The NG9-1-1 Data Model Requires Every Record to Name a Discrepancy Agency 

At a glance 

1. A cross-system address discrepancy is one physical location carrying different address values across systems inside a single agency. 
2. NENA 71-501 recommends a minimum 98 percent match rate between GIS and the MSAG before the GIS data goes into the routing database. 
3. The same document lists eight recurring causes of mismatch. Six of the eight are disagreements about how to write a name down. 
4. NENA-STA-006.2-2022 requires a Discrepancy Agency ID on every record in every layer, and section 8.4 records that guidance for reporting and resolving discrepancies has yet to be written. 
5. ACE compares the master address point dataset against the systems that disagree with it and returns the conflicts with a confidence score. 

What is a cross-system address discrepancy? 

A cross-system address discrepancy occurs when a physical location has different address values across various systems within a single agency. GIS holds one version, computer-aided dispatch holds another. The MSAG, the tax roll, the permit system and utility billing each contain a version that was correct when it was entered, but no longer. 

Every one of those values was entered by someone competent, following the conventions of the office they work in. 

What is the MSAG? 

The Master Street Address Guide is the E9-1-1 table of street names and house number ranges, grouped by community, that ties each range to an Emergency Service Number so a call reaches the right responders (NENA Knowledge Base). An MSAG address permits no variants. Directionals, spellings, street suffixes and community names each have one accepted form. 

NENA reserves the term MSAG discrepancy for the moment that rule fails in public: an ALI record retrieved during an actual 9-1-1 call carries wrong information because an MSAG field was wrong. In NG9-1-1, the jurisdiction’s GIS data takes over the validation and routing work the MSAG, ALI and selective router did under E9-1-1 (NENA-STA-006.2a-2022). Every disagreement between GIS and the MSAG is a disagreement about which version the routing system will trust. 

Why do address records drift between systems? 

Departments update addresses on their own cycles and for their own reasons. Assessment changes a situs address when a parcel splits, and permitting assigns a number to a lot months before anything gets built on it. Meanwhile, at three in the morning, a telecommunicator corrects an address in CAD because a responding unit could not find the house, and that correction never gets relayed back to anyone. 

Over the course of a decade, the cycles diverge. Each office keeps working within its own system, so the drift stays invisible until something forces two of them to agree. One House, Six Records traces that pattern through a single property. 

ACE uncovers this issue more often than any other. Moving data between those systems without first reconciling it just propagates the disagreement into a new place, the trap described in GIS for NG9-1-1 ETL. 

Where do GIS and MSAG values diverge? 

NENA 71-501, the information document for synchronizing GIS databases with the MSAG and ALI, issued May 26, 2009, lists the recurring causes: 

• Different road naming conventions 
• Inaccurate address ranges 
• Improper ESN assignments 
• Street type missing or misspelled 
• Post directional transposed 
• Community names differ 
• One word versus two-word names 
• Incorrect standardization 

Six of the eight are disagreements about how to write a name down. Two offices acknowledge the house exists but disagree on whether the street is Saint Andrews or St Andrews, and whether the community name should be the incorporated municipality or the postal city. The fact there’s a house was never in dispute. 

What match rate does NENA recommend between GIS and the MSAG? 

NENA 71-501 recommends “a minimum match rate of 98% be set prior to using the GIS data in the Emergency Routing Data Base (ERDB) or the Location to Service Translation (LoST) Protocol services.” Run the arithmetic on a county comparing 60,000 records and the tolerance permits 1,200 that don’t agree when the data goes live. 

The threshold was drafted for routing and sets the floor for publishing the dataset, well below where a well-maintained dataset typically stands after several years. 

Fifteen years on, the problem is still with us. NENA’s GIS Data Transition Information Document, NENA-INF-046.1-2024, published in September 2024, notes that as the transition progresses, conflicting information between the MSAG, ALI and GIS databases may arise. It also tells agencies to keep legacy attributes such as MSAG Community and ESN alive in the GIS layers throughout, because both environments have to work at once. 

A required field, but no procedure behind it 

NENA’s Standard for NG9-1-1 GIS Data Model, NENA-STA-006.2-2022, requires a Discrepancy Agency ID, DiscrpAgID, on every record in every layer. The field is mandatory and holds up to 100 characters, on road centerlines and address points alike. 

Section 8.4 of the same standard, Recommendation for Additional Development Work, records that the GIS Data Model Working Group “should consider developing guidance for the reporting and resolution of discrepancies identified through the quality control process.” 

The field naming who should be informed is mandatory. The procedure for informing them sits on a list of topics the working group may (or may not) take up. So, it’s been settled as to who gets told, but how they get told is still wide open.  

How can a county find the disagreements before a 9-1-1 call does? 

Pick the two systems most likely to disagree and compare them on a sample. The usual starting point is GIS against the tax roll, because both are maintained continuously and neither consults the other. 

The parcel identifier is the only usable key for that comparison, since the street address is the field in dispute and cannot be trusted to join anything. 

Count the categories before you count the records. Two hundred mismatches produced by a single street-type convention are a thirty-minute fix. Two hundred mismatches spread across two hundred separate causes may mean bringing on extra staff. 

Every record needs a discrepancy agency under the NG9-1-1 data model. The comparison produces that answer as a byproduct: each conflict category points to the office responsible. Record it in the DiscrpAgID field during the reconciliation, because rebuilding that mapping later means repeating the work. 

What ACE does with them 

DATAMARK’s Address Comparison Evaluation (ACE) compares the master address point dataset against additional sources the client supplies, and in most engagements, those sources are the systems above. Cross-system discrepancies come back alongside missing addresses and missing subaddresses, so an addressing office can separate the conflicts with evidence behind them from a single system’s typo. The terms are defined in the NG9-1-1 Glossary. 

Communities across the U.S. have identified more than 2M address candidates through ACE. Many clients report a second result: the counties are creating closer working relationships with the various addressing coordinators. Reconciliation leaves behind a standing relationship between offices, and that relationship keeps the next decade of updates aligned. 

The 98 percent threshold is a recommendation that grants exceptions for 2 percent of communities, which are just as important as the 98 percent. No standard sets an equivalent for the assessment roll or the utility file, and a resident who learned their address from a lease or a prior occupant may be reciting one that originated in a system error years ago. They’ll give that version to a call taker with full confidence, because they’ve got no reason to think it was ever wrong. Offices that want those files compared can start with the questions addressing offices ask before commissioning an ACE comparison.