We Demoted Wage Progression. The Index Is Better for It.

Last week, I changed my own instrument.

The Durability Index measures household stability across five domains. Until July 23, one of those domains was wage progression. It is not anymore.

Wage progression now lives one level down, as a sub-measure inside employment retention. Its old seat at the domain table belongs to Family Connection.

Here's the pattern across fifty years of workforce measurement: the instruments never change. They calcify. This issue is about why mine did, and what the change reveals about everything the field still gets wrong.

A Raise Is Not a Result

The field treats a raise as an unambiguous win. Wages went up, so the story is progress, so the file closes on a high note.

But households do not experience raises as arithmetic. They experience them as tradeoffs.

A raise that comes with a schedule change can cost more in child care than it delivers in pay. A raise that crosses a benefits threshold can leave a family with less total resources than before. A raise built on overtime can buy income with the very hours that held the household together.

Picture it in one composite household, the Thompsons. Lamar earns a promotion that moves him to a rotating schedule. His pay rises. But the rotation breaks the child care arrangement that Keisha built around her own shifts, so she cuts her hours to cover the gaps with Jaylen and Nia.

The program that placed Lamar records a wage gain and closes the file as a success. The Thompson household, measured as a household, just lost ground: less combined income, a strained care arrangement, and two parents with less margin than they had before the promotion.

No one in that story did anything wrong. The failure is architectural: the instrument saw one pay stub and called it the whole family.

When the instrument promotes wage progression to a headline domain, it teaches every program using it to chase raises. And chasing raises, in a system full of cliffs and tradeoffs, can destabilize the households the instrument exists to protect.

The Design Decision

The revised Durability Index carries five domains: Employment Retention, Housing Stability, Financial Resilience, Family Connection, and Justice-System Stability. Twenty sub-measures underneath, four per domain. One composite score from 5 to 25, administered at five points: intake baseline, then months 6, 12, 18, and 24.

Wage progression did not disappear. It became Sub-measure 1.2, inside Employment Retention, where it belongs: one signal among several about whether work is holding, not a headline about whether a family is.

Family Connection took the domain seat because the design logic of two-generation work points one direction. Household coherence predicts durability. A family that holds together absorbs shocks that no raise can absorb for them.

Look at what the promotion changes. Wage progression as a domain asks: did the paycheck grow? Family Connection as a domain asks: is the household still holding the routines, the care arrangements, and the mutual support that let any paycheck matter?

The second question is harder to measure. It is also the one that predicts whether the worker is still employed at month eighteen, because the household is where retention is actually decided. Month eighteen is deliberate: far enough past placement to see whether stability held, too deep for a ninety-day check to fake.

Old instrument: five domains, one of them a wage line that programs could game with overtime and job-hopping. New instrument: five domains, every one of them a condition of the household rather than a line on a pay stub.

Readers of LESI will notice its printed instrument reflects the earlier architecture. The revision you are reading about lands fully in Book Three, and LESI's revised edition will carry it home.

Why Architects Publish Their Changelogs

The field is full of frameworks that arrive finished, polished, and frozen. Nobody shows their revisions, because revisions look like admissions.

I think the opposite. An instrument that never changes is an instrument nobody is checking against reality.

A frozen instrument is a snapshot of what its designer believed on the day it shipped, preserved against every lesson the field learned afterward. And year by year, the gap between what it measures and what matters grows in silence.

The $107,000 household stability threshold for a family of four did not come from wishing. It came from research, and research keeps moving. A measurement architecture either moves with the evidence or becomes one more compliance ritual.

I earned two degrees in a place built on the assumption that I could not change. I do not get to build an instrument that refuses to.

So this is the standard I hold: when the evidence says the instrument is wrong, the instrument changes, and the change is published with its reasoning. That is what separates architecture from branding.

There is a second reason to publish the changelog. The moment the reasoning is public, every practitioner using the instrument can check the logic against their own caseload. Revision stops being the architect's private act and becomes the field's shared discipline.

For Funders

Ask every program you fund one question this quarter: when did your outcome instrument last change, and why? A portfolio measured by instruments nobody revises is a portfolio measured by habit.

Then ask what their instrument would say about a raise that costs a family its child care arrangement. If the instrument calls that a win, the instrument is scoring pay stubs, not households.

Your ninety-day move: pick your three largest grants and request each program's current instrument with its revision history. No history means no revision, and no revision means the instrument has never met reality. Make the next renewal conversation about what the instrument misses, not just what it reports.

For Policymakers

Wage records are the easiest data you can get, which is why they dominate public reporting. Easy is not the same as true.

Median earnings tell you what a program's completers were paid. They do not tell you whether the households behind those wages were stable at month eighteen. Fund the measurement infrastructure that can answer the second question across the 24-month administration schedule, and the first question will finally have context. Durability across 24 months versus compliance at day 90: that is the choice your reporting requirements are already making.

Your ninety-day move: identify one contract up for renewal and add household stability reporting alongside the wage measures it already requires. Do not replace the wage data. Pair it, so every earnings figure arrives with the household context that tells you what it cost.

For Operators

Start this quarter. Pull your last twenty wage-gain outcomes and trace what happened to each household in the six months after the raise.

Count the schedule disruptions, the benefits cliffs crossed, the child care arrangements that broke. If even three of twenty raises destabilized the family behind them, your headline metric is hiding your true results.

Assign the audit an owner, give it thirty days, and put the findings in front of your leadership with one question attached: which of our celebrated outcomes would survive a household-level look? The answer becomes your case for measuring differently, in your funders' language, from your own files.

What Changes When the Metric Changes

We measure raises. We optimize for raises. We get raises, and we call the churn beneath them personal failure.

We measure households. We optimize for households. We get durability, and the wages follow on ground that can hold them.

That is the whole argument of the revision, and it is the whole argument of this newsletter. The unit of change is the household. Every instrument that forgets this will keep producing wins that families experience as losses.

Next week: who measures the programs themselves. For fifty years, the instruments have pointed at justice-impacted individuals. It is time one pointed back.

Take the free Durability Index Self-Assessment. Twenty questions. Five domains. One score.

Take the Assessment →

Until next time, keep building what they said couldn't be built.

Khalil Osiris

Author & Founder, Khalil Osiris Consulting | Market Architect, 2Gen Economy Workforce Ecosystem | Fair-Chance Hiring · Household Stability · Workforce Durability | Publisher, The Durability Economy

Subscribe to The Durability Economy for workforce redesign, fair-chance hiring, and household stability. For leaders who measure what lasts.

Change the metrics. Change the outcomes.