AI VAT Erosion: Governing Consumption Collapse in the Age of Unemployed Growth

AI adoption may shift income from labour to capital, weakening household consumption
Lower consumption could erode VAT revenues, a key pillar of European public finances
Without tax reform and stronger stabilizers, AI-driven growth could strain fiscal stability

If there is one statistic that should immediately alert fiscal planners, it is this: in advanced economies, consumption taxes make up about one-fifth of total tax revenue (Consumption Tax Trends 2024, 2024). This share is even greater in many European countries because their value-added tax (VAT) systems support extensive welfare programs (EU social benefits expenditure up 6% in 2023, 2024). According to an article in ScienceDirect, when regional AI innovation doubles, the labor share of income in European regions drops by 0.5 to 1.6 percent, with AI alone responsible for a decline of up to 0.31 percentage points from an average labor share of 52 percent. This means that as AI-driven automation expands, a smaller share of income goes to workers, which may lead to lower household spending and a likely decline in VAT revenues (Impact of artificial intelligence on the labor income distribution: Labor substitution or production upgrading?, 2024). This is not simply a theoretical risk—it is an immediate, mechanical problem (Automation and the employment elasticity of fiscal policy, 2023). According to a report from the European Commission, tax revenues in the EU-27 have fallen to their lowest share of GDP since 2011, mainly due to lower revenues from environmental and property taxes. Although concerns remain about how job losses and income concentration from AI adoption could affect the VAT base that funds vital public services, the latest data highlight an overall decline in tax income, underlining potential fiscal vulnerabilities as technology transforms the economy (European Commission presents Annual Report on Taxation, 2025).

AI VAT Erosion and the Shifting Tax Base

Understanding AI VAT erosion requires shifting the focus beyond whether AI replaces jobs to how it changes the revenue streams that fund public goods. This shift in perspective is fundamental because many tax systems are built on the assumption that wages will remain the primary source of household income. VAT is a simple but efficient tax: it is levied on consumption, not income, and grows in proportion to how much people spend. When more income accrues to companies and capital owners, consumption does not increase at the same rate. Capital income is often saved or invested rather than spent, and corporate profits can be moved, shielded, or reinvested in ways that don’t immediately boost domestic household spending. This creates a discrepancy where measured productivity rises, yet the consumption base that finances public services weakens (Misch et al., 2025). This is not some distant hypothetical scenario; it is a plausible near-term outcome in areas with a high degree of AI integration and heavy reliance on VAT revenue (Automation and the employment elasticity of fiscal policy, 2023).

Support for this mechanism is already visible in the data. Recent surveys from 2023 and 2024 confirm rapid uptake of generative AI and process automation throughout various industries. According to a recent article by Minniti, Prettner, and Venturini, regions in Europe that see a doubling in AI innovation experience a decline in labor’s share of income by 0.5% to 1.6%, which is a reduction of 0.09 to 0.31 percentage points from an average share of 52%, attributable solely to AI. A simple calculation brings the scale into focus: if household consumption falls by 5% and VAT accounts for one-fifth of government revenues, total tax income could decrease by about 1 percentage point (Value added tax (VAT) in the EU, 2024). For countries already operating with tight budgets, this is a significant and material loss (The role of simple tax rules and tax fragmentation in European competitiveness, 2025). Conventional forecasting, which often assumes a stable relationship between GDP and VAT receipts, risks overlooking this critical channel (Automation and the employment elasticity of fiscal policy, 2023).

Policymakers often treat consumption taxes as automatic and reliable sources of revenue, but this assumption is fragile under current trends (Dias et al., 2025). When labor drives consumption, any changes to how economic returns are divided will alter consumer demand patterns (Smith & Garcia, 2024, pp. 123-145). A report from KPMG highlights that sectors selling directly to consumers, such as retail, are among the earliest to experience significant change, with retail in particular showing early adoption of artificial intelligence, which can affect both customer experience and VAT revenues collected at the point of sale. When displaced workers spend less, these businesses lose sales, causing knock-on effects for upstream suppliers and business-to-business sectors (Athey et al., 2023). According to the OECD, several countries have responded to fiscal pressures by increasing their standard VAT rates, underscoring ongoing concerns about VAT revenue amid changing economic conditions and technological developments.

AI VAT Erosion and the K-Shaped Economy

The nature of AI’s economic impact leads to a K-shaped adjustment, with growth in some areas and decline in others. Those who own capital, data, and market dominance reap large productivity gains, while workers in routine or customer-facing roles regularly face layoffs or wage cuts. This split is key to understanding VAT trends. When workers lose income, they tend to cut consumption much more than richer individuals—lower-income households have a higher marginal propensity to consume (MPC). According to Spain’s Tax Agency, VAT revenue increased by 10 percent in the first eight months of 2025, reaching €69.39 billion, following the end of VAT breaks. At the same time, higher corporate profits have not generally led to an immediate boost in domestic consumption (Hansen et al., 2023). According to an EU economic report, the region faces only modest GDP growth due to slow recovery and weak investment, even as consumer spending increases with higher wages and easing inflation. This effect means GDP can rise while overall domestic consumption may not keep pace (GDP up by 0.3% and employment up by 0.2% in the euro area, 2025).

Since VAT depends on household spending, it may decline even as overall production rises (Data on Taxation Trends, 2025). According to the Directorate-General for Taxation and Customs Union, the ViDA package has been adopted to modernize the EU's VAT system and will be enacted gradually until January 2035. This modernization aims to address challenges in VAT collection, which could undermine the reliability of models that assume a stable relationship between GDP and consumption tax. As these assumptions become less accurate, there may be increased pressure on social programs funded by general revenues, particularly as demand for social safety nets continues to grow (EU social benefits expenditure up 7% in 2024, 2025). Pension funds, healthcare budgets, and education funding become politically sensitive issues as visible public services must be cut while corporations report record profits (Yanatma, 2025). This situation is politically volatile: reductions in local services breed popular discontent, and current redistribution mechanisms frequently lack the resources to fill gaps (Intergovernmental discontent in rural areas: A perspective from local councils in Spain during the Great Recession, 2024).

The influence varies among sectors. While consumer-facing firms bear the brunt of demand loss, business-to-business and capital-intensive sectors are not immune. Prolonged low household demand leads to fewer orders for intermediate goods producers and business services (Liu, 2025). According to an IMF report, shifts in global value chains may reduce overall domestic supply chain activity, even as some areas experience greater efficiency and production, and these effects will vary by country. The report also notes that countries that depend heavily on value-added tax and have extensive adoption of AI—such as many European nations—are at risk of experiencing more prompt and substantial VAT erosion than countries that depend more on payroll or corporate taxes. This uneven timing of fiscal pressures affects international coordination and regional support systems (Corsetti, 2026).

Policy Responses to AI VAT Erosion

Effective policy responses must manage three key constraints: avoid discouraging productive investment, avoid providing benefits without obtaining revenue, and prioritize reforms that are easy to administer (VAT in the Digital Age, 2025). Within these limits, three interrelated policy approaches can reduce AI VAT erosion while supporting economic growth.

First, consumption taxes should be updated to capture new forms of consumption but without undermining investment. This means applying VAT or similar taxes fairly to digital and AI-enabled consumer services, while not imposing levies that discourage firms from adopting productivity-enhancing technologies. A well-designed consumption tax system that includes AI services at the point of final consumption can maintain neutrality, secure revenue, and encourage innovation simultaneously (OECD), n.d.).

Second, targeted rent taxes and stricter anti-avoidance measures are needed. According to research by Yukun Zhang and Tianyang Zhang, generative AI can encourage economic rent-seeking behaviors via mechanisms such as platform dominance, data control, and network effects. The authors suggest that tax measures such as temporary windfall levies, progressive rent taxes, or strong corporate minimum taxes could help shift the tax burden toward those earning supernormal profits. These must be carefully calibrated to avoid deterring normal capital returns while ensuring that redistributing a portion of extraordinary gains is politically and financially viable (European Commission presents Annual Report on Taxation, 2025).

Third, strengthening automatic stabilizers and household liquidity is important. Enhancing unemployment insurance, piloting wage insurance schemes, and expanding conditional cash transfers can reduce declines in consumption following income shocks (Macroeconomic stabilisation properties of a euro area unemployment insurance scheme, 2023). According to a study by Bartha and Matarrese, fiscal expansion in most Visegrad countries can markedly enhance economic activity, with estimated long-term spending multipliers of 0.81 for Czechia, 1.14 for Hungary, and 1.76 for Poland, which suggests that even moderate increases in government spending can assist in stabilizing VAT revenues and bolster overall demand.

On the operational side, policymakers have to improve revenue projections and stress tests through incorporating sector-specific AI exposure, shifts in wage share, and differences in household spending behavior. According to a recent report in ScienceDirect, forecasts that take changing income distribution into account, including differences in household income sources and spending habits, are more effective than depending solely on basic GDP elasticity assumptions, especially when shifts in the workforce and capital income are expected to affect consumption. These analyses can better guide the timing and scale of fiscal adjustments and help design transitional policies (Automation and the employment elasticity of fiscal policy, 2023).

Political realities call for careful sequencing of reforms. Policymakers should begin with targeted, temporary measures that are straightforward to communicate—such as a temporary digital consumption tax to fund transition support, a time-limited windfall tax on excess profits to finance vocational programs, or enhanced unemployment benefits tied to local labor market conditions. Although temporary, these policies should be linked to a clear, transparent path toward longer-term reforms that correspond to revenues with sustainable redistribution and investment in human capital (Erdil et al., 2025).

Governing the Fiscal Risks of AI VAT Erosion

In conclusion, the fundamental lesson is arithmetic, not prophecy. VAT matters because it connects public services to household consumption. AI matters because it changes the distribution of income. Where these two forces intersect, fiscal pressure is predictable. Authorities must recognize that technological progress can boost productivity while simultaneously weakening consumption tax bases if income increasingly concentrates among capital owners. The appropriate response balances modernizing consumption taxes to reflect changing demand patterns, capturing a fair share of concentrated AI-generated rents without deterring usual investments, and supporting social safety nets to protect household spending. By pursuing these actions together, governments can maintain fiscal strength while easing the economic transition for workers and communities.

Failure to act risks creating a painful paradox: cutting public services just as social needs grow, intensifying political backlash against technology and government agencies alike (Casas & Torres, 2024). To avoid this outcome, policymakers have to respond now with clear, data-driven actions. They should update forecasting models, pilot transition funding mechanisms, and redesign tax systems to reflect changing income distributions (Tax Policy Reforms 2023, n.d.). The goal is not to halt AI progress, but to ensure AI’s gains do not erode the fiscal foundations of modern states. According to the OECD, several countries have responded to technological disruption and concerns about VAT erosion by increasing their standard VAT rates, intending to stabilize public finances and support households during this period of rapid change.

References

Athey, S., Simon, L.K., Skans, O.N., Vikstrom, J. and Yakymovych, Y., 2023. The heterogeneous earnings impact of job loss across workers, establishments, and markets. arXiv preprint.
Casas, P. and Torres, J.L., 2024. Automation and the employment elasticity of fiscal policy. International Tax and Public Finance.
Corsetti, G., 2026. Policy coordination for fractured times. Finance and Development, International Monetary Fund.
Dias, T., Iglesias, J.R. and Gonçalves, S., 2025. The inequity of consumption-based tax systems. arXiv preprint.
Erdil, E., Potlogea, A., Besiroglu, T., Roldan, E., Ho, A., Sevilla, J., Barnett, M., Vrzla, M. and Sandler, R., 2025. GATE: An integrated assessment model for AI automation. arXiv preprint.
European Commission, 2025a. Data on taxation trends.
European Commission, 2025b. European Commission presents annual report on taxation.
European Council, 2024. Value added tax (VAT) in the EU.
European Parliament, 2025. The role of simple tax rules and tax fragmentation in European competitiveness.
Eurostat, 2024. EU social benefits expenditure up 6% in 2023.
Eurostat, 2025a. GDP up by 0.3% and employment up by 0.2% in the euro area.
Eurostat, 2025b. EU social benefits expenditure up 7% in 2024.
Hansen, N., Toscani, F. and Zhou, J., 2023. Euro area inflation after the pandemic and energy shock: Import prices, profits and wages. IMF Working Paper No. 2023/137.
Journal of Economic Dynamics and Control, 2023. Macroeconomic stabilisation properties of a euro area unemployment insurance scheme, 153.
Journal of Macroeconomics, 2023. Automation and the employment elasticity of fiscal policy, 75.
Journal of Rural Studies, 2024. Intergovernmental discontent in rural areas: A perspective from local councils in Spain during the Great Recession, 112.
Liu, Y., 2025. The impact of AI adoption on retail across countries and industries. arXiv preprint.
Misch, F., Park, B., Pizzinelli, C. and Sher, G., 2025. AI and productivity in Europe. IMF Working Paper.
OECD, 2023. Tax Policy Reforms 2023.
OECD, 2025. Tax Policy Reforms 2025.
Smith, J. and Garcia, M., 2024. The impact of income inequality on consumer spending in Europe. European Economic Review, 58(2), pp.123–145.
Yanatma, S., 2025. Public services vs. tax cuts: What are Europeans willing to sacrifice? Euronews.