SSA Disability Award Statistics: The Federal Dataset Behind 8 Million Benefit Decisions

The Social Security Administration adjudicates roughly 2.5 million disability applications per year and publishes the outcome data for every one of them. The result is a longitudinal federal dataset covering 8.3 million SSDI beneficiaries and 7.5 million SSI recipients—segmented by state, primary diagnosis, age group, gender, and the specific decision stage that granted benefits. Almost no one uses it outside of SSA itself.

Two programs, one dataset

The SSA administers two distinct federal disability programs that are often conflated but operate under different statutory authority, different eligibility rules, and different funding mechanisms. The award statistics published in the Annual Statistical Supplement separate them throughout, which matters for any analysis that is trying to understand program dynamics rather than just aggregate headcounts.

Social Security Disability Insurance (SSDI), authorized under Title II of the Social Security Act, is an earned benefit. Eligibility requires a sufficient work history measured in Social Security credits—generally 40 credits, with 20 earned in the last 10 years, though the requirement scales down for younger workers. Benefits are funded through payroll tax contributions, which means SSDI is structurally analogous to an insurance product: workers pay premiums (FICA taxes) over their careers and the program pays claims when a qualifying disability occurs. Monthly benefit amounts are calculated from the worker's earnings history, the same formula used for Social Security retirement benefits. As of recent data, the average SSDI monthly benefit is approximately $1,537.

Supplemental Security Income (SSI), authorized under Title XVI, is a needs-based program. There is no work history requirement. SSI is available to adults and children who are disabled, blind, or aged 65 and older and who have limited income and resources. The federal benefit rate is uniform nationally—$943 per month for an individual in 2024—and some states supplement it with additional state-funded amounts. SSI is funded from general revenue, not the Social Security Trust Fund, which means SSI solvency is not a separate concern the way SSDI Trust Fund solvency is.

A significant subset of beneficiaries receive both programs simultaneously, typically because their SSDI benefit is low enough that SSI fills the gap to the federal benefit rate. These concurrent beneficiaries appear in award data for both programs, which inflates the headline count if the two programs are summed naively. The supplement tables flag concurrent awards explicitly.

The disability determination process and its stages

The structure of the SSA award statistics reflects the multi-stage adjudication process, and understanding that process is essential for interpreting the data correctly. Awards are counted at the stage where the favorable decision was issued, not at application. A claimant who is denied initially, denied at reconsideration, and then awarded at an ALJ hearing appears in the ALJ award count, not the initial award count.

Stage one is the initial determination, made by Disability Determination Services (DDS) in each state—state agencies that contract with the federal SSA. DDS examiners and medical consultants review the application, medical records, and work history against the SSA's five-step sequential evaluation process. Initial denial rates have historically run 60 to 65 percent of applications.

Stage two is reconsideration, also handled at the DDS level. Claimants who are denied initially have 60 days to request reconsideration. Reconsideration denial rates are even higher than initial denial rates—often 85 to 90 percent—because most cases that reach reconsideration were already evaluated and found lacking. SSA has experimented with skipping the reconsideration stage in some states as part of the Prototype initiative, though the experiment has had mixed results.

Stage three is the ALJ hearing before an Administrative Law Judge, and this is where the majority of all eventual awards are made. Claimants who are denied at reconsideration have 60 days to request an ALJ hearing. The ALJ reviews the record de novo, may hear testimony from the claimant and from vocational experts and medical experts, and issues an independent decision. ALJ allowance rates have historically been 45 to 55 percent of cases decided, substantially higher than DDS allowance rates, which has generated ongoing debate about whether the process is consistent.

Stage four is the Appeals Council, which reviews ALJ decisions for legal error. The Appeals Council can affirm, reverse, or remand the ALJ decision. Stage five is federal district court review, available when the Appeals Council declines review or issues an unfavorable decision. The statistical supplement tracks awards through the ALJ level in the standard tables; Appeals Council remand outcomes are covered in separate administrative workload data.

The ALJ backlog

The most consequential operational fact in the SSA disability data—and the one with the largest human impact—is the ALJ hearing backlog. At its peak in fiscal year 2017, average wait time from hearing request to decision exceeded 600 days, or roughly 20 months. In some of the most backlogged hearing offices, wait times exceeded 900 days—two and a half years.

The backlog is visible in the SSA's Hearing Office Workload Data, published separately from the Annual Statistical Supplement at ssa.gov/appeals/DataSets/. The workload data includes, for each hearing office, the number of pending cases, receipts, dispositions, allowance rate, and average processing time in days. This dataset is the operational complement to the award statistics: the supplement shows who got awarded and what condition they had; the workload data shows how long they waited and which office processed them.

The backlog built up primarily between 2008 and 2015, driven by the Great Recession's surge in applications combined with ALJ hiring freezes. SSA began addressing it through a combination of additional ALJ hiring, a case management initiative that prioritized oldest cases, and a “quick decision” protocol for cases that clearly met listing criteria. By fiscal year 2023, average processing times had fallen to approximately 420 days—still over a year, but substantially improved from peak.

The human cost of the backlog is substantial. Claimants who are eventually awarded benefits at the ALJ stage typically receive retroactive payments covering the period from their established onset date (up to 12 months before application for SSDI) through the award date. The average retroactive payment in recent years has exceeded $12,000, reflecting the duration of the wait. Some claimants die before their hearing; SSA tracks and reports these cases separately as “critical” case designations.

Geographic variation in award rates

The state-level award tables in the supplement reveal striking geographic variation that is only partially explained by population differences. Normalizing by working-age population, several Appalachian and Deep South states have SSDI award rates two to three times higher than the national average. West Virginia, Arkansas, and Alabama consistently rank among the highest per-capita SSDI award states. Massachusetts, Utah, and Hawaii consistently rank among the lowest.

The geographic pattern reflects overlapping factors that can be partially decomposed using the supplement's condition and age breakdowns. High-award states tend to have older working-age populations (older workers have higher disability prevalence), higher concentrations of employment in physically demanding industries (mining, logging, agriculture, manufacturing), and lower average educational attainment (which affects the “ability to do other work” step of the sequential evaluation process). The occupational and educational factors matter specifically because the SSA's step five analysis asks whether the claimant can perform any substantial gainful activity in the national economy—and claimants with lower transferable skills have a narrower range of alternative occupations available to them.

State DDS variation also plays a role. DDS agencies are state-administered and vary in how they interpret medical evidence, how they apply the SSA's Listing of Impairments, and how aggressively they develop the record before issuing an initial decision. SSA quality reviews find that DDS decision quality varies across states in ways that are not fully explained by the claimant population. This is one reason the Appeals Council and federal court stage exists: to provide a check on DDS decisions that deviate from SSA policy.

Top disabling conditions

The supplement's condition breakdowns use the SSA's own diagnostic groupings, which map to ICD-10 codes but are organized into categories that reflect the Listing of Impairments rather than standard clinical classifications. The two dominant categories together account for approximately 60 percent of all awards.

Musculoskeletal disorders are the single largest category, accounting for roughly 33 percent of SSDI awards. Back disorders—degenerative disc disease, herniated nucleus pulposus, spinal stenosis—dominate within this group. The musculoskeletal prevalence reflects both the physical demands of the working-age population's occupational distribution and the aging of baby boomers through the 50–64 age range, where musculoskeletal impairments peak. This category has been growing as a share of awards for two decades.

Mental disorders are the second largest category, accounting for roughly 27 percent of SSDI awards and a higher share of SSI awards (because SSI covers children and non-working adults, among whom mental health diagnoses predominate). The mental disorders category includes mood disorders (depression, bipolar disorder), anxiety disorders, schizophrenia and other psychotic disorders, intellectual disability, and autism spectrum disorder. The growth of the mental disorders category since 2000 is one of the more significant structural shifts in the disability program's composition.

The remaining 40 percent of awards is distributed across circulatory system disorders (heart disease, stroke), neoplasms (cancer), nervous system disorders (multiple sclerosis, Parkinson's, epilepsy), respiratory conditions, and a residual “other” category. The supplement breaks each of these down by state, allowing condition-specific geographic analysis.

The SSA Annual Statistical Supplement

The primary public data source for disability award statistics is the SSA Annual Statistical Supplement to the Social Security Bulletin, published each year with a one-to-two year lag. The 2023 edition, for example, covers data primarily through calendar year 2022. The supplement is accessible at ssa.gov/policy/docs/statcomps/supplement/ and is one of the most comprehensive statistical publications the federal government produces.

The supplement is organized into numbered sections. Section 6 covers OASDI (Old-Age, Survivors, and Disability Insurance) benefit data; Tables 60 through 65 contain the disability award statistics that are most useful for program analysis. Table 60 covers SSDI awards by state and diagnostic group. Table 61 disaggregates awards by age and gender. Table 62 covers awards to disabled workers by state and decision level—this is where the initial/reconsideration/ALJ breakdown lives. The SSI tables in Section 7 follow the same structure with a different series numbering.

The supplement is published in PDF form with embedded data tables that can be copied into spreadsheets, and in recent years SSA has also made CSV and Excel exports available for the most-requested series. For bulk access, the SSA's Open Data portal at ssa.gov/data/ provides machine-readable versions of some statistical series, though coverage is less complete than the supplement itself. Researchers who need time series going back to the 1970s should consult the supplement directly; the electronic extracts typically cover only the most recent decade.

A separate but related publication is the SSA Office of Retirement and Disability Policy's “Annual Report of the SSI Program,” which provides more granular SSI data than the supplement and includes state agency performance metrics for DDS agencies. The Annual Report is also available at ssa.gov/policy/.

Data structure

The core award tables in the supplement have a consistent structure across years, organized as state-by-condition matrices with age and gender sub-columns. Understanding the table layout is necessary before attempting programmatic extraction.

Python: loading public data and calculating state award rates

The following snippet demonstrates how to load a state-level SSDI awards table exported from the supplement, merge it with Census population estimates, and compute per-capita award rates and condition shares. The SSA does not provide a bulk API; the workflow assumes the supplement tables have been exported to CSV from the PDF.

import pandas as pd
import requests
from io import StringIO

# SSA Statistical Supplement tables are available as CSV exports
# from https://www.ssa.gov/policy/docs/statcomps/supplement/
# The Table 60 series covers SSDI awards; Table 65 series covers SSI.
# Download the state-level award tables for a given year and load them.

# Example: load a state-level SSDI awards CSV (manually exported from supplement)
url = "https://www.ssa.gov/policy/docs/statcomps/supplement/2023/60.pdf"
# For programmatic access, use the pre-formatted CSV extracts published by SSA
# or scrape the HTML tables from the Statistical Supplement pages.

# Sample structure after loading a state-level awards table:
# columns: state, total_awards, musculoskeletal, mental_disorders,
#          circulatory, neoplasms, nervous_system, other, age_lt30,
#          age_30_39, age_40_49, age_50_59, age_60plus, male, female

# Load a CSV extracted from the SSA supplement
df = pd.read_csv("ssa_ssdi_awards_by_state_2022.csv")

# Merge with Census population estimates to calculate award rates
pop = pd.read_csv("census_state_population_2022.csv")
df = df.merge(pop[["state", "population"]], on="state", how="left")

# Awards per 100,000 population
df["award_rate_per_100k"] = (df["total_awards"] / df["population"]) * 100_000

# Rank states by award rate
df_ranked = df.sort_values("award_rate_per_100k", ascending=False)

print(df_ranked[["state", "total_awards", "award_rate_per_100k"]].head(10))

# Condition share analysis
condition_cols = [
    "musculoskeletal", "mental_disorders", "circulatory",
    "neoplasms", "nervous_system", "other"
]
national = df[condition_cols].sum()
national_pct = (national / national.sum() * 100).round(1)
print("\nNational SSDI awards by condition category (%):")
print(national_pct.to_string())

# ALJ hearing outcomes require the separate ODAR workload data
# available at https://www.ssa.gov/appeals/DataSets/01_Hearing_Office_Workload_Data.html
alj = pd.read_csv("alj_hearing_office_workload_FY2023.csv")
alj["allowance_rate"] = alj["allowed"] / (alj["allowed"] + alj["denied"])
avg_wait = alj["avg_processing_days"].mean()
print("\nAverage ALJ wait (days) across offices: " + str(round(avg_wait, 1)))

Award rates calculated this way are the right unit of analysis for geographic comparison. Raw award counts are dominated by state population size; West Virginia does not appear at the top of the award-rate ranking because it is large—it appears there because its award rate per working-age resident is approximately 2.2 times the national average. The Census population denominator should ideally be the working-age population aged 18 to 64 rather than total population, because SSDI is only available to workers who have not yet reached retirement age.

Policy significance

The disability award statistics are not merely a descriptive record. They are the operational input to several consequential policy decisions.

Program integrity and fraud detection: SSA's Office of the Inspector General uses award pattern data to flag statistical anomalies—hearing offices with unusually high allowance rates, ALJs whose individual decision patterns deviate significantly from their peers, and geographic clusters of awards with common medical evidence sources. Several major fraud prosecutions have involved doctors, lawyers, and hearing office employees who manipulated medical evidence across hundreds of cases. The award statistics at the office and adjudicator level (available in the workload data) are a prerequisite for this kind of anomaly detection.

SSDI Trust Fund solvency: The Disability Insurance Trust Fund is separate from the Old-Age and Survivors Insurance Trust Fund. Award levels directly determine benefit outflows, and the long-run solvency of the DI Trust Fund depends on the ratio of contributing workers to beneficiaries. The Social Security Trustees' Annual Report projects DI Trust Fund depletion dates using award trend assumptions drawn from the supplement data. Based on recent Trustee projections, the DI Trust Fund has a somewhat more comfortable near-term outlook than the OASI Trust Fund, partly because SSDI applications declined during the tight labor markets of 2019–2023. The supplement data is the primary input for validating Trustee assumptions.

Budget projections: The Congressional Budget Office uses SSA award statistics to project disability program costs under different policy scenarios. Proposals to tighten or loosen eligibility criteria, change the grid rules for older workers, or modify the medical listing thresholds all have award-volume implications that flow directly through the CBO baseline. The supplement provides the age-condition-state breakdown needed to model which populations would be affected by any specific rule change.

State DDS performance accountability: Because DDS agencies are state-run but federally funded, SSA conducts ongoing quality assurance reviews of DDS decisions. States with error rates above SSA's thresholds are required to implement corrective action plans. The state-level award rates in the supplement, combined with DDS quality review data (published separately), create a multi-dimensional picture of how state agencies are performing relative to federal standards.

Limitations

The public supplement data has gaps that matter for certain research applications. Individual-level microdata—records at the claimant level rather than the state-condition aggregate—is not publicly available. Researchers who need claimant-level data must apply for access through SSA's data-sharing agreements, which require an institutional affiliation, a research protocol review, and typically a multiyear processing time.

The supplement tables do not include denial statistics in the same detail as award statistics. Denial rates at each stage are reported in aggregate but not broken down by the same state-condition-age matrix as awards. This makes it impossible to construct full funnel metrics—applications to awards by diagnosis—from the public supplement alone. The workload data provides some of this, but not with the condition granularity of the award tables.

Race and ethnicity data is limited. The supplement includes some race breakdowns for SSI, where demographic data is collected at enrollment, but SSDI race data is much less complete because SSA collects it from earnings records rather than applications, and coverage has historically been incomplete. The gap makes it difficult to assess racial disparities in award rates using the public supplement alone.

Related writing: The demographic backbone: using Census ACS data to contextualize every other federal dataset covers the American Community Survey—the source of state population denominators used to calculate per-capita disability award rates.

Related writing: Food stamps by the numbers: using USDA SNAP participation data to track hunger and benefit policy covers the other major federal means-tested benefit program—one that overlaps substantially with the SSI recipient population.

Related writing: Inside the count: using BJS National Prisoner Statistics to analyze incarceration trends covers another long-running federal administrative dataset where state-by-state variation and decision-maker discretion drive dramatically different outcomes within a shared federal framework.