
- Historical context in China:
  - By 2010, PPPs had two generations; LGFVs belong to the second generation (2000s), with SOEs as major private-sector players.
  - First generation (1990s) featured foreign companies and led to problems with guaranteed minimum returns and large renegotiations; central government intervened in 2002 to prohibit fixed-return guarantees to foreign companies.
- Reasons LGFVs resemble unregulated PPPs:
  - Long-term project responsibilities beyond annual budgetary cycle.
  - Structural similarity to PPPs with special purpose vehicles.
  - State-owned enterprise status and lack of regular budget constraints.
- Incentives and risks:
  - Pre-2015 balanced budget rule forbade borrowing without central approval, incentivizing SNGs to use LGFVs to deliver infrastructure off budget.
  - Before 2010, LGFVs’ firm and contingent liabilities were typically not disclosed in annual budgets, financial statements, and fiscal statistics.
  - Rapid land-price appreciation masked liquidity issues; fiscal risks from expanding LGFV debt rose to prominence starting in 2010.
- Policy implication articulated: all PPP projects should be subject to a strong regulatory framework to balance infrastructure development and fiscal sustainability.

### Policy response since 2013 (measures and objectives)
- Three policy dimensions undertaken since 2013:
  1. Overall fiscal management reform.
  2. Relaxing SNG fiscal rules and developing domestic municipal bond market.
  3. Introducing a new PPP regulatory framework and promoting a new generation of PPPs.
- Key changes:
  - 2014 budget law revision:
    - Lifts the prohibition on local government borrowing.
    - Provincial governments allowed to borrow up to a central-government-set ceiling for capital spending only.
    - Expanded information disclosure requirements and fiscal responsibility oversight.
    - Requires inclusion of government-management funds, state-owned assets, and social security funds in the budget document.
    - Prohibits LGFVs from financing local governments going forward.
    - Introduces a three-year medium-term fiscal framework.
- PPP promotion objectives stated by government:
  1. Accelerate government reform to reduce microeconomic role and enhance market regulatory capacity.
  2. Remove red tape and encourage private capital to provide public services.
  3. Improve fiscal management and efficiency of budget spending; PPPs expected to reduce current-year budgetary spending needs and distribute public investment financing over generations.
- Implementation activity:
  - Since 2013, over 40 PPP regulatory documents issued by State Council, NDRC, MOF, and others.
  - MOF introduced 233 pilot PPP projects in 2014-2015, worth about 800 billion RMB.
  - NDRC and MOF databases (by April 2016):
    - NDRC: over 2,000 projects worth about 3.5 trillion RMB.
    - MOF: about 7,700 projects worth about 8.8 trillion RMB.
  - MOF and various local governments created PPP units to centralize PPP regulation.

### Issues and implementation challenges with new measures
- Potential bias toward PPPs:
  - PPPs change timing of cash flows rather than reduce net present value (NPV) of government spending; PPPs can change form of liabilities but not necessarily lower total fiscal cost.
  - Budget and medium-term frameworks are shorter than PPP lifetimes, which can bias authorities toward PPPs.
  - Recommendation: apply Value for Money (VfM) tests to determine suitability of PPPs versus traditional procurement.
- Limited participation by truly private partners:
  - “Social capital” often includes SOEs; many private investors remain hesitant.
  - SOEs enjoy advantages: experience, better bank access, and stronger dispute-position given government connections.
  - A market dominated by SOEs undermines goals to expand private-sector role; a level playing field is needed.
- Coordination and regulatory fragmentation:
  - State Council, NDRC, and MOF have issued differing framework documents; roles of NDRC and MOF over PPP lifecycle remain unclear.
  - Fragmentation risks: projects approved by NDRC/local DRCs without MOF/local BOFs fiscal checks—mirrors past LGFV-related fiscal risks.
- Post-contract oversight weakness:
  - Current focus on pre-contract appraisal; post-contract regulation, including approval of renegotiations and ex-post audits, lacks detail.
  - Recommendation: PPP unit approval required for renegotiations, and audit offices should strengthen post-contractual and ex-post audits.

### International best-practice framework and implications for China
- Four recommended elements for effective PPP fiscal-risk management:
  1. Good project selection:
     - Integrate PPPs and traditional investments into capital budget cycle, medium-term fiscal framework, and public investment strategy.
     - Three-step decision process:
       - Step 1: economic and social appraisal (cost-benefit) and inclusion in investment planning and medium-term fiscal framework.
       - Step 2: choose procurement form (PPP vs traditional) based on VfM.
       - Step 3: if PPP chosen, apply a gateway process (due-diligence gates where projects can be stopped if fiscally unaffordable or too risky).
     - Clarification of NDRC and MOF roles across these steps is paramount; MOF should ensure VfM and fiscal-affordability checks at each gateway.
     - Potential consolidation: a national PPP pipeline for projects above thresholds consistent with national priorities.
  2. Good institutional framework:
     - A dedicated PPP unit (or public investment unit) is helpful and preferably located in the MOF to assess fiscal affordability and control the gateway process.
     - China has established a PPP center in MOF in 2014; recommendation to coordinate closely with NDRC and consider joint MOF/NDRC arrangements or a “one-stop shop.”
     - PPP unit coordination with PBoC and CBRC is necessary because major state-owned banks are key creditors of LGFVs/PPPs.
     - Strengthen post-contractual regulation and audits; PPP unit should approve renegotiations and audit offices should perform ex-post audits.
  3. Good laws:
     - Best practice favors a PPP framework law (or consolidated high-level regulatory document) covering scope, budget integration, role assignment (MOF as gatekeeper), competitive bidding, renegotiation/termination guidelines, aggregate exposure limits, and transparent accounting/reporting.
     - Recommendation for China: consolidate existing framework documents into a single high-level PPP regulatory document; later consider elevation to PPP law after piloting.
     - Clarify whether disputes are governed by civil or administrative law; civil law generally places public and private partners on equal commercial footing and may encourage private participation.
  4. Good accounting and reporting:
     - International standards: IPSAS 32 (accounting), IMF GFSM 2014 (reporting), and PSDS 2011 (debt statistics); all accrual-based.
     - Adoption of IPSAS 32, GFSM 2014, and PSDS 2011 would, in practice, lead to most PPPs being treated on-budget.
     - Under IPSAS 32, projects where government controls/regulates services, prices, or recipients should be considered public and affect fiscal aggregates.
     - China’s cash-based government accounting tends to underestimate PPP fiscal risks, especially in early project stages.
     - Transition to accrual standards will take time and should be gradual, carefully sequenced, and tailored to China’s circumstances; initial steps could include disclosure in budget or complementary budget documents of contract value and long-term implications.
     - The IMF–World Bank PPP Fiscal Risk Assessment Model (PFRAM) exists to assess fiscal risks from individual PPPs based on IPSAS 32.

*Source: CEIC China database, IMF staff estimate; content excerpted from the provided IMF working paper text.*

### 32. As more and more projects are disclosed in this way, a partial government balance sheet

### 32. As more and more projects are disclosed in this way, a partial government balance sheet could be gradually compiled

### Partial balance sheet and disclosure
- Appendix 4 describes a suggested proposal of disclosure requirements for PPPs and guarantees.
- With enhanced disclosure, a partial government balance sheet could be gradually compiled to capture PPPs and related contingent liabilities.
- The MOF has classified SNG debt identified by the 2013 NAO report into three types: (1) general obligations, (2) obligations arising from specific revenue-generating projects, and (3) debt converted to company debt through PPPs.
- The MOF documents require the disclosure of government commitments in PPPs in the government’s comprehensive fiscal report, when such a reporting system is ready.

### PPP ceilings to complement fiscal rules
- Rationale:
  - Fiscal rules (ceilings on debt, deficit, and/or spending) can be circumvented by PPPs, especially under cash-based accounting and limited coverage of headline fiscal indicators.
  - In China, before 2015 SNGs effectively faced a rule of “no borrowing without central government approval”; post-2015, provincial governments are subject to debt ceilings set by the central government.
  - SNGs used LGFVs to circumvent borrowing prohibitions in the past and retain incentives to use PPPs to circumvent recently imposed debt ceilings.
- Potential PPP ceilings for SNGs in China could include:
  - PPP contract value over current revenue ratio (excluding land sale proceeds) or over GDP ratio.
  - PPP debt over current revenue or GDP ratio.
  - Government commitments in PPPs over current revenue or GDP ratio.

### A three-stage strategy to establish a PPP regulatory framework in China
- Overview:
  - Strategy follows a three-stage sequencing: near-term, medium-term, and longer-term.
  - Starts at the central government level and gradually expands to SNGs once capacity is strengthened.
  - Careful policy sequencing is emphasized.

- Near term (1-2 years) — central government aims to:
  - Classify and disclose existing LGFV projects and new PPPs as part of budget documents.
  - Ensure that converting LGFVs to PPPs is not limited to “relabeling” or “reclassification”; more effective regulation is required.
  - Evaluate the introduction of ceilings/limits for PPPs at central and subnational levels, considering current outstanding debt, infrastructure needs, and economic growth prospects.
  - Gradually introduce PPP regulations in line with the seven key areas discussed in the report; note that since 2013 over 40 regulatory documents have been issued by central government entities, and consolidation—especially of three framework documents into a single framework document—is urgently needed.
  - Improve coordination between MOF, NDRC, PBoC, CBRC over LGFVs and other PPP issues.

- Medium term (2–5 years) — elements to take shape:
  - Issue a consolidated PPP regulatory document by the State Council; new central and SNG PPP projects should follow national PPP regulation, be integrated into the normal budgetary process, and be subject to people’s congress scrutiny.
  - Expand the PPP unit in the MOF into a national PPP/public investment unit as a “one-stop shop” for PPP regulatory issues; include representatives of NDRC and line ministries, and coordinate with PBoC and CBRC.
  - Mandate the unit to oversee fiscal risks of all major PPP projects above a certain threshold, regardless of level of government; require key documents of projects below the threshold to be submitted for monitoring.
  - Consolidate existing PPP contract guidelines (the two PPP contract guidelines already issued by the MOF and NDRC) into standardized contracts and procedures for selecting, evaluating, and approving PPP projects.
  - Allow SNGs to outsource evaluation of main PPP/investment projects to the national unit while subnational capacity is developed.
  - Enforce PPP ceilings at both central and subnational government levels; enforcement for SNGs would be supervised by both the central government and the subnational people’s congress.
  - Gradually improve reporting of PPP transactions in line with international standards and the plan of statistical improvement.

- Longer term (5-10 years) — framework aligned with international best practice:
  - After 5-10 years of experience, upgrade the consolidated PPP regulatory document into a PPP framework law passed by the NPC.
  - Make monitoring of PPP projects a regular part of the annual budget and the medium-term fiscal framework of all government levels.
  - Enable the national PPP/public investment unit to effectively oversee fiscal risks of major projects, with provincial PPP/public investment units overseeing smaller projects within a consistent oversight framework.
  - Regularly disclose detailed information about PPP projects and their fiscal impact in the near-, medium-, and long-term in budget documents.

### Findings on risks, biases, and the need for regulatory strengthening
- PPPs:
  - Are long-term contracts where a private contractor builds, operates, and maintains an asset; government remains accountable for service provision and promises payments or allows user fees.
  - Tend to offer potential efficiency gains through private-sector innovation, but these can be offset by higher private borrowing costs and higher transaction costs.
  - Can create a government bias in favor of PPPs under cash-based budgeting because PPPs may appear “off-budget” in the short term even though they do not change total net present value (NPV) ceteris paribus.
- Empirical examples of fiscal risk materialization:
  - Mexico in mid 1990s: government eventually assumed about US$7.7 billion in debt after taking over private toll road concessions.
  - Hungary in late 1990s: M1 highway traffic forecasts proved too optimistic and the government assumed traffic risk.
  - Portugal in early 2000s: highways built using PPPs were reclassified as central government liabilities after the 2009 crisis, adding substantially to deficit and debt.
  - Spain in early 2000s: local airports and railways built by local governments using PPPs faced traffic shortfalls, closures, and persistent long-term commitments financed by transfers.

### Simple PPP versus government procurement model (cash-flow and NPV illustration)
- Government procurement scenario assumptions:
  - Government spends $100 to construct the road in Year 1 and $3 each year during Year 2 to 15 to maintain the road.
  - Discount rate of 5 percent.
  - NPV of total government commitment in Year 0 is $123.5.
- Government-funded PPP scenario assumptions:
  - Private company spends $100 to construct in Year 1 and $3 each year during Year 2 to 15.
  - Government compensates the company with an annual fee from Year 3 to Year 15.
  - If there is no efficiency difference, no company profit, no borrowing costs, NPVs of government commitment using both methods are exactly the same: $123.5.
  - Under these assumptions, the annual fee paid by the government to the company will be $14.5 per year.
- Additional model insights:
  - If the road charges user fees, NPVs remain the same because user fees reduce the government procurement NPV and reduce the PPP annual fee by the same amount.
  - If borrowing is required, NPVs remain the same assuming identical interest rates for government and company; both NPVs increase by the interest costs.
  - Efficiency gains are possible if the private company is more efficient, but can be offset by company profits, higher private interest costs, and PPP transaction costs.
  - Cash-based budgeting creates a bias in favor of PPPs; accrual-based accounting (e.g., IPSAS 32) is needed to capture true fiscal implications.

### Concluding remarks and policy implications
- There remains substantial room to strengthen China’s PPP regulatory framework.
- The rise in local government debt through LGFVs—essentially unregulated PPPs—highlights major regulatory weaknesses.
- The government’s promotion of a third generation of PPPs and related regulatory steps are welcome but key issues persist:
  - Vigilance against government bias in favor of PPPs.
  - Ensuring a level playing field for SOEs and truly private companies competing for PPPs.
  - Improving coordination within the government, particularly between the MOF and the NDRC.
  - Consolidation of framework regulatory documents to remove confusion and uncertainties.
- The authors propose a four-pillar regulatory framework promoting:
  - Good project selection.
  - Good institutional framework.
  - Good laws.
  - Good accounting and reporting practices.
- The proposed three-step strategy aims to balance promoting infrastructure and containing fiscal risks.

*Source: _wp16187 - 32. As more and more projects are disclosed in this way, a partial government balance sheet*

### Appendix 3. International Public Sector Accounting Standard 32

### Appendix 3. International Public Sector Accounting Standard 32

### Overview
- Formal name: “Service Concession Arrangements: Grantor”.
- Issued by the International Public Sector Accounting Standards Board (IPSASB) in October 2011.
- Compatible with IFRIC 12, “Service Concession Arrangements” (operator standard).
- Both standards are accrual based.
- Scope: covers both government-funded and user-funded PPP contracts as defined in Appendix 2.

### Definition of a service concession arrangement (PPP contract)
- Binding arrangement between a grantor (the government) and an operator (private sector contractor) where:
  - (a) The operator uses the service concession asset to provide a public service on behalf of the grantor for a specified period of time; and
  - (b) The operator is compensated for its services over the period of the service concession arrangement.
- Covers both government-funded and user-funded PPP contracts.

### Recognition criteria (when assets and liabilities are recorded on the grantor’s balance sheet)
- IPSAS 32 requires assets of a PPP and corresponding liabilities be recorded on the grantor (government)’s balance sheet if both conditions are met:
  - (a) The grantor controls or regulates what services the operator must provide with the asset, to whom it must provide them, and at what price; and
  - (b) The grantor controls—through ownership, beneficial entitlement or otherwise—any significant residual interest in the asset at the end of the term of the arrangement.
- For a whole-of-life asset (asset used in a service concession arrangement for its entire useful life), only the conditions in paragraph (a) need to be met.

### Accounting treatment: government-funded projects (financial liability model)
- Initial recognition:
  - The grantor records the same amount of asset and liability at the fair value.
- Subsequent accounting:
  - Expenses recorded: asset depreciation (consumption of fixed capital), finance charge (interests), charges for services paid to the operator.
  - Financing treatment: repayment of principal is accounted as reduction in liability.

### Accounting treatment: user-funded projects (grant of a right to the operator model)
- Initial recognition:
  - The grantor records the same amount of asset and liability at the fair value.
- Subsequent accounting:
  - Expense recorded: asset depreciation only.
  - Revenue: the grantor records the revenue accrued during the contract period (imputed revenue).
  - Financing treatment: reduction in liability is accounted as equivalent to the accrued revenue in each year.
- Cash flow implication: there are no cash flows in user-funded PPPs for the grantor while government-funded PPPs require cash payments by the government to the private partner.

### Mixed funding (combination of government and users)
- IPSAS requires the project to be divided into a government-funded part and a user-funded part.
- Apply the accounting rules for each part respectively.

### Illustrative example (IPSAS 32 BOT road example)
- Contract: build a road through a build-operate-transfer (BOT) contract lasting for 10 years.
  - Construction: first 2 years.
  - Operation by private-sector company: remaining 8 years.
- Construction costs:
  - Road base cost: $940.
  - Road surface cost: $110.
- Usable life:
  - Base: 25 years.
  - Surface: 6 years (implying surface will be re-constructed in year 8).
- Two financing approaches in example:
  - (1) Government-funded: the government pays the operator $200 per year during years 3-10.
  - (2) User-funded: the government allows the operator to charge users $200 per year during years 3-10.
- Illustration outcomes (as described):
  - Government nonfinancial assets recorded identical in both user-funded and government-funded PPPs.
  - Government liabilities broadly similar in both cases; difference arises from different treatment of interest and amortization. Government liability in user-funded PPP is only slightly lower because it does not incur interest cost.
  - Government cash flows: no cash flows in user-funded PPP; government-funded PPP requires cash payments by the government to the private partner.
  - Government net lending/borrowing (accrual surplus/deficit): higher in user-funded PPP due to the imputation of accrued revenue (revenue is imputed as if the government collects the user fees directly and uses these fees to repay the private partner).

### Implications for fiscal reporting and policy
- IPSAS 32 treats PPPs similarly to traditional government procurement, which reduces government bias in favor of PPPs.
- Assets and liabilities for both government-funded and user-funded projects are recorded on government balance sheet because the government is ultimately responsible for the public services, even in user-funded projects.
- Practical rationale: when many user-funded projects encounter difficulties, governments often take over those projects and consolidate their assets and liabilities onto the government balance sheet.
- Effect: IPSAS 32 places PPPs and government procurement on an equal basis so that the decision to choose between the two will be mainly based on efficiency gains.

*Source: IPSAS 32: Service Concession Arrangements: Grantor (IPSASB), as presented in Appendix 3 of the source PDF.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16187.pdf_