Illustration of balooning and disintigrating dollar bills
Economic Risks

The Impact of Deficits on Costs for Households

Abhi Gupta
March 9, 2026
Download the Data
Printable PDF

Key Takeaways

  1. Federal fiscal policy is a significant but often overlooked factor in the affordability of vehicles, homes, and small business loans. More debt leads to higher interest rates, making credit less affordable.

  2. Since 2015, the cumulative effects of fiscal policy, as measured by CBO’s estimated costs of enacted legislation, have raised 10-year-ahead projected federal debt by about 49 percentage points of GDP. As a result, we estimate long-term Treasury yields have risen about 97 basis points.

  3. For a family taking out a 30-year mortgage at the 2025Q3 median home price, this rise in long-term interest rates has raised borrowing costs by about $2,500 per year or roughly $76,000 over the life of the loan compared to a world without this additional federal debt.

  4. Annual borrowing costs on a typical auto loan are about $120 higher and on a typical small business loan about $770 higher than they would be absent these fiscal-policy changes.

  5. This post introduces a new Budget Lab tracker that cumulates the interest-rate effects of all enacted legislation across Congressional Budget Office (CBO) projection releases. This release is based on the February 2026 budget and economic forecasts, which incorporate tariff policies and enacted legislation as of late 2025. We plan to update these estimates each time CBO publishes a new set of projections.

Deficits and Affordability

Recently, public discussion has focused on whether households can afford the big investments — a car, a house, funds to grow a small business — that help unlock new economic opportunities.  One underdiscussed factor affecting the affordability of these things is the ongoing and rapid growth of federal debt. Research has shown that when Congress decides to persistently raise deficits (e.g., by cutting taxes or increasing spending without making offsetting adjustments), then interest rates rise for the overall economy (Gomez-Cram et al., 2025). That suggests that the consistent pattern of deficit spending that has been the norm in the United States for many years has driven up costs for families. In this piece and companion interactive tool, we estimate how much deficit-financed fiscal policy has increased loan costs for families and reduced the affordability of important goods and investments. We plan to track how fiscal decisions are affecting costs for families moving forward.

A Running Tally of Fiscal Policy’s Effect on Interest Rates

In March 2025, The Budget Lab (TBL) published a study documenting the inflationary risks and household cost effects of rising federal deficits and debt. That analysis found that deficit-financed spending raises interest rates through multiple channels, such as crowding out of private investment, higher term premia demanded by bondholders, and eventual monetary-policy tightening. In June, TBL applied that framework to the One Big Beautiful Bill Act, translating the bill’s projected interest rate effects into dollar costs for mortgages, auto loans, and small business loans.

This post introduces a complementary approach. Rather than scoring the effects of a single piece of legislation using a full macroeconomic simulation, we track the cumulative interest-rate impact of all enacted legislation over time using the Congressional Budget Office’s (CBO’s) own decomposition of changes in its budget projections. Each time CBO updates its baseline projection, it reports how much of the change in projected deficits is attributable to new legislation versus economic and technical revisions. By isolating the legislative component across 22 CBO projection vintages since August 2015, we can build a running account of how much enacted fiscal policy has changed the debt outlook and, by extension, contributed to higher borrowing costs. Figure 1 summarizes the headline household cost results. We also include the fiscal impacts of this administration’s tariff policy in our accounting. More detail on the methodology and a discussion of the differences between our previous and current approaches is available in the appendix.

The Impact of Rising Federal Debt on Household Borrowing Costs

To translate increased federal borrowing into rises in interest rates we use a standard rule-of-thumb approximation based on recent research and assume that for every 1 percentage point forecasted debt-to-GDP rises, interest rates on 10-year Treasury bonds will rise 2 basis points.1 These increases in Treasury yields then pass through to interest rates on household mortgages, auto loans, and small business loans. Table 1 below summarizes these estimated borrowing-cost effects for all enacted legislation since 2015 and since 2022, applying the cumulative impact to typical loan amounts and interest rate levels as of late 2025Q3. Figure 2 shows the running sum of interest rate effects since 2015 (and since 2022, alternatively).

Sensitivity to Assumptions about Interest Rate Response

The true responsiveness of interest rates to federal debt may be different than researchers have calculated. The table below shows how the since-2015 fiscal policy effects and resultant mortgage rate effects change if Treasury rates are more or less sensitive to rising federal debt than we assume in our baseline calculations. The range of annual mortgage cost effects from cumulative fiscal legislation since 2015 is approximately $1,900 to $3,750 per year.

Our chosen interest rate sensitivity is typical for this literature, though some have found larger effects. For example, Plante, Richter & Zubairy (2025) use revisions to CBO forecasts of medium-term debt/GDP and find a response of 3bps/pp. Other researchers have exploited plausibly exogenous variation in debt levels generated by close election results and other events. Ehrlich, Kay, and Thapar (2025) take such an approach and find, looking at parliamentary elections abroad, a much larger effect of long-term interest rates on government debt of about 6bps/pp, at about triple our baseline assumption.

Regardless of the estimated interest rate sensitivity, an important assumption we make is that each increment of additional federal debt has the same effect on interest rates, i.e., the relationship is linear. Deviations from that assumption could be important when considering such a large increase in federal debt as has been legislated since 2015.

Key Drivers of Rising Federal Debt

Figure 3 shows CBO’s projected debt-to-GDP has changed due to legislation for each published forecast since August of 2015. Each bar represents the approximate change in projected 10-year-out debt-to-GDP attributable to legislation enacted between one CBO projection and the next. Figures 4 and 5 translate these debt-to-GDP changes into basis-point effects on long-term Treasury yields, for the full since-2015 and shorter since-2022 windows respectively. In each figure, the upper panel shows the per-CBO-projection interest rate change and the lower panel shows the running cumulative total.

Major Pieces of Legislation Often Have Large Long-Run Impacts

The largest contributions to projected debt reflect the major legislative episodes of the past decade. The April 2018 CBO update, which captured the effects of the Tax Cuts and Jobs Act (enacted December 2017) and the Bipartisan Budget Act of 2018, added approximately 9 percentage points of projected debt-to-GDP and raised long-term Treasury yields roughly 19 basis points. The TCJA alone was scored by CBO and JCT at approximately $1.5 trillion in 10-year deficit costs.

The COVID-19 response legislation of 2020-2021 collectively represents the largest cumulative legislative contribution in the series. The September 2020 CBO update, which captured the CARES Act and other pandemic relief enacted through mid-2020, added about 9 percentage points of projected debt-to-GDP. The American Rescue Plan Act (March 2021) and the December 2020 Consolidated Appropriations Act together added another roughly 11 percentage points across subsequent forecasts.2

The Fiscal Responsibility Act of 2023 is the largest legislative reduction in the series, reflecting the spending caps negotiated as part of the June 2023 debt-ceiling agreement. That vintage reduced projected debt-to-GDP by about 7 percentage points, partially offsetting prior increases.

Fiscal Policy Over the Past Year Has Modestly Increased Forecasted Debt

The most recent data point (the February 2026 CBO projection, which was released before the Supreme Court invalidated the IEEPA tariffs and the new Section 122 tariffs were announced) shows a modest net increase in projected debt-to-GDP in this tracker, reflecting two large policy changes that move in opposite directions but do not fully offset. The result is a modest increase in projected debt-to-GDP. To the extent that IEEPA tariffs are not fully reconstituted under other legal authorities (and reimbursements are made for previously collected tariff revenues), deficits will be correspondingly higher than in our calculations. As such, our estimates of interest rate effects are likely to be somewhat conservative.

CBO's legislative changes category for this vintage shows an approximately $3.4 trillion increase in projected 10-year deficits, driven almost entirely by the 2025 reconciliation act (One Big Beautiful Bill Act), enacted in July 2025. This figure reflects the law's static budgetary effects on revenues and outlays (the direct scoring by CBO and the Joint Committee on Taxation) plus approximately $0.7 trillion in associated debt-service costs.

Working in the opposite direction, CBO projects a roughly $3 trillion increase in projected customs duty revenue from the administration's tariff actions through November 20, 2025. Because these tariffs were implemented through executive action rather than new legislation, CBO classifies the resulting revenue as a technical and economic rather than legislative change.3 Since the administration considers tariffs to be of similar importance as the tax and spending policies in the reconciliation bill, we classify the economic and technical customs revenue lines in the February 2026 forecast as “legislative changes” for our purposes and incorporate them in our tracker.4

Conclusion

Federal deficits, and the borrowing they necessitate, tend to raise the cost of private borrowing. As interest rates rise throughout the economy in response to increasing federal debt, households find it more expensive to finance their homes, vehicles, and small businesses. Determining how much more expensive is complicated by the fact that interest rates are constantly changing for reasons unrelated to federal debt.

To better understand this, the Budget Lab has calculated affordability impacts for households based on CBO’s scoring of legislated debt increases and research on the responsiveness of interest rates to federal borrowing. We find that the post-2015 increase in federal debt has led to almost a 1 percentage point increase in Treasury yields, which translates to an annual increase of $2,500 for the median home, $120 for the average auto loan, and $770 for the average small business loan.

Appendix

Comparison to Prior Analysis

TBL’s June 2025 analysis of the One Big Beautiful Bill Act used a different methodology, relying on full macroeconomic simulations (from FRB/US, MA/US, and other models) to estimate the interest-rate effects of a single piece of legislation over a 30-year horizon. Those models capture a wider range of channels than this tracker, including crowding out, deficit-inflation feedback, monetary-policy tightening, and general-equilibrium effects on the capital stock. That approach projected that, as enacted, the OBBBA alone would raise the 10-year Treasury yield by 1.4 percentage points by 2054.

This tracker takes a narrower but more frequently updatable approach. The Laubach (2009)-type framework captures primarily the fiscal risk premium channel (the effect of higher projected debt on Treasury term premia) and, to a lesser extent, expected future short rates. It does not account for the full general-equilibrium response of the economy to higher debt. As a result, this tracker’s estimates should be interpreted as an approximate lower bound on the total interest-rate impact of fiscal policy, particularly over longer horizons where inflation and monetary-policy feedback become quantitatively important.

The two approaches are complementary. This tracker provides a timely, regularly updated running total that can be published with each new CBO projection release. The full macro-simulation approach, as used in the OBBBA analysis, provides a richer but less frequently updated picture of the long-run effects of specific pieces of legislation.

Methodology

Overview

Since the 1980s, CBO has regularly decomposed changes in its budget projections into three categories: legislative changes (the static budgetary effects of newly enacted laws), economic changes (revisions to the economic forecast, including estimated effects of new laws on the economy), and technical changes (all other revisions). To capture the intended medium-run trajectory of fiscal policy, we focus on changes in the projections of debt-to-GDP 10 years out5 from each projection vintage that CBO attributes to legislation. For the February 2026 projection, to reflect the deficit effects of the administration’s tariff policies, we also include the CBO’s estimates of customs duties changes due to economic and technical factors.6 We cumulate these changes in these medium-run forecasts from 22 consecutive CBO projection vintages from August 2015 through February 2026. The result is a running total of how much enacted fiscal policy has added to the projected debt-to-GDP path. These impacts are visualized in Figure A1, which shows the cumulative change in projected debt-to-GDP since 2015 and since 2022.

To estimate the impact of increased government borrowing on Treasury yields we draw on recent research on the topic. To convert this debt accumulation into a Treasury interest rate effect, we assume long-term Treasury yields rise by 2 basis points for each percentage point increase in the forecasted debt-to-GDP ratio. This is the central estimate from Neveu & Schafer (2024, CBO Working Paper 2024-05), which extends the approach of Laubach (2009) through 2023. This estimated sensitivity captures both the fiscal risk premium that bondholders demand when the debt trajectory worsens, other term premia effects, and, to a lesser extent, the expected monetary-policy response. Plante et al. (2025) estimate that about 75% of the response appears in term premia and 25% in short-run rates. This methodology does not capture the full inflation-feedback and general-equilibrium channels modeled in TBL’s prior macroeconomic simulations and should therefore be interpreted as a lower bound on the total interest-rate effect of fiscal policy.

Lastly, we convert this rise in Treasury rates to rises in consumer borrowing rates and therefore borrowing costs. Based on the full USMM model used in our OBBBA analysis, we estimate that a 10 basis point rise in Treasury yields raises 30-year mortgage rates by a full 10 basis points, raises auto loan rates by 5 basis points, and small business loan rates by 2.5 basis points. Using recent data on typical loan amounts, durations, and current interest rates we estimate monthly payments in a counterfactual world where consumer rates are lower by the previously estimated amounts.

Detailed Steps

The methodology leverages the Laubach (2009) framework. The steps are:

  1. Extract the legislative component. From each CBO Budget and Economic Outlook projection starting in August 2015, we extract the portion of the change in projected deficits that CBO attributes to enacted legislation (as reported in CBO’s decomposition table).
  2. Harmonize to a ten-year window. Each vintage’s legislative deficit change is standardized to a ten-year projection horizon (t through t+9),7 summed to give an approximate change in debt,8 and then divided by CBO’s projected nominal GDP at the horizon year. Nominal GDP projections are included in CBO’s economic projections release, which is often but not always released contemporaneously with the budget projections. When a contemporaneous release is not available, we use the most recent prior vintage. The result is a legislative change in projected debt-to-GDP over the next 10 years, in percentage points, for that vintage.
  3. Cumulate across vintages. The per-vintage legislative debt-to-GDP changes are summed across consecutive CBO projections to form a cumulative series. We have 22 vintages between August 2015 and February 2026 in our sample.
  4. Apply the debt-to-rates sensitivity. The cumulative change in projected debt-to-GDP is multiplied by an estimated sensitivity of 2 basis points per percentage point to give a change in long-term Treasury yields. This is the central estimate from Neveu & Schafer (2024, CBO Working Paper 2024-05). We report a range using 1.5 basis points per percentage point (rounded from Furceri, Goncalves & Li 2025, fully-controlled specification in table 2) and 3 basis points (rounded from Plante, Richter & Zubairy 2025’s estimated impact on 10 year Treasury yields in table 5) per percentage point as lower and upper bounds.9
  5. Pass through to consumer rates. We calibrate the passthrough to consumer lending rates using the USMM model. In USMM, changes in long-term Treasury rates pass through to consumer interest rates directly. This passthrough is approximately 1-to-1 for 30-year mortgages and 50% for auto loans. In contrast, in USMM small business loan rates move nearly 1-for-1 with short-term interest rates. Following Plante et al., we assume that for every 10 bps the 10-year rate rises, short term rates rise 2.5bps. This gives an effective passthrough rate of 25% to small business loan rates.
  6. Translate into dollar costs. Rate effects are converted into annual payment differences using standard amortization, comparing observed rates to a counterfactual rate that removes the estimated fiscal-policy contribution. That is, we compute annual payments at the observed rate and at the counterfactual rate (observed rate minus the estimated fiscal-policy effect) and report the difference. All loan parameters reflect late 2025Q3 market data and SBA’s reported 7(a) average loan size for fiscal year 2025 so we can approximately match the data cutoff for the February 2026 CBO projection cycle.
    1. A 30-year mortgage taken out at the Q3 2025 median home sale price ($426,800, per the National Association of Realtors) with 20% down results in a principal of $341,440. We use the Freddie Mac PMMS data as of September 25, 2025 for our reference mortgage rate of 6.30%.
    2. The average SBA 7(a) small business loan principal was about $477,571 in fiscal year 2025 per the SBA’s lender report. We assume a 10 year term and that rates on small business loans equal the prime lending rate plus a 3% spread. We use the September 30th prime rate of 7.25% (as reported in series DPRIME in FRED) to give an assumed loan rate of 10.25%.
    3. The average new auto loan principal in Q3 2025 was $42,332, with an average term of about 69 months and an average interest rate of 6.56%, as reported in Experian’s 2025Q3 data release.

Footnotes

  • 1

    This estimate is from Neveu & Schafer (2024, CBO Working Paper 2024-05) and is in line with the long-run interest rate response to federal debt used by CBO in its analyses.

  • 2

    It is worth noting that CBO’s legislative decomposition reflects the static budgetary cost of enacted laws (the direct effects on revenues and outlays as scored at the time of enactment) and does not capture any resulting “dynamic” macroeconomic effects. CBO records these effects separately as economic, rather than legislative, changes.

  • 3

    CBO records the decrease in customs revenue due to tariffs’ trade-discouraging effects in their economic changes breakout. The technical component reflects the direct revenue effects of implemented tariffs.

  • 4

    See CBO, The Budget and Economic Outlook: 2026 to 2036, Chapter 5 and Box 5-1 for further discussion. CBO's projections assume that tariff rates in place as of November 20, 2025 are permanent and continue unchanged.

  • 5

    One alternative would be to use changes in realized debt issuance since that does rely on potentially biased or unrealistic forecasts; another would be to use changes in projected debt from each year out to some distant horizon (e.g., expected change in debt between the projection year and 2050) since the forward-looking nature of financial markets means that rate changes should be driven more by fiscal news than the current stock of debt. Our choice to use a fixed 10 year horizon represents a middle ground and aligns our exercise with Congress’ chosen fiscal policy planning horizon and the empirical literature on debt and interest rates.

  • 6

    In its projections, CBO assigns the increased customs duty revenue due to the administration’s tariff policies to its technical changes category and the projected reduction in import volumes to its economic changes category because these tariffs were implemented through executive action (primarily under IEEPA) rather than new legislation. For more discussion, see CBO, The Budget and Economic Outlook: 2026 to 2036, Chapter 5 and Box 5-1.

  • 7

    This is the furthest horizon CBO regularly forecasts as part of its regular Budget and Economic Outlook updates; the Long Term Budget Outlook gives 30 year forecasts but is only released annually.

  • 8

    Cumulated deficits are not exactly equal to changes in outstanding debt. Debt can also change due to other financing adjustments (for example, changes in Treasury cash balances and certain credit-financing mechanics) that are not part of reported deficits. These additional effects are quantitatively quite small in practice.

  • 9

    These papers’ main estimates use the 5-year-ahead forecasted debt/GDP ratio as their independent variable. Plante et al. also report some results using the 10-year-ahead projection of debt/GDP, like what we use in our calculations, and find similar responses.