GL provides independent structural diagnosis of government administrative systems — using your own public audit data to measure what your office already knows but cannot say.
Government audit offices operate inside institutional constraints. GL operates outside them.
Independent GL analysis of publicly available state audit records.
Massachusetts spent millions automating its unemployment system. Appeals backlogs grew nine times over. On-time payments collapsed to 49% — worst in the nation. This occurred while new claims were at a six-year low.
GL diagnosis: denominator acceleration. Speed was applied to a broken classification logic. The system produced errors faster, routed them to appeals faster, and depleted adjudicator capacity faster. Automation did not fix the friction. It amplified it.
California EDD failed federal timeliness standards for all but one month since January 2013. Its 43% appeal overturn rate is not a fraud signal. It is a friction signal: the system is denying valid claims at the front end, forcing claimants through appeals, and reversing those decisions at the back end.
GL diagnosis: the system spends more processing appeals for correctly eligible claimants than it would have spent paying them in the first place. Rework is the invisible denominator.
Texas Workforce Commission holds a $1B UI Trust Fund balance and presents it as fiscal health. But only 23% of unemployed Texans receive benefits — ranking 28th nationally. The fund stays solvent partly because the system's friction excludes eligible claimants before they can reach it.
GL diagnosis: a system that achieves solvency through exclusion has not solved its efficiency problem. It has transferred the cost to unemployed workers who never appear in the denominator.
A 2015 audit identified 31 problems. A 2019 audit found over 50% unresolved. A 2021 audit found 14 issues still outstanding — including processing errors from 2014. A $244M modernization over five years produced no follow-up audit. The last progress report: June 2024. Then silence.
Florida has the worst or second-worst UI recipiency rate nationally. Governor DeSantis himself called the system a "jalopy" with "pointless roadblocks." The data is there. The pattern is visible. What is missing is an independent voice to say what the data shows.
$5.2B in fraudulent and improper payments while legitimate claimants experienced delays. This is the signature pattern of a high-friction system: it fails in both directions simultaneously, because the underlying process design cannot distinguish between valid and invalid claims.
GL diagnosis: fraud at that scale is not primarily a criminal problem. It is a process design problem. The same structural failure that let $5.2B flow incorrectly is the same failure that blocked legitimate claimants from accessing what they were owed.
Because GL does not set the goal. It only asks: you said you would do X. Did you do X? How efficiently?
You say small government, low spending, no dependency — fine. But when an eligible claimant is wrongly denied, appeals, gets re-adjudicated, then finally paid, TWC spent 3× the admin cost. That is not protecting taxpayers. That is wasting taxpayer money on rework. GL quantifies that waste.
You say protect vulnerable populations, ensure benefits reach those in need — fine. But when IDES has $5.2B in fraud while legitimate claimants wait, the system design is letting money flow to the wrong places. GL quantifies that leakage.
"GL does not judge whether the benefit should exist.
It only measures whether the system served the purpose it was built for."
Forthcoming March 13, 2026. This article introduces the GL Framework's application to public administrative systems — arguing that institutional failure is not a resource problem but a denominator design problem. The Massachusetts DUA serves as the central case study.
Request Pre-Publication Copy →This is a research conversation, not a sales call. I reach out to you. You decide whether to continue.
Your name, office, and state. One or two sentences on what you are observing in your system. No formal documentation required.
I will indicate whether your state's public audit data supports a GL analysis, and what that analysis could produce.
To confirm methodology fit. The analytical work and any public exposure remain mine. Your underlying audit record remains yours.
政府審計辦公室在體制內部運作。GL 在體制外部運作。
基於公開州審計記錄的獨立 GL 分析。
麻州花費數百萬美元自動化其失業系統。上訴積壓案件成長了九倍。準時付款率崩潰至 49%——全國最差。而這發生在新申請量處於六年低點之際。
GL 診斷：分母加速。速度被套用在一個破損的分類邏輯上，更快地產生錯誤、更快地導向上訴、更快地耗盡裁決者能量。
加州 EDD 自 2013 年 1 月以來幾乎每個月都未達聯邦及時性標準。43% 的上訴推翻率不是欺詐信號，而是摩擦信號：系統在前端拒絕了有效申請，迫使申請人通過上訴，然後在後端推翻那些決定。
GL 診斷：系統花在處理合格申請人上訴的費用，比直接付款給他們還要多。返工是不可見的分母。
2015 年審計發現 31 個問題。2019 年審計發現超過 50% 未解決。2021 年審計發現 14 個問題仍未處理——包括 2014 年的處理錯誤。2.44 億美元的五年現代化改造沒有產生後續審計。最後一份進度報告：2024 年 6 月。然後是沉默。
數據就在那裡。模式清晰可見。缺少的是一個獨立的聲音來說出數據所顯示的事。
因為 GL 不設定目標。它只問：你說你會做 X。你做到了嗎？效率如何？
你說小政府、低支出、不依賴福利——好。但當一個合格申請人被錯誤拒絕、上訴、重新裁決、最終獲得給付，TWC 花費了三倍的行政成本。這不是保護納稅人。這是在浪費納稅人的錢。GL 量化那個浪費。
你說保護弱勢群體、確保福利到達需要的人——好。但當 IDES 有 52 億美元欺詐而合法申請人在等待，系統設計讓錢流向了錯誤的地方。GL 量化那個漏損。
「GL 不判斷福利是否應該存在。
它只測量系統是否服務了它被建立的目的。」
即將於 2026 年 3 月 13 日發表。本文介紹 GL 框架在公共行政系統中的應用——論證制度失敗不是資源問題，而是分母設計問題。麻州 DUA 是核心案例。
申請預發表版本 →這是一個研究性對話，不是銷售電話。由我主動聯繫你。你決定是否繼續。
你的姓名、辦公室和所在州。一兩句說明你在系統中觀察到的現象。無需正式文件。
我會說明你的州的公開審計數據是否支持 GL 分析，以及這個分析能產生什麼結果。
確認方法論的適用性。分析工作和任何公開曝光由我承擔。底層審計記錄仍屬於你們。
政府审计办公室在体制内部运作。GL 在体制外部运作。
基于公开州审计记录的独立 GL 分析。
麻州花费数百万美元自动化其失业系统。上诉积压案件增长了九倍。准时付款率崩溃至 49%——全国最差。而这发生在新申请量处于六年低点之际。
GL 诊断：分母加速。速度被套用在一个破损的分类逻辑上，更快地产生错误、更快地导向上诉、更快地耗尽裁决者能量。
加州 EDD 自 2013 年 1 月以来几乎每个月都未达联邦及时性标准。43% 的上诉推翻率不是欺诈信号，而是摩擦信号。GL 诊断：系统花在处理合格申请人上诉的费用，比直接付款给他们还要多。返工是不可见的分母。
2015 年审计发现 31 个问题。2021 年审计发现 14 个问题仍未处理——包括 2014 年的处理错误。2.44 亿美元的五年现代化改造没有产生后续审计。数据就在那里。模式清晰可见。缺少的是一个独立的声音来说出数据所显示的事。
因为 GL 不设定目标。它只问：你说你会做 X。你做到了吗？效率如何？
你说小政府、低支出——好。但当一个合格申请人被错误拒绝、上诉、重新裁决、最终获得给付，TWC 花费了三倍的行政成本。这是在浪费纳税人的钱。GL 量化那个浪费。
你说保护弱势群体——好。但当 IDES 有 52 亿美元欺诈而合法申请人在等待，系统设计让钱流向了错误的地方。GL 量化那个漏损。
「GL 不判断福利是否应该存在。
它只测量系统是否服务了它被建立的目的。」
即将于 2026 年 3 月 13 日发表。本文介绍 GL 框架在公共行政系统中的应用——论证制度失败不是资源问题，而是分母设计问题。
申请预发表版本 →这是一个研究性对话，不是销售电话。由我主动联系你。你决定是否继续。
你的姓名、办公室和所在州。一两句说明你在系统中观察到的现象。无需正式文件。
我会说明你的州的公开审计数据是否支持 GL 分析，以及这个分析能产生什么结果。
确认方法论的适用性。分析工作和任何公开曝光由我承担。底层审计记录仍属于你们。