How to close gaps for any group, including underrepresented ones, in a way that is legally sound, broadly trusted, and built to last
One of the most consequential questions in workforce inclusion right now is not whether to address disparities. It is how.
The political and legal environment has shifted enough that programs designed to lift specific groups through restricted access are drawing scrutiny in ways they did not a few years ago. But the underlying gaps in promotion rates, attrition, leadership representation, and development access have not disappeared. They are still measurable. They still cost organizations real money and real talent. And the pressure to abandon the work entirely, while loud, is not supported by the law as courts have actually applied it.
What is needed is a framework that is honest about where the gaps are, disciplined about how to address them, and designed to hold up regardless of who is in the White House or what the EEOC says in a given news cycle.
That framework exists. Fair360 has been building toward it in our work on governance, measurement, and program design. And it was articulated with particular clarity at a July 2026 forum on the human-centered workplace hosted by NYU’s Meltzer Center for Diversity, Inclusion, and Belonging, where legal scholar and author Kenji Yoshino presented a framework that we think every HR and talent leader should understand: the shift from lifting to leveling, with triage as the organizing logic for where to focus first.
What lifting means, and why it is now legally complicated
Lifting strategies are programs designed to advance specific protected groups through restricted access. A fellowship limited to Black employees. A mentoring cohort restricted to women. A sponsorship program available only to employees from underrepresented racial groups.
These programs were designed with good intentions and often produced real results. The problem is that the legal and political landscape has shifted in ways that make restricted-access programs increasingly difficult to defend, particularly those that confer what Yoshino has called “palpable benefits” on the basis of protected characteristics. The Supreme Court’s decision in Students for Fair Admissions, while focused on higher education, has been interpreted by courts and agencies as a signal about preferences more broadly. Executive orders, DOJ settlements, and EEOC enforcement positions have all targeted programs that limit access by race or sex.
This does not mean lifting was wrong. It means the terrain has changed, and organizations that continue running restricted-access programs without legal review are carrying more risk than they may realize. The core argument Yoshino has made consistently is that programs conferring preferences based on protected characteristics will remain legally risky regardless of the political cycle, because the Supreme Court’s current composition is not temporary.
The answer is not to abandon the work. It is to redesign it.
What leveling means, and why it is both stronger and safer
Leveling strategies do not restrict access by protected characteristic. Instead they remove bias from the processes that determine who gets seen, who gets developed, and who gets advanced.
The case Yoshino and his colleagues at the Meltzer Center have built for leveling rests on two arguments that reinforce each other. First, it is far more difficult to legally challenge a program framed around making processes more rigorous, structured, and consistently applied than one framed around access for a specific group. Second, leveling reclaims the connection between inclusion and merit that some lifting strategies obscured, which makes the argument both easier to defend and easier to sustain across changing political environments.
In practice, leveling strategies include structured interviews with consistent, job-related criteria, calibration sessions that check for consistency across evaluators, transparent high-potential criteria that are documented and applied uniformly, mentoring and sponsorship programs open to all employees who meet defined development criteria, ERG structures that are open to everyone who wants to support the group’s purpose, and performance evaluation frameworks that reduce the influence of informal networks and subjective impressions.
These are not weaker versions of inclusion work. In many cases they are stronger, because they change the system rather than routing around it. And they are the programs that Fair360’s benchmarking data consistently shows are associated with better workforce outcomes across all employee groups, not just those the programs were originally designed to serve.
Where triage comes in
Here is the tension that many practitioners feel acutely: if programs have to be open to everyone, how do you still address the specific gaps that data shows are concentrated in particular communities?
The answer the Meltzer Center has developed is triage. The analogy they draw is to medicine: a physician goes wherever the illness is, not just to patients who are easy to treat, while still making judgments about where to deploy resources most urgently. Inclusion practitioners should apply the same discipline to workforce data.
Triage means using data to identify where the gaps are most acute, then designing open interventions targeted at the conditions, processes, and points in the talent pipeline where those gaps are being generated, without restricting access by protected characteristic.
In practice this looks like the following. If promotion data shows a gap at the manager-to-director transition, the triage response is not a restricted fellowship for underrepresented groups at that level. It is a structured review of the promotion process at that transition point: Are the criteria clear and documented? Are all eligible employees informed of their eligibility? Are calibration sessions being conducted consistently? Is sponsorship being distributed through informal networks that favor people who are already visible? Fixing those process failures benefits everyone in the pipeline, but because the gap was concentrated at that transition point for specific groups, those groups are likely to benefit most.
This is the core insight: well-designed leveling interventions, targeted at the right process failures through triage, can produce outcomes that are similar to or better than restricted lifting programs, while being far more legally defensible and far more broadly trusted by employees across demographic groups.
The data argument: why you cannot triage what you cannot see
There is a temptation in the current environment to stop collecting demographic data as a way of reducing legal exposure. That instinct, while understandable, is exactly backwards.
The panel at the NYU forum was emphatic on this point. Stopping data collection leaves organizations exposed to traditional discrimination claims from employees and regulators. State laws in California, Colorado, and elsewhere still require demographic reporting regardless of what happens to federal EO-1 requirements. And without data, triage is impossible: you cannot identify where the gaps are, you cannot design targeted interventions, and you cannot measure whether those interventions are working.
The more sophisticated approach involves what practitioners have described as mature privacy protocols: individual-level collection with aggregate-only access. This means collecting the data you need to do the work, restricting who can see individual-level data, and using aggregate analysis to identify process failures and measure outcomes. That approach is lawful, well-established, and gives organizations both the protection they need and the intelligence they need to act.
Fair360’s own benchmarking model is built on exactly this logic. The data from 150-plus organizations in our dataset does not expose individual employees. It enables organizations to see where they stand, where the gaps are, and whether their programs are producing the outcomes they are designed to produce. That is what makes triage possible at scale.
How this applies to mentoring and sponsorship specifically
Mentoring and sponsorship are two of the places where the lifting versus leveling distinction plays out most concretely, and where the redesign opportunity is clearest.
A restricted mentoring cohort for employees from specific racial groups is a lifting strategy. It may have produced real development outcomes, but it is now legally and politically vulnerable. The leveling equivalent is a mentoring program with open eligibility, clear matching criteria tied to development goals and career stage, and outcome tracking that shows whether participants are advancing at higher rates. If triage shows that certain groups are underrepresented in the program despite open eligibility, the question becomes why: Are they less likely to be nominated? Are they less connected to the informal networks that generate nominations? Are managers in certain parts of the organization less likely to encourage participation? Those are process failures that can be fixed without restricting or expanding access by protected characteristic.
The same logic applies to sponsorship, which Fair360 has written about extensively. Informal sponsorship, where advocacy flows through personal relationships and existing networks, is one of the most reliable ways to reproduce existing power structures while calling it development. Structured sponsorship programs with open, documented criteria are a leveling strategy: they make the advocacy process visible, consistent, and based on contribution and potential rather than who already knows whom. When triage shows that sponsorship relationships are clustering too narrowly, the answer is better design, not restricted access.
Our pieces on how to build a mentoring program that drives promotions, what high-performing sponsorship programs measure, and how to avoid favoritism and opacity in sponsorship systems all reflect this leveling logic, even if we did not use that specific vocabulary before. The framework the Meltzer Center has articulated is one we have been building toward in our program design and governance work.
The cohort-to-content and cohort-to-character moves
Yoshino and his colleagues have developed two additional frameworks that give practitioners specific tools for preserving mission without restricting access, and both are worth understanding in practical terms.
The first is cohort to content. Rather than organizing a program around who can participate, organize it around what the program is trying to accomplish. A program whose mission is advancing racial equity or closing the gender gap in leadership can remain open to all participants, as long as the mission is genuine and not a pretext for restricted access. An ERG whose purpose is advancing racial equity can welcome allies, sponsors, and employees from any background. A mentoring cohort focused on leadership development for employees who have faced systemic barriers can define eligibility around demonstrated barriers rather than protected characteristics. The mission remains. The access broadens.
The second is cohort to character. Rather than having the program itself carry the weight of protected-group identity, individuals can surface their own experience through their own narratives. An employee who is the first in their family to hold a professional role, or who faced documented obstacles to advancement, can bring that context into a sponsorship or development conversation without the program itself being restricted by protected class. This preserves the human reality of unequal experience without creating organizational liability.
What this looks like in practice: a triage-based approach
For organizations trying to put this framework into practice, here is a starting point.
Map the gaps first. Use workforce data to identify where disparities exist in promotion rates, attrition, high-potential designation, development program participation, and leadership representation. Be specific: which transition points, which functions, which manager cohorts are generating the largest gaps. That is your triage map.
Identify the process failures behind the gaps. For each gap, ask what process is generating it. Unclear promotion criteria? Inconsistent calibration? Informal nomination processes that favor the already-visible? Sponsorship relationships concentrated in certain networks? The gap is the symptom. The process failure is what needs to be fixed.
Design open interventions targeted at the process failures. Structured interviews. Documented promotion criteria. Calibration sessions with demographic review built in. Mentoring programs with transparent matching. Sponsorship programs with defined criteria and outcome tracking. ERGs with open participation and clear governance. These are leveling strategies. They fix the system.
Measure outcomes and adjust. Track whether the interventions are closing the gaps. If they are not, go back to the triage map and look for process failures you missed. The data is not just a compliance tool. It is how you know whether the work is working.
Involve legal from the start. Program design, eligibility criteria, and data use all benefit from legal review before implementation rather than after a concern arises. This is especially true for any program that touches promotion, compensation, or employment decisions.
Why this framework is built to last
One of the central questions at the NYU forum was whether inclusion strategies are built to last a change of administration or only built to survive the current moment.
The lifting-to-leveling framework, anchored in triage and measurement, is the closest thing available to an evergreen approach. It does not depend on a favorable executive branch. It does not require a particular EEOC enforcement posture. It is grounded in process fairness, outcome measurement, and the long-standing principle that employment decisions should be job-related and merit-based rather than driven by arbitrary factors. Courts have consistently upheld that principle regardless of which party controls the executive branch.
It is also, importantly, the approach most likely to build and sustain the broad coalition that makes inclusion work durable. As Fair360’s own series on rebuilding inclusion for everyone has argued, programs that employees across demographic groups can see as fair to them are programs that survive political pressure. Programs that feel exclusive, even when well-intentioned, are the ones that become targets.
Leveling, done well, is the argument that inclusion and merit are the same thing. That is not a retreat from the work. It is the strongest version of it.
A brief self-check
Before closing, leaders and practitioners may want to ask: do we know where our gaps are, at which transition points and in which functions? Do we know what process failures are generating those gaps? Are our current programs designed around open access and documented criteria, or do they restrict participation in ways that now carry legal risk? Are we collecting and protecting the demographic data we need to triage effectively? And when we talk about these programs internally, are we making the case in merit and process terms, or in terms that invite the preference critique?
Closing
The goal of addressing workforce disparities has not changed. The tools that are legally sound, broadly trusted, and built to last have. Lifting strategies that served organizations well in a different environment are now carrying more risk than most leaders realize. Leveling strategies that fix the underlying processes, targeted through triage at the places where gaps are most acute, are both more defensible and, in many cases, more effective.
Kenji Yoshino and his colleagues at the NYU Meltzer Center for Diversity, Inclusion, and Belonging have developed this framework with precision and legal grounding. Fair360’s role is to help organizations put it into practice: measuring where the gaps are, designing programs that hold up, and building the governance infrastructure that makes this work sustainable regardless of what the political weather does next.
That is the work. And it is worth doing right.
Further reading from Fair360:
- How to Build a Mentoring Program That Drives Promotions
- What High-Performing Sponsorship Programs Measure Beyond Participation
- How to Avoid Favoritism and Opacity in Sponsorship Systems
- How to Identify High Potential Talent Without Bias
- How to Keep ERGs Open, Lawful, and Business-Relevant
- We Lost the Thread: What DEI Got Wrong, Who It Left Behind, and What Comes Next
