From a distance, the picture looks beautiful: public debt at ~34% of GDP - among the lowest in the region - growth of +8.02% (2025), disciplined deficits, and borrowing mostly in local currency. But look closer, layer by layer, and the picture gains more shading: the tax-to-GDP ratio of 11.4% of GDP is just a third of the OECD benchmark, 2025 public investment disbursement reached only ~83.7% despite unusually high political pressure to spend, the social insurance fund has been warned it could face sustainability pressure sooner than the population itself ages, and pieces of the iceberg that sit below the waterline keep getting flagged by the IMF in its Article IV reports. This piece lays four layers - revenue, spending structure, public debt, and structural themes - side by side so the fuller picture can emerge.
Scope: An overview of Vietnam's fiscal architecture across the layers of revenue, spending, public debt & decentralization, and structural themes. Figures are drawn from the Ministry of Finance, the General Statistics Office, the IMF's 2024 Article IV, the World Bank, Fulbright Vietnam, and international credit rating agencies.
Note: The official public-debt figure of 34.7% of GDP reflects the scope defined under the Public Debt Management Law. The contingent liabilities - implicit obligations from SOEs, BOT projects with implicit guarantees, and bank recapitalization costs - sit outside the published number, and the IMF has recommended greater transparency around them.
The Big Picture - Six Numbers That Say a Lot
The ceiling is 60%, and the current level is ~34%. That number says there's plenty of room left to borrow. It does not say that revenue collection, spending, and governance capacity are enough to turn that headroom into actual development - that's a story that has to be read in the layers that follow.
Compared with peers at the same income level, Vietnam has a genuinely enviable set of macro indicators: low debt, high growth (GDP per capita hit $5,026 in 2025 - officially crossing into the upper-middle-income bracket), disciplined deficits, adequate reserves, and borrowing in local currency. This is the real outcome of years of budget discipline combined with rapid nominal GDP growth.
But those same numbers reveal a different picture: the (narrow) tax-to-GDP ratio of 11.4% - only a third of the OECD level, two-thirds of the ASEAN level; the informal sector at 25–30% of GDP (Fulbright's estimate) still sits outside the tax system; and ~57% of the labor force works informally. This structure creates a static constraint: the budget has room to borrow, but doesn't have room to collect enough revenue to fund long-term reforms without borrowing. Those two things are not the same.
The Revenue Engine - Three Main Pillars, One Narrow Base
Targeted 2026 total state budget revenue: ~VND 2,530 trillion (~$96B), around 18–20% of GDP on a broad basis (including fees and non-tax revenue). The three main tax pillars - VAT, CIT, PIT - show the classic profile of a developing economy: indirect taxes make up >50%, direct taxes stay thin.
Accounts for ~43% of total tax revenue. The standard rate is 10%, temporarily cut to 8% through 12/31/2026 as a stimulus measure. Cost to the budget: ~VND 121.7 trillion - a temporary measure now in its fifth straight year.
Standard rate 20%. Since 10/2025, tiered: 15% / 17% / 20% by revenue size. But: the effective rate for the FDI sector after incentives is only 5–7%, and lower still for firms with maximum incentives.
The new law takes effect 07/01/2026: brackets cut from 7 to 5 (5–35%). The personal/family deduction rises to VND 15.5 million/month. A flat 5% digital-services tax applies. The PIT base stays narrow because most informal workers remain outside the system.
Vietnam collects only 11.4% of GDP in (narrow) tax revenue - versus an ASEAN average of 17%, an OECD average of 34%. Why? Fulbright and the IMF point to three mutually reinforcing reasons: (1) the informal sector is simply too large (25–30% of GDP, ~57% of the workforce), (2) FDI incentives push the effective rate down to 5–7%, and (3) digital tax-collection capacity is still limited. The 2025–2026 tax reform is a step toward gradually closing these three gaps - but whether the pace of closing them keeps up with the pace at which spending is expanding is a story worth watching.
Spending Structure - When Recurrent Spending Eats Into Investment Headroom
Targeted 2026 total state budget expenditure: ~VND 3,159 trillion (~$120B, up 23% YoY). The spending structure is being pulled by two opposing forces: public investment is prioritized to rise sharply on paper, while recurrent spending - especially after the 07/2024 base-salary adjustment - is expanding even faster.
Recurrent spending in Q1/2025: VND 316,500 billion (20.2% of plan, up 16.8% YoY). Investment spending over the same period: VND 78,700 billion (just 10.0% of plan, down 2.5% YoY). When two line items in the same budget run in opposite directions - one outpacing plan, the other lagging it - the 65/29 split isn't an equilibrium; it's a ratio sliding further out of balance.
The constitutional target is 20% of the state budget; the actual figure is ~15.5% (~3–4% of GDP). This multi-point gap has persisted for years. The OECD average is ~5% of GDP.
2026 plan: ~VND 995 trillion (~$38B). Focused on the North-South expressway and high-speed rail. The plan is ambitious - execution is a separate story (see Theme 01).
~6.8% of recurrent spending. Health insurance covers >93% of the population. Most spending flows through health insurance, with the budget covering the shortfall - a pressure that grows as the population ages.
~12.3% of state budget spending (~1.7–2% of GDP). Growing by an average of ~8.5%/year - faster than GDP growth. Rose substantially over 2003–2018.
Public Debt - Attractive by Comparison, Fragile by Definition
This is where Vietnam genuinely stands out: lower than almost the entire region. But this is also exactly where the IMF and credit rating agencies keep making the same point: the ~34% figure (end-2025) only covers public debt under a narrow definition - the contingent liabilities sit outside it.
Vietnam has shifted decisively toward domestic VND-denominated borrowing - which made up the bulk of new borrowing in 2024 - helping it avoid the kind of currency risk that hit Argentina, Turkey, and Sri Lanka. This is one of the most important debt-structure reforms since Doi Moi.
The IMF's Article IV reports keep flagging the same point: the 34.7%-of-GDP figure excludes a meaningful chunk: (1) unguaranteed SOE debt; (2) the cost of recapitalizing state-owned commercial banks; (3) implicit guarantee obligations tied to BOT projects; (4) roughly 30% of power generation capacity (~13 GW) built under government-guaranteed BOT contracts; and (5) off-budget funds that aren't fully controlled. The iceberg has a visible part and a submerged part - the regional league table only looks at the visible part.
The chart above reveals a paradox: Vietnam's debt sits at 34.7% of GDP - the lowest in the group - yet it's the only major Southeast Asian economy that hasn't reached investment grade. Indonesia carries 39% debt and is rated BBB. The Philippines carries 61% and is still BBB+. Why? Because S&P, Moody's, and Fitch don't just look at how much debt there is - they look at the system surrounding that debt.
- Average income is still low relative to peers. GDP per capita hit $5,026 in 2025 - just crossing into upper-middle-income territory, but still below Thailand (~$7,300) and Malaysia (~$12,400). Rating agencies treat income as a proxy for shock-absorption capacity: lower-income economies have thinner buffers when a crisis hits.
- Fiscal transparency doesn't yet meet the standard. Public finance statistics don't fully comply with GFS 2014, data is published with a longer lag than peers, and - as noted above - the contingent liabilities sit outside the official figures. When agencies can't measure hidden risk, they respond by keeping the rating lower.
- Institutional quality. The World Bank's Worldwide Governance Indicators (WGI) show Vietnam scoring notably lower than Indonesia and the Philippines on two pillars agencies weight heavily: regulatory quality (~38th vs ~55th percentile) and voice & accountability (~15th vs ~45th–50th). This is also the hardest factor to improve quickly.
- Shallow capital markets. Government bonds are heavily concentrated in the hands of the social insurance fund and banks (68–91%), with little foreign or diverse institutional participation. A lack of depth means a lack of price discovery - agencies read this as a sign of an immature market.
- Monetary policy independence is not fully established. Bank Indonesia shifted to inflation targeting in 2005, and the Philippines' BSP did so in 2002. The State Bank of Vietnam still operates with multiple simultaneous objectives (the exchange rate, credit growth, system stability) - a framework agencies view as less predictable.
In other words: low debt is a necessary condition, not a sufficient one. Investment grade requires the whole ecosystem - transparency, institutions, capital markets, monetary policy - to clear a threshold. Vietnam is improving on each pillar, but the gap to BBB remains a systemic gap, not a numerical one.
The Great Tax Reform of 2025–2026
In 2025, the National Assembly passed new laws covering nearly every major tax: CIT, PIT, VAT, special consumption tax, and the Tax Administration Law. It's the most comprehensive reform since Doi Moi - and, indirectly, an admission that the current tax base is too narrow for the fiscal needs of a country aiming for high-income status by 2045.
Tiered rates of 15% / 17% / 20%. Adds incentives for R&D, innovation, and digital transformation.
Cuts brackets from 7 to 5. The personal/family deduction rises to VND 15.5 million/month. Adds a 5% digital services tax.
A 15% QDMTT applies to MNCs with revenue ≥ EUR 750M. The OECD recognized it as "transitional qualified" starting 08/2025.
Comprehensive digitalization. Links data across tax, customs, and banking. Strengthens tools for taxing e-commerce and digital platforms.
COVID-19 - A Fiscal Stress Test, and a "Temporary" Measure That Wouldn't End
COVID-19 was the first large-scale stress test of Vietnam's fiscal capacity since Doi Moi. Thanks to low debt headroom and years of discipline, the government had enough resources to roll out stimulus packages without having to borrow from the IMF. That's a result of discipline - not luck.
The 2-point VAT cut was designed as a temporary measure - but it has now run for five straight years through the end of 2026. It's a textbook case of "temporary becomes permanent" in fiscal policy: a measure enacted for a crisis becomes hard to unwind, because the political cost of "raising taxes back" always outweighs the fiscal cost of "keeping taxes low for one more year." Each year of extension costs the budget ~VND 121.7 trillion.
Ten Structural Themes - The Submerged Part of the Iceberg
Beneath the surface of these attractive headline numbers, the IMF, World Bank, Fulbright, and credit rating agencies keep coming back to a set of structural themes. Not all of them are bad - reform programs are already underway for many. But they share one trait: none of them operates in isolation - these themes connect and reinforce each other, and together they create a gap between "fiscal headroom on paper" and "actual fiscal capacity."
01 Public Investment Disbursement Bottlenecks The money, the plan, and the directives are all there - but the money moves slowly A Chronic, Multi-Year Problem
It's a strange but real mechanism: the government issues bonds (which carry interest) to fund public investment. The capital gets allocated, but the money can't get out the door to projects on schedule. The result: the budget is paying interest on money sitting idle in the treasury. It's one of the quietest and most persistent forms of fiscal waste - in 2025 the rate jumped to 83.7% under unusually direct pressure from the Prime Minister, but it still fell short of the 100% target and still left many projects without disbursement for the year.
The country's economic engine was allocated more than VND 79,200 billion. By mid-to-late 2024, disbursement was running well below plan. 63 of 115 ministries, agencies, and localities disbursed below the national average; 12 units disbursed less than 20%. The cause can't be blamed on a lack of funding or a lack of political will - both were already there.
The contributing factors have already been identified: slow land clearance · volatile construction material prices · late-issued implementing guidance · the 2025 administrative merger disrupting project management boards · unrealistic planning · excessive caution in the approval process (see Theme 10). These factors have been listed in resolution after resolution for years - the question is no longer "why," but "why hasn't it been fixed yet."
02 The Informal Sector & Transfer Pricing A quarter to a third of the economy sits outside the tax system Narrow Tax Base
03 SOE Efficiency - Still a Meaningful Gap 29% of GDP, 40% of investment capital - but efficiency and bad debt tell a different story Capital Allocation
The SOE sector accounts for ~29% of GDP and ~40% of total social investment capital, yet generates only ~30% of GDP growth, and SOEs are linked to ~60% of banking-sector bad debt. The input-to-output ratio shows an efficiency gap relative to the FDI and private sectors - something research from the Ministry of Planning & Investment, the Ministry of Finance, and the World Bank has pointed out for years.
The Ministry of Industry and Trade's Twelve "Trillion-Dong" Projects
04 Dependence on Land Revenue - A Wave That Drags the Whole Local Budget With It When real estate cools, provincial and city budgets cool along with it An Unsustainable Revenue Source
Land-related taxes and fees made up ~15% of total state budget revenue in 2024. For many localities, land-use fees are the single largest revenue source. That creates a dangerous correlation: when the real-estate cycle turns, local budgets turn with it - and this is the exact budget line that funds local infrastructure.
The spiral that needs breaking: real estate cools → land auctions fail → local budgets fall short → local infrastructure slips behind schedule → the regional economy slows → real estate cools further. The new laws (Land Law 2024, Housing Law 2023, Real Estate Business Law 2023), effective 2024–25, have tried to cut this spiral from the legal side, but the real-world impact still needs time to be confirmed.
05 FDI Incentives - When the Effective Rate Is Just a Quarter of the Nominal Rate 70% of export value, but only ~7–8% of state budget revenue via CIT A Disproportionate Contribution
Samsung - Vietnam's largest FDI enterprise - is estimated to owe up to $6.5B in additional tax over its full incentive period. If Vietnam hadn't proactively applied the 15% QDMTT, that money would have gone to South Korea instead. Applying the QDMTT was the right move - and it's also an indirect admission that the earlier incentives had gone deep enough to effectively become another country's tax revenue.
The FDI paradox in the fiscal balance sheet: FDI generates ~70% of Vietnam's export value and pays 39–41% of total CIT revenue. But CIT only makes up 18–21% of the total state budget, so the multiplication works out this way: the entire FDI sector - the backbone of exports - contributes only about 7–8% of total state budget revenue via CIT. The question isn't "should FDI be given incentives" - it's "how deep have those incentives gone, and is it time to redesign them."
06 Sustainability Pressure on the Social Insurance Fund Warned to run dry sooner than the population ages · 38% coverage · 4 million lump-sum withdrawals Demographics
Over 2016–2021: more than 4 million people took a lump-sum withdrawal - permanently leaving the pension system. Only 19% of surveyed workers said they would keep contributing if they became unemployed. The 2024 Social Insurance Law tightened lump-sum withdrawal conditions, but the pivotal event has already happened: the contribution base has already been eroded in the middle of the pyramid.
Most of the social insurance fund's assets are invested in government bonds - meaning the pension fund is effectively lending to the very budget it depends on, at a negative real interest rate. That does two things at once: (1) it creates the illusion of cheap borrowing costs for the state budget, and (2) it gradually erodes the pension fund's long-term returns. As the fund edges toward depletion, the budget will have to both service the bond debt and cover the pension shortfall - the same dong being asked to do two jobs.
07 Contingent Debt & BOT - The Submerged Part of the Iceberg 34.7% of GDP is only the defined, published portion Contingent Liabilities
In its Article IV reports, the IMF has repeatedly noted: "contingent and guaranteed liabilities are not fully disclosed." The official public-debt figure of 34.7% of GDP follows the Public Debt Management Law - but that definitional framework is narrower than the GFS/IMF standard, and so it excludes several significant items.
There is no official public estimate of the total for these items - and that's precisely the point the IMF recommends improving in order to move up to investment grade.
08 Recurrent Spending Eating Into Investment Headroom The IMF warns of crowding out · Q1/2025 ran opposite to plan Spending Structure
The +30% base-salary adjustment (07/2024) was a necessary response to the wave of ~40,000 public employees who quit between 2020 and mid-2022, mostly in education and healthcare (where pay lagged the private sector significantly). But the measure also carries fiscal consequences the IMF flagged from the start.
09 The Bond Market - Concentrated Demand, No Foreign Investors Social insurance + banks hold 68–91% of government bonds · negative real yields · the 2022 corporate-bond crisis Capital Markets
10 When Execution Quality Becomes the Bottleneck Necessary discipline + a side effect: approval hesitancy Public Governance
Institutional discipline is a necessary condition for any reform. But a public-governance principle that many countries have already experienced applies here: when penalties for execution mistakes are heavy while protections for officials who act correctly are not correspondingly robust, behavior that is individually rational can produce outcomes that are undesirable at the system level.
Recent Central Committee resolutions have already called for a mechanism to protect officials who dare to think, act, and take responsibility for the common good. Turning this principle into concrete, applicable regulations with clear protections is the key link that determines the outcome of many of the other themes in this piece (from disbursement to SOEs to the implementation of the new tax laws).
The Feedback Map - When the Themes Amplify Each Other
The ten themes above aren't independent. They connect, amplify each other, and together create the gap between the picture on paper and actual fiscal capacity. Five main feedback loops:
Fiscal Health Dashboard
Putting it all together: the picture has genuinely bright spots, areas to watch, and areas needing medium-term reform. Laying the three layers side by side makes it clearer:
The Takeaway - A Kind of "Execution Ceiling"
Vietnam isn't facing a debt-crisis risk of the kind seen in Greece, Sri Lanka, or Argentina - debt is low, reserves are adequate, and borrowing is mostly in local currency. That's a genuine strength, the result of years of discipline. This piece isn't arguing against that point.
But a different kind of risk is emerging - harder to see, harder to fix, and far less written about: a kind of "execution ceiling." The room to borrow exists (25 percentage points below the ceiling), but tax-collection capacity, the ability to spend effectively, and the quality of public governance are three bottlenecks that make it hard to convert that headroom into real development. Collecting only 11.4% of GDP in tax revenue isn't enough to self-fund long-term reforms. Disbursing at 73% means money moves more slowly than expected. Pressure on the social insurance fund means the future budget will have to cover the shortfall. A decision-making environment still marked by hesitancy means every well-designed law slows down once it reaches the execution layer.
The high-income-by-2045 goal isn't blocked by a lack of money or a lack of vision. Both of those exist. It's being challenged by the quality of execution at the middle layer - and the window to solve that problem is closing fast, in step with population aging and shifting global economic structure.
03 Discussion
Leave a note
A considered space for questions, counterpoints, and useful additions. Civil, on-topic, signed.
Reader notes
...Loading notes...