NYS Tax Databases

Author: Dave
Date: 04.01.26 - 12:41am

So I ended up having to build a couple databases on NY tax Advisory Opinions, Tribunal decisions, memorandums, and Division of Tax Appeals decisions to run some research queries.

Download databases & Source files (1.2gb unpacked), ( read me / examples )

This project converts a scattered body of New York State tax materials into a structured research corpus: adjudicated decisions, Advisory Opinions, and Technical Services Bureau memoranda, all searchable through SQLite full-text indexing. The result is a practical research tool for examining how New York personal income tax law has actually been articulated and applied over time. :contentReference[oaicite:2]{index=2}

The real value of the dataset is that it makes pattern detection possible. Rather than relying on anecdote or isolated citations, users can measure outcome distributions, locate references to specific regulations, and evaluate how often particular sourcing or residency concepts appear across the published administrative record. :contentReference[oaicite:3]{index=3}

Not the most fun to build, but they scratched the itch. I will be filing for an Advisory Opinion soon I will post a copy of it here once submitted.

Database	Records	Date Range	Description
`dta.db`	5,926 cases	~1980–2026	All published ALJ determinations / Tax Appeals Tribunal
`advisories.db`	395 opinions	1980–2024	Personal income tax Advisory Opinions
`memoranda.db`	272 memos	1978–2018	NYS Technical Services Bureau Memoranda (TSB-M series)

For the advisory opinions...Prior to ~2019, submission dates were recoverable and often encoded in the petition number. In more recent opinions, submission dates are no longer disclosed in the published record (text, identifiers, or metadata), so processing time can no longer be independently measured. As a result, all post-~2019 opinions lack observable submission dates. In this dataset that affects ~30 recent entries:

I see two oddities I want to investigate before finalizing — they affect the credibility of the table:

1. **1997 shows 0.0 months for 9 opinions, and there are 18 zeros total in fastest list** — that's almost certainly a parsing artifact in the original data, not actual same-day processing 2. **1999 jumps to 54.0 months** because of those 4 outliers received in 1991 (the `.1` subscripts suggest reissued/corrected versions, not original receipts)

Let me check.Confirmed bug in the original ingestion: the 1997-1998 opinions don't have a date in the header (e.g. TSB-A-97(10)I jumps straight from "Income Tax" to "STATE OF NEW YORK" — no date), so whatever upstream parser populated `opinion_timing` for those used the *received_date* as a fallback for the issued_date. That's not data we can fix here without finding the actual issue dates.

The 1991/1999 outliers are also legitimately weird — those `.1` suffixed IDs look like reissued/corrected versions of much older opinions, so 105 months is real but a different kind of event.

Both of these were already in your existing data (not introduced by the FOIL inserts), so I'll flag them and report cleanly. Let me regen the tables and add a few extra cuts I think are meaningful given the FOIL data dramatically changes the post-2019 picture.Numbers are solid. The 2024 batch finding is striking — 13 of the 14 opinions issued in 2024 came out within a 6-day window in late April. Let me write this up cleanly.

--- ## Regenerated tables (now with full FOIL data)

A note on data hygiene before the numbers: 18 rows in your existing data have `received_date == issued_date` (months=0). Spot-checking the 1997-98 source documents, those opinions don't have an issue date in the header at all — looks like the original ingestion fell back to receipt date when it couldn't parse one. There are also 4 `(N.1)` reissue/correction IDs that show 105-month "delays" but are really republications of much older opinions. **I excluded both groups (23 rows total) from the clean stats below — they're noise, not signal.** That brings the working set to 372 opinions.

======================================================================
PROCESSING TIME BY DECADE (issued year)
======================================================================
Decade     Count    Avg        Min    Max    Median
----------------------------------------------------------------------
1980s      113      5.8 mo     1      34     4.0
1990s       92      3.7 mo     0      12     3.0
2000s       78      7.2 mo     1      28     6.0
2010s       61     10.4 mo     1      36     8.0
2020s       28     41.1 mo     8      85    42.5

The 2020s line is the headline finding. With your old data there was 1 opinion in that decade (42 months). With the FOIL data there are 28, and the average is essentially unchanged at 41 months — confirming the 2020(7)I outlier was representative, not an anomaly.

======================================================================
SLOWEST 20 OPINIONS
======================================================================
   85 mo  [2024] TSB-A-24(1)I,(7)S   recv 2017-06-15  issd 2024-07-16  [FOIL]
   71 mo  [2024] TSB-A-24(8)I        recv 2018-05-07  issd 2024-04-24  [FOIL]
   58 mo  [2020] TSB-A-20(5)I        recv 2015-12-11  issd 2020-10-20  [FOIL]
   57 mo  [2024] TSB-A-24(2)I        recv 2019-07-05  issd 2024-04-18  [FOIL]
   57 mo  [2024] TSB-A-24(9)I        recv 2019-06-27  issd 2024-04-24  [FOIL]
   53 mo  [2024] TSB-A-24(3)I        recv 2019-11-12  issd 2024-04-19  [FOIL]
   52 mo  [2020] TSB-A-20(12)I       recv 2016-06-16  issd 2020-10-27  [FOIL]
   51 mo  [2020] TSB-A-20(13)I,(4)C  recv 2016-07-18  issd 2020-11-03  [FOIL]
   51 mo  [2024] TSB-A-24(13)I       recv 2019-12-20  issd 2024-04-18  [FOIL]
   49 mo  [2024] TSB-A-24(5)I        recv 2020-03-05  issd 2024-04-24  [FOIL]
   49 mo  [2024] TSB-A-24(12)I       recv 2020-03-11  issd 2024-04-24  [FOIL]
   44 mo  [2020] TSB-A-20(3)I        recv 2016-06-06  issd 2020-02-10  [FOIL]
   44 mo  [2020] TSB-A-20(9)I        recv 2017-01-12  issd 2020-10-06  [FOIL]
   43 mo  [2020] TSB-A-20(4)I        recv 2016-10-20  issd 2020-06-16  [FOIL]
   42 mo  [2020] TSB-A-20(7)I        recv 2017-03-24  issd 2020-09-29
   41 mo  [2020] TSB-A-20(8)I        recv 2017-04-13  issd 2020-10-06  [FOIL]
   38 mo  [2020] TSB-A-20(6)I        recv 2017-07-17  issd 2020-09-29  [FOIL]
   38 mo  [2020] TSB-A-20(11)I       recv 2017-08-23  issd 2020-10-27  [FOIL]
   36 mo  [2017] TSB-A-17(2)I        recv 2014-03-06  issd 2017-03-01
   35 mo  [2020] TSB-A-20(2)I        recv 2017-02-17  issd 2020-02-04  [FOIL]

The previous slowest list was dominated by those 4 phantom 105-month entries. With those removed and the FOIL data added, the slowest-20 is now an almost solid block of 2020 and 2024 issuances. Pre-FOIL, you had **1** entry over 36 months in the entire post-2000 era. Post-FOIL: **18**.

======================================================================
DISTRIBUTION OF PROCESSING TIMES
======================================================================
  0-3 mo        125  (33.6%)  #################
  4-6 mo        111  (29.8%)  ###############
  7-12 mo        70  (18.8%)  #########
  13-24 mo       34  ( 9.1%)  #####
  25-36 mo       14  ( 3.8%)  ##
  37-60 mo       16  ( 4.3%)  ##
  61-120 mo       2  ( 0.5%)

63% of opinions historically came out within 6 months. The "37-60 mo" tail is almost entirely the FOIL set.

======================================================================
BY YEAR (received) — when did the wait start?
======================================================================
  2013   15.0 mo  ( 5)  ###############
  2014   17.8 mo  ( 8)  ##################
  2015   22.0 mo  ( 9)  ######################
  2016   34.0 mo  ( 6)  ##################################
  2017   38.2 mo  ( 9)  ######################################
  2018   28.7 mo  ( 3)  #############################
  2019   44.0 mo  ( 5)  ############################################
  2020   30.2 mo  ( 4)  ##############################
  2021   34.5 mo  ( 2)  ##################################
  2022   21.5 mo  ( 4)  ######################

I'd argue this view is more meaningful than "by issue year" for what you're probably looking at — it shows what a petitioner who filed in year X actually waited. Petitions submitted in 2013 waited about 15 months. By 2019 it's 44 months. There's a clear monotonic climb 2013→2019 with a clean inflection point.

======================================================================
THE 2024 BATCH — were these issued together?
======================================================================
  2024-04-18:  3 opinions  TSB-A-24(2)I, (4)I, (13)I
  2024-04-19:  1 opinion   TSB-A-24(3)I
  2024-04-24:  9 opinions  TSB-A-24(5)I, (6)I, (7)I, (8)I, (9)I, (10)I, (11)I, (12)I, (14)I
  2024-07-16:  1 opinion   TSB-A-24(1)I,(7)S

This is the new finding I think is most worth flagging: **13 of the 14 opinions issued in 2024 came out in a single 6-day window (April 18-24)**. These are petitions that had been sitting for an average of 50 months. The Department wasn't spreading these out as they got resolved — they were processed as a batch dump. That pattern is hard to explain by ordinary workload management; it looks more like a clearance operation.

======================================================================
SUMMARY
======================================================================
Total opinions:       395
Clean for stats:      372  (excluded: 18 parsing artifacts, 4 reissues)
Overall average:      9.0 months
Overall median:       5.0 months

1980–1999 average:    4.8 months  (205 opinions)
2000–2009 average:    7.2 months  (78 opinions)
2010–2019 average:   10.4 months  (61 opinions)
2020–2024 average:   41.1 months  (28 opinions)

The headline shift: **average processing time roughly 8.5x'd between the 1990s and the 2020s.** The 2010s already showed a doubling, and then the 2020s break the chart entirely.

A couple of notes on what these numbers can and can't say:

The 2020–2024 average is 41.1 months across 28 opinions, but that's only the opinions we have receipt dates for — meaning those obtained via FOIL plus the one self-disclosed 2020(7)I. Any 2020+ opinions that were issued *quickly* and weren't in your FOIL request wouldn't appear in this set. So this is biased toward slow opinions — but it's also the *only* observable processing-time data for that era, since the Department stopped publishing receipt dates in identifiers around 2019. The bias cuts both ways: you could argue the true 2020s average might be lower if there are fast-resolved opinions you haven't captured, but you could equally argue the FOIL request specifically returned the cases where the department had something to disclose, so the "lost" cases are likely the older ones still unresolved (which would push the average higher).

The "By year (received)" data only goes through 2022 because anything received in 2023 or later that hasn't yet been issued isn't in the database at all — those represent ongoing/unresolved petitions whose true wait times are still accumulating.

Comments: (2)

On 04.03.26 - 9:00pm dave wrote:

NYS Tax Databases: Outcome Statistics

A few people have asked about win rate numbers so here is the full outcome breakdown from dta.db.

Full Corpus Outcomes

Across all 5,926 published ALJ determinations and Tax Appeals Tribunal decisions:

Outcome Count %

division_won 3,363 57.3%

taxpayer_won 704 12.0%

mixed 566 9.6%

dismissed_procedural 557 9.5%

unknown 161 2.7%

motion_denied 145 2.5%

dismissed_jurisdiction 139 2.4%

dismissed 105 1.8%

remanded 73 1.2%

motion_granted 27 0.5%

unreviewed 19 0.3%

dismissed_moot 8 0.1%

The 12% taxpayer win rate is the all-in figure including procedural and jurisdictional dismissals. The more useful number for anyone deciding whether to litigate is the merits-only rate. Restricting to taxpayer_won + division_won + mixed gives 4,633 cases where substance was actually decided. Against that denominator: taxpayers won outright ~15% of the time, the Division ~73%, mixed ~12%.

Nonresident / Remote Work Subset

727 cases are flagged as involving nonresident sourcing, remote work, or the convenience of the employer rule:

Outcome Count %

division_won 460 63.3%

taxpayer_won 114 15.7%

mixed 86 11.8%

unknown 23 3.2%

dismissed_procedural 22 3.0%

remanded 7 1.0%

motion_denied 7 1.0%

dismissed_jurisdiction 5 0.7%

motion_granted 3 0.4%

On merits-decided cases in this subset (460 + 114 + 86 = 660): the Division prevailed in ~70%, taxpayers won outright ~17%, mixed ~13%. The Divisions win rate is slightly higher in nonresident/remote cases than in the general corpus, consistent with the Departments historically aggressive sourcing positions.

Note: the boolean flags driving this subset (is_nonresident_case, is_remote_work_case, is_convenience_rule) are regex-generated and not manually verified at scale. Reliable for triage but not exhaustive — some relevant cases will not be flagged.

show query

SELECT outcome, COUNT(*) as cnt FROM case_analysis WHERE is_nonresident_case = 1 OR is_remote_work_case = 1 OR is_convenience_rule = 1 GROUP BY outcome ORDER BY cnt DESC;

On 04.06.26 - 7:46pm Dave wrote:
The FOIL results for the missing 31 Advisory Opinion submission dates are now integrated 5.5.26

Leave Comment:

Name:

Email: (not shown)

Message: (Required)

Math Question: 64 + 53 = ? followed by the letter: C

About Me
More Blogs
Main Site

Posts:

2026 (5)

Micro Sample Handling

NYS Tax Databases

Sclerotic Dentin Fragment

Solar education