Will this address land in an Opportunity Zone 2.0?

How we calculate this score

Eligibility. A tract can be nominated only if it is a low-income community under IRC §1400Z-1(c)(1), as amended by P.L. 119-21 §70421(b): median family income at or below 70% of the area median (the metropolitan area median for metro tracts, the statewide median otherwise), or a poverty rate of at least 20% with median family income not above 125% of that area median. We use Treasury's September 23, 2026 eligible-tract data (25,461 tracts, ACS 2020–2024), not our own calculation. Some state notices count differently (Alaska's lists 32; Treasury lists 33).

Status. On a state's published nomination list: 90% (nominated, pending Treasury certification). The state published its list and the tract isn't on it: 2%. Not on Treasury's eligible list: no estimate (it can't be nominated, so the result shows "Not on Treasury's eligible list" with no percentage). Certified by Treasury: 100% (no tract is certified yet). These status-based percentages come from published lists, not the model, so they carry no confidence label. Status comes from our nomination tracker.

Computed estimate (everything else). A logistic model fitted on the 4,936 eligible tracts in the seven states whose tract lists were public when it was fitted (Delaware, Georgia, Indiana, South Carolina, Texas, Utah, Washington; 1,242 nominated). What raises the odds: being a current OZ 1.0 tract (the strongest signal; partly overlapping OZ 1.0 tracts count less), a higher poverty rate relative to the rest of the state, and a state cap that covers a larger share of its eligible tracts. What lowers them: high population density relative to the rest of the state, many other eligible tracts in the same county (states spread picks across counties), a population under 1,000, and rural status once density is accounted for. The probability is rounded and held between 3% and 85%; no computed score can look like a certainty. Local nominations, project readiness, shortlists and state priorities drove most picks and are not in the model; no shortlist adjustment is applied.

Bands. For tracts without a published state list we show a band rather than a percentage: Low (model probability under 18%), Medium (18–30%) and High (30% or more). Tested leave-one-state-out, tracts in those bands were nominated 15%, 25% and 40% of the time.

How well it predicts. We tested the model leave-one-state-out: fitted on six of the seven states, scored the seventh, and repeated. The average within-state AUC was 0.70 (0.5 is a coin flip, 1.0 is perfect) and the pooled AUC 0.65; on average 49% of each held-out state's nominated tracts fell in the model's top picks, against about 25% by chance. That is a modest signal: treat a band as a rough guide, not a forecast. Accuracy varied by state, from 0.95 in Delaware to 0.55 in Georgia (close to chance).

Updates. Scores for all 85,529 census tracts are precomputed from the latest tracker status and rebuilt whenever our data updates, so a newly published state list replaces model estimates with 90% or 2%. The model is refitted when more states publish lists. ZIP results are weighted by each tract's share of the ZIP's land area (Census ZCTA relationship file).

Sources: P.L. 119-21 §70421 · 26 U.S.C. §1400Z-1 · CDFI Fund OZ 1.0 list · Census Geocoder · TIGERweb. Not tax or investment advice.

FAQ

Is OZ 2.0 live?

No OZ 2.0 designations exist yet; Treasury's certifications take effect January 1, 2027.

How is the likelihood calculated?

Tracts not on Treasury's September 2026 eligible list can't be nominated, so they get no estimate: the result says "Not on Treasury's eligible list" with no percentage. Tracts on a state's published nomination list score 90 (pending Treasury certification; 100 is reserved for Treasury certification), and eligible tracts left off a published list score 2. Every other eligible tract gets a model estimate from a logistic model fitted on the seven states whose tract lists were public when it was fitted, shown as a Low, Medium or High band. In leave-one-state-out testing it separated nominated from other tracts with an average within-state AUC of 0.70 (0.65 pooled across states), a modest signal. Scores are rebuilt from the latest tracker status whenever our data updates.

What does nominated mean?

A nominated tract is on the list a governor submitted to Treasury. It is pending Treasury certification, and it is not an Opportunity Zone until Treasury certifies it.

Is my address stored?

No. A street address goes to our geocoder (the U.S. Census Geocoder, with Mapbox as a fallback) to find its census tract and is not logged by us. Analytics record only the result type and score band, never the address. Only if you sign up for tract alerts under a result do we save your email with that tract and the address you searched, so we can email you when its status changes.