When Governor Josh Shapiro took office in 2023, Pennsylvania had a reputation few governors would want: a state where projects went to wait.

By late 2023, the Department of Environmental Protection (DEP) counted a backlog of more than 2,400 permit applications. Applicants often couldn’t tell where their permit stood, how long it would take, or whom to call.

Shapiro made fixing that a signature agenda item.

In his first month, he ordered every state agency to catalog its permits and commit to a processing deadline. Over the next three years, his administration and a divided legislature added a money-back guarantee, a paid third-party review option, a concierge program for major projects, and statutory deadlines after which some environmental permits are approved automatically.

“Government at the speed of business” became a slogan, then a record, and eventually a national brand. DEP says the backlog hit zero in October 2025, and the administration says average processing time fell from 53 days in 2022 to 38 days in 2025. Then, in August 2026, Shapiro pulled data centers, the most visible projects his reforms had courted, out of his own Fast Track program.

Pennsylvania has shown that Democratic leadership can make the permitting bureaucracy measurably faster. It has also shown that faster was never the only thing people wanted from it.

Pennsylvania’s Permitting Problem Was Really Three Problems

It’s worth being precise about what was broken, because each of Shapiro’s reforms targeted a different failure.

1. The first was volume. Applications piled up faster than DEP could review them, and nothing forced the department to prioritize.

2. The second was uncertainty. Agencies had no public timelines. An applicant waiting four months had no way to know whether that was normal, slow, or stuck.

3. The third was coordination. A major project needing permits from DEP, PennDOT, and local authorities had no single point of contact and no one responsible for moving the whole package.

A reform aimed at one of these problems won’t necessarily solve the others. That distinction turns out to matter a great deal.

What Did Pennsylvania Actually Build?

The reforms came in four layers.

1. PAyback (2023): Executive Order 2023-07 directed every agency to document each permit type, its legal basis, its fee, and a target processing time, and to publish them. If an agency misses its own target without an extension, the applicant gets the fee back.

2. SPEED (2024–25): The Streamlining Permits for Economic Expansion and Development (SPEED) program lets applicants pay a DEP-approved outside professional (an engineer, surveyor, or environmental scientist meeting licensing and conflict-of-interest standards) to conduct the first technical review of eligible permits. DEP still issues the final decision. The applicant pays both the state fee and the reviewer.

3. PA Permit Fast Track (2024): Executive Order 2024-04 created a program, run by the new Office of Transformation and Opportunity, to coordinate review of major economic-development and infrastructure projects. Developers can ask to join, but the administration decides which projects get in. Participants get a coordinated permitting timetable, a public dashboard, and a single point of contact across agencies.

4. Deemed approval (2025): The 2025–26 budget, Act 45, expanded SPEED to more permit categories. It also set statutory clocks: 30 days for certain air-quality permits and 60 days for stormwater and groundwater permits. After that, a permit is automatically approved if DEP hasn’t acted or sought an extension.

This is not deregulation in the usual sense. The environmental standards largely stayed the same. What changed was the clock, the visibility, and who does the first read.

So, Is It Working?

Measured against the three problems above (volume, uncertainty, coordination), the record is strongest where the reforms were least glamorous. One asterisk also applies throughout: nearly every number comes from the administration itself, and none has been independently audited.

1. Volume: largely solved, by the state’s account. DEP says the backlog is gone, and the administration says average processing time fell from 53 days to 38. It reports similar gains elsewhere: business filings that averaged 14 days now take one, and new-teacher certification waits fell by about 10 weeks.

2. Uncertainty: better, but unproven. Published targets and statutory clocks now tell applicants what to expect. But since 2023 the state has issued just five PAyback refunds out of more than 40 million applications. Either agencies almost never miss their deadlines, or the deadlines were set loosely enough that almost nothing counts as late. Until the state shows how its targets compare with past processing times, there’s no way to know which.

3. Coordination: the weakest link. Fast Track, the most heavily branded reform, was built for this problem. But a September 2026 analysis by the conservative outlet Broad + Liberty, republished by PoliticsPA, found nothing in the executive order that lets a Fast Track project finish ahead of a comparable project outside it. The administration cites wins, like a Lackawanna County project it says saved 18 months, but what the program guarantees is coordination, not acceleration.

Key Policy Takeaway — The Boring Parts Did the Work

If Pennsylvania’s reforms delivered, the credit belongs to the least glamorous pieces: published deadlines, backlog triage, and statutory clocks.

A target posted publicly creates internal pressure. A deemed-approval deadline creates legal pressure. Clearing a backlog frees staff to work on new applications instead of old ones.

The flagship program and the effective machinery aren’t the same thing. States copying Pennsylvania should prioritize copying the machinery.

The Strongest Critique Is About Who Gets the Fast Lane

The most serious objection to Pennsylvania’s reforms isn’t that they made government faster. It’s that speed became something the state handed out, and not always by rules the public could see.

Some applicants can buy it. SPEED gives applicants who can afford an outside reviewer a faster first review. Environmental groups, including PennFuture, warn that DEP staff will feel pressure to defer to a paid recommendation rather than re-examine it. DEP calls the program “staff augmentation,” extra capacity rather than weaker review. Whether that holds depends on how often DEP reviewers actually push back, and no public data shows how often they do.

One applicant got it privately. Internal emails reported by Heatmap News show that in April 2025, Shapiro’s office offered Amazon “exclusive early access” to SPEED. The offer came two months before SPEED opened to the public, and before Amazon announced a $20 billion data-center investment that Shapiro called the largest private investment in state history. The memo was marked as subject to a nondisclosure agreement. Whether that access sped up a single permit matters less than the signal: reforms sold on transparency gave their most powerful applicant a side door.

The public absorbs the risk when the clock runs out. Deemed approval works when DEP is staffed to meet its deadlines. If it isn’t, because of a budget squeeze, a hiring freeze, or a flood of applications, the default outcome is approval without full review. That trade-off may be defensible. It should be made openly.

And communities got a say last. Data centers are where these critiques converge. By July 2026, 74% of Pennsylvania voters opposed an AI data center in their own community, including 82% of Democrats, according to Quinnipiac. In August, Shapiro pulled every data-center project out of Fast Track, banned nondisclosure agreements with developers, and barred DEP from issuing permits without local approval. Because Fast Track never guaranteed faster permits, the move was largely symbolic. But the symbolism was the point: the fight was never about permitting efficiency. It was about who gets a say before the process starts.

The Blueprint for Other States

Pennsylvania’s experience offers four lessons.

1. Start with deadlines and publication. Public target times, backlog triage, and statutory clocks are the unglamorous core of permitting reform. They’re also the part most likely to be doing the work.

2. Make the guarantee verifiable. A money-back guarantee that pays out five times in three years proves nothing on its own. Publish the historical baseline behind each target, and let an auditor, a university, or the legislature’s fiscal office check the results.

3. Build the fast lane in the open. If a state offers expedited coordination, admission criteria should be public, access should be open, and there should be no side doors or nondisclosure agreements. The Amazon emails did more damage to the reforms’ credibility than any processing statistic could repair.

4. Separate speed from consent. Permitting reform answers “how fast?” It doesn’t answer “should this be built here?” States that don’t settle the second question upstream, through siting standards, community-benefit rules, and local input, will find it resurfacing as a revolt against the first.

Bottom Line

Pennsylvania has built something Democrats should want to copy.

A faster, more transparent permitting system isn’t a concession to business. It’s a precondition for building the housing, transmission, and clean energy that a progressive agenda depends on.

But speed is a means, not a mandate. The data-center fight revealed what the efficiency agenda had skipped: whether communities had been asked, and whether every applicant was playing by the same rules.

Pennsylvania learned that a state can process a permit in 38 days and still not have permission. Other states should copy its deadlines, and build their fast lanes in public.

Mississippi’s reading gains complicate the usual partisan story about public education.

A Republican-led state took responsibility for a stubborn statewide problem, treated implementation as the state’s job, and stayed with the effort long enough to produce measurable improvement. That is why Mississippi can serve as a Red State Lesson when it comes to education reform.

Democrats do not need to embrace every part of the Mississippi model to confront the challenge it presents: ambitious policy is only as progressive as the public institutions capable of delivering it.

The task is to borrow that governing seriousness without importing punitive features or claiming more than the evidence supports.

The Mississippi Context

Before reform, Mississippi routinely appeared near the bottom of national education rankings. Low performance reflected poverty, unequal opportunity, weak early-learning access, inconsistent instructional practice, and a state system that had not built the capacity to improve teaching at scale.

The policy response developed over years. Mississippi raised academic standards, strengthened its school accountability system, invested in early childhood education, and built a more active state education department. The 2013 Literacy-Based Promotion Act sat inside that larger structure.

That context matters because “science of reading” laws can sound deceptively simple. Legislatures can require evidence-based instruction – they cannot create thousands of teachers who understand how to deliver it by changing statutory language alone.

Importantly, Mississippi invested in implementation.

The state trained educators, deployed literacy coaches—initially concentrating them in low-performing schools—approved screening tools, produced instructional guidance, and required schools to communicate with parents. The third-grade promotion rule created urgency, but the state also built supports intended to prevent children from reaching the gate unprepared.

The politically visible reform, a “3rd Grade Gate,” was a mandate, the administratively important reform was a delivery system.

How the Mississippi Model Works

The Literacy-Based Promotion Act requires schools to screen students in kindergarten through third grade for reading deficiencies and notify parents when a child is substantially behind. Students receive an individual reading plan and targeted intervention.

Teachers and school leaders receive training aligned with evidence about how children learn to read, including explicit instruction in foundational skills. State literacy coaches work with schools on classroom practice, data, and instructional planning.

The most controversial provision, the “3rd Grade Gate,” arrives at the end of 3rd grade when students generally must demonstrate a minimum level of reading proficiency through the state assessment or an approved alternative to advance. The law provides limited “good cause” exemptions, including for some English learners, students with significant disabilities, and students who demonstrate proficiency through an alternative route.

Students who are retained are supposed to receive intensive intervention rather than repeat the identical year. The theory is that fourth grade marks a shift from learning to read toward using reading to learn across subjects. Proponents argue that promoting a child without foundational skills can turn one year of difficulty into years of accumulated academic failure.

The counterargument is equally serious: retention imposes social and emotional costs, affects older and disadvantaged students disproportionately, and can become a substitute for adequate support.

Mississippi’s model rests on the claim that retention is a last-resort accountability measure within a much broader literacy system—not the system itself.

What Mississippi Got Right

Mississippi made the state responsible for more than issuing rules.

It created common expectations, screened children repeatedly, gave parents clearer information, trained teachers, sent coaches into classrooms, and built a state department capable of monitoring implementation. It then sustained the work across political cycles.

This is a useful lesson for progressive education policy reformers because progressive policymaking sometimes places great faith in the ambition of a statute while paying less attention to the public machinery needed to carry it out. Mississippi’s approach recognized that implementation is itself a policy choice.

The sequence also matters. A state should not impose a high-stakes consequence on a child before it can show that the school provided high-quality curriculum, trained instruction, early identification, and meaningful intervention.

Accountability without capacity punishes. Capacity without accountability can dissipate. Mississippi tried to build both and states looking to build off their successes should do the same.

What the Evidence Says and Does Not Say

On the measure most closely associated with the education reform, Mississippi produced tangible results.

Its fourth-grade reading scores improved substantially over the decade following enactment while national performance stagnated or declined. The gains reached beyond a single district and appeared among historically underserved students as well as the statewide average.

But attribution requires discipline.

The literacy law coincided with stronger standards, accountability changes, early-childhood investments, and a sustained administrative campaign. The third-grade retention rule can also change the composition of the tested fourth-grade population: some struggling students remain in third grade and take the fourth-grade assessment later.

Researchers disagree about how much that mechanical effect explains. Emerging work suggests retention accounts for some, but not most, of Mississippi’s relative gains. The results also weaken as students age. Mississippi’s eighth-grade reading performance has not produced the same celebrated national standing as fourth grade. That could mean early gains fade, older-grade instruction does not build on them, or the fourth-grade measure overstates the durable effect.

This is not evidence that nothing improved – it is evidence that early literacy is a foundation, not a complete K–12 strategy.

What to Borrow from the Mississippi Model

Policy designs and principles to borrow from Mississippi model:

1. A state role in implementation. District autonomy should not mean handing schools a vendor list and walking away. States can build expertise, vet materials, train teachers, and target coaching where local capacity is thinnest.

2. Intervention before crisis. The promotion decision should never be the first moment a family learns that a child is struggling. Regular screening should trigger timely, specific support.

3. Policy alignment. Curriculum, teacher preparation, coaching, assessment, family communication, and accountability should reinforce the same instructional strategy.

4. Staying power. Mississippi treated early literacy as sustained governing work, not a one-session legislative achievement. The gains developed over years and across political cycles.

5. A willingness to measure results. The state set a legible goal and made performance visible. Democrats should pair that focus with a broader dashboard extending through middle school and graduation.

The broader lesson is that Mississippi’s model works as a system, not as a menu. States cannot separate the visible mandate from the less visible public capacity that gives it a chance to succeed.

A third-grade gate without early intervention, trained teachers, and sustained support shifts responsibility for institutional failures onto children.

Democratic education reformers across the country should borrow Mississippi’s seriousness about implementation while rejecting the idea that consequences can substitute for capacity. The question for other states is not whether to copy Mississippi, but whether they are prepared to build—and improve upon—the conditions that made its gains possible.

The Adaptation Test

Before borrowing from Mississippi, policymakers should be able to answer five questions:

1. Capacity: Has the state funded teacher training, coaching, strong materials, screening, and intervention—or merely required them?

2. Timing: Will children receive help well before a third-grade promotion decision?

3. Fairness: Are exemptions, appeals, accommodations, and family communication clear and accessible?

4. Continuity: Does the literacy strategy continue into later grades, when students must use reading to learn increasingly complex material?

5. Evidence: Will the state track not only fourth-grade scores, but also repeated retention, attendance, special-education identification, discipline, middle-school achievement, and graduation?

If the answer to those questions is no, a state is not adapting Mississippi’s system – it is borrowing its slogan.

Red State Takeaway

Mississippi did not discover a miracle. It built a system.

The state set a clear goal, aligned policy around it, invested in educators, monitored children early, involved parents, and maintained pressure to improve. The “3rd Grade Gate” was part of that system, but it was not a magic ingredient.

Democratic policymakers should be willing to learn from a Republican-led state that produced meaningful gains. They should also improve on the model: fund the supports fully, protect children from crude high-stakes consequences, measure outcomes beyond a single test, and connect literacy reform to the conditions that shape learning outside the classroom.

The lesson from Mississippi is not that government should become more punitive. It is that government must become more capable.

Washington did not merely pass an ambitious climate law. It put that law before voters after fuel prices rose—and won.

The 2021 Climate Commitment Act (CCA) created a statewide cap-and-invest system. Large emitters must hold an allowance for each metric ton of covered greenhouse-gas emissions. The total supply of allowances declines over time, while auction revenue finances transportation, clean energy, environmental justice, tribal projects, wildfire resilience, building efficiency, and assistance for households.

The first auctions began in 2023 and generated far more revenue than early budget writers expected. They also intensified a basic political problem: when a carbon price flows through fuel and other markets, voters can see the cost immediately while the climate benefit arrives gradually and the public investment is spread across hundreds of projects.

Opponents responded with Initiative 2117, a 2024 ballot measure that would have repealed the cap-and-invest program and barred agencies from establishing a similar system. Washington voters rejected it by a decisive margin.

That vote made the state’s experiment nationally important.

While Washington demonstrated that cap-and-invest can survive direct democratic review, it has not yet demonstrated that the program will deliver emissions reductions efficiently, protect households from regressive costs, and spend billions with enough discipline to retain public trust.

From Climate Target to Carbon Market

Before the Climate Commitment Act, Washington had statutory emissions goals and a growing set of sector-specific policies. What it lacked was an economy-wide mechanism that forced covered emissions downward while producing durable revenue for transition.

Washington enacted the CCA in 2021, when Democrats controlled the governor’s office and both chambers of the state Legislature. The program covers large facilities, fuel suppliers, natural-gas utilities, and other entities meeting statutory thresholds. Electricity imported into the state is also addressed, with rules intended to limit leakage.

The law’s politics depended on reinvestment. Cap-and-invest was not sold simply as a price signal. Auction revenue would fund the infrastructure and household transition needed to make lower emissions possible.

It also embedded environmental-justice requirements. At least 35% of specified investments must provide direct and meaningful benefits to vulnerable populations within overburdened communities, with a target of 40%. Additional provisions direct resources and consultation toward tribes.

What Does the Climate Commitment Act Do?

Under the CCA, the Department of Ecology sets an emissions cap and issues a limited number of allowances.

Each allowance authorizes its holder to emit one metric ton of greenhouse gases covered by the program. Covered businesses obtain allowances through quarterly auctions, secondary markets, or limited free allocation.

The auctions operate differently from a traditional highest-bidder-wins sale. Participants submit sealed bids specifying how many allowances they want and what they are willing to pay. Ecology ranks those bids from highest to lowest and distributes allowances until the available supply is exhausted. The lowest successful bid becomes the settlement price paid by every winning bidder.

Companies that do not obtain enough allowances at auction can purchase them from other market participants—or reduce their emissions and lower the number they need. This structure turns the state’s emissions limit into a market price.

As Washington reduces the number of available allowances over time, continued pollution should become more expensive, giving businesses a financial incentive to invest in cleaner operations. Meanwhile, auction proceeds flow into state accounts where the Legislature appropriates them for transportation electrification, public transit, building efficiency, clean-energy development, air-quality improvement, natural and working lands, climate resilience, and other authorized uses under the CCA.

Key Policy Takeaway — Carbon Pricing Needs a Public-Benefit Strategy

Economists often describe carbon pricing as elegant: make pollution more expensive and allow the market to find the cheapest reductions. But politics is less elegant.

Consumers do not experience an allowance auction. They experience gasoline, heating, food, and freight prices. If the state cannot show where the money goes and who benefits, cap-and-invest can look like a complicated tax whose environmental result is impossible to see.

Washington’s most replicable choice was to treat investment as integral to the policy.

By fiscal year 2025, the state reported that 61% of initial spending benefited vulnerable populations in overburdened communities—above the 35% statutory minimum. Projects included energy-bill assistance, public and active transportation, clean energy, habitat, and wildfire resilience.

That does not by itself prove the investments are effective. However, it does make the distributional promise measurable.

So, Is It Working?

Financially and politically, the program has cleared two major tests while another significant test remains.

Auctions have generated more than $4.5 billion since the program began. The revenue has moved cap-and-invest from an abstract climate instrument into a major part of the state budget.

Then voters considered repeal. Initiative 2117 lost with roughly 62% voting no. The result did not prove that every voter supports carbon pricing. The campaign also focused on the transportation, clean-air, and community investments that repeal would eliminate. But the outcome showed that opponents could not convert anger over fuel costs into a statewide majority for dismantling the law.

The environmental test will take longer. Covered entities operate within multiyear compliance periods, allowance banking affects when reductions occur, and statewide emissions data arrive with a lag. Revenue is not the same as reduced pollution. Nor does selling every allowance demonstrate that each funded project is cost-effective.

Accordingly, one would be wise to resist declaring victory from Washington’s auction totals alone. The central outcome is the emissions trajectory, followed by local air quality and the household cost of achieving it.

The Strongest Critique Is About Cost and Accountability

The case against the CCA begins with fuel prices.

Suppliers can pass allowance costs to consumers. The precise effect is contested, but pretending there is no effect undermines trust. Carbon pricing is designed to make high-emission activity more expensive. The relevant questions are how much costs rise, who bears them, and whether investments create affordable alternatives.

Critics also argue that unexpectedly high revenue can weaken spending discipline. When billions are distributed across agencies and legislative accounts, projects may be branded as climate investments without demonstrating material emissions reductions or resilience benefits.

Free allowances create another tension. They can protect jobs and prevent emissions leakage, but they also reduce the immediate incentive for covered industries. Market linkage may reduce costs while making Washington’s system harder for voters to understand.

Finally, a cap can be weakened through excessive offsets, reserve allowances, exemptions, or accounting choices. The law’s credibility depends on the integrity of the declining limit, not the sophistication of the auction platform.

The Blueprint for Other States

Washington and the CCA offer six lessons.

1. Legislate the emissions cap, not only the revenue mechanism. The environmental obligation must remain clear when auction receipts fluctuate.

2. Make household and community benefits visible. Bill assistance, transit, building upgrades, and local air-quality investments should be traceable by place and population.

3. Report costs honestly. States should publish independent estimates of consumer incidence rather than dismissing every price effect or accepting the largest industry claim.

4. Build spending discipline before revenue arrives. Every appropriation should identify an emissions, resilience, affordability, or environmental-justice outcome.

5. Protect trade-exposed industries conditionally. Free allowances should decline or change when leakage risk falls and should not become a permanent entitlement.

6. Prepare for repeal politics. Washington survived because supporters defended both the climate goal and the tangible investments financed by the program.

Bottom Line

Washington has built one of the most consequential state climate policies in the country.

The Climate Commitment Act has created a binding carbon market, raised billions, directed substantial resources toward overburdened communities, and survived an explicit statewide repeal vote. Those are real accomplishments.

But the program’s durability now creates a higher standard.

Washington must show that emissions are falling because of a credible cap, that investments are more than a list of worthy projects, and that working families receive affordable alternatives to the behavior the state is making more expensive.

The lesson for other states is not simply “put a price on carbon.” It is to build the political and administrative system around that price—and keep proving what the public receives in return.

For decades, Oregon has been one of America’s leading laboratories for state land-use policy. Now it is testing a more difficult proposition: if the state is willing to tell cities where housing should be allowed, should it also help them build the infrastructure—and enforce the rules when they do not comply?

That is the bet behind Senate Bill 1537, Governor Tina Kotek’s 2024 housing production package.

The law does not rely on one grand solution. It creates a Housing Accountability and Production Office, gives qualifying projects temporary flexibility from some local development standards, makes certain housing decisions faster, opens a limited route for cities to expand urban growth boundaries, and commits state resources to infrastructure and affordable housing.

In other words, Oregon is trying to address the entire gap between zoning a home and completing one.

The early evidence is necessarily incomplete. Several major provisions took effect only in 2025. Oregon is still producing far fewer homes than Kotek’s target, and expensive credit and construction costs are suppressing projects that have little to do with local permitting. But the design of SB 1537 already offers a useful blueprint: housing mandates work better when the state pairs enforcement with technical assistance and money.

It also offers a warning. Government can make a home legal and easier to permit. It cannot make an uneconomic project finance itself.

Oregon’s Housing System Stopped Producing Enough Homes

Oregon entered the decade with an unusually strong statewide planning system and an increasingly severe housing shortage.

The state had already legalized duplexes and other “middle housing” in many communities and required larger cities to plan for housing needs. But legal capacity did not reliably become completed housing. Projects could still encounter local design standards, infrastructure constraints, uncertain interpretation of state law, and lengthy permitting. Meanwhile, high interest rates and rising material, labor, and insurance costs weakened the financial case for construction.

Kotek took office in 2023 and set a target of 36,000 new homes per year. The target was intentionally ambitious: Oregon needed to increase production substantially, not merely return to a weak baseline.

SB 1537 followed a failed 2023 attempt to pass some of the same ideas. The earlier proposal drew opposition from local governments and environmental and land-use advocates, particularly over required adjustments to local rules. The 2024 package narrowed and revised those provisions, added infrastructure support, and preserved limits around Oregon’s urban growth system.

That legislative history matters. Oregon did not abandon statewide land-use planning to pursue supply. It tried to make that system deliver more housing.

What Does SB 1537 Do?

The law’s most important institutional change is the Housing Accountability and Production Office, or HAPO.

The office is jointly operated by the Department of Land Conservation and Development and the state Building Codes Division. It helps local governments and developers interpret housing, land-use, building-code, and permitting requirements. It can coordinate state agencies, investigate complaints, provide technical assistance, and take enforcement action when local governments violate state housing law. Its enforcement authority became operative July 1, 2025.

That combination is deliberate. A state office that only punishes cities can produce defensiveness and litigation. An office that only advises can be ignored. HAPO is designed to do both.

SB 1537 also requires local governments to grant qualifying housing projects adjustments from certain development and design standards. Those may include rules governing setbacks, lot coverage, building orientation, and some dimensional requirements. The policy does not erase building codes or every local standard. It creates a limited escape valve when specified rules make otherwise lawful housing difficult to build.

The law makes other procedural changes: applicants with pending projects may opt into updated local rules; certain housing applications receive a more streamlined form of land-use review; and prevailing parties in some housing appeals have broader access to attorney fees.

On land supply, qualifying cities receive a one-time, time-limited process to add or exchange a constrained amount of land at the urban edge, subject to affordability, planning, and review requirements. On infrastructure, the law creates support for water, sewer, transportation, and other investments needed to turn planned land into buildable sites. It also directs funding toward affordable and moderate-income housing.

This is not blanket deregulation. It is a state attempt to identify—and intervene at—several separate choke points.

Key Policy Takeaway — Enforcement and Capacity Belong Together

The strongest feature of Oregon’s approach is not any single zoning rule. It is the decision to combine accountability with implementation support.

States often issue housing targets or preempt a narrow local restriction, then leave cities and builders to navigate everything else. SB 1537 recognizes that housing can fail at multiple handoffs: a city may misunderstand state law, a building department may lack staff, an approved site may lack sewer capacity, or a legal project may be blocked by an inflexible design standard.

HAPO gives participants one place to surface those problems. Infrastructure funding gives cities a reason—and a practical ability—to resolve them. Enforcement creates consequences when the barrier is not capacity but resistance.

Other states should pay attention to that institutional architecture. The housing shortage is partly a rules problem, but it is also a delivery problem.

So, Is It Working?

It is too soon to answer confidently.

The mandatory-adjustment and streamlined-review provisions began operating in January 2025. HAPO’s enforcement powers followed in July. The office’s initial statutory reports are due to the Legislature in September 2026. A fair evaluation therefore has to distinguish implementation evidence from housing-market outcomes.

The immediate implementation evidence is clear: Oregon has created the office, issued detailed guidance, opened an inquiry and complaint process, and required local governments to apply the new adjustment rules. The machinery exists.

The production evidence is more sobering. Oregon remains far below its 36,000-home annual goal. Multifamily permitting fell sharply during the broader financing slowdown. Portland’s experience captures the mixed picture: middle-housing reforms have produced thousands of duplex, fourplex, accessory-dwelling, and cottage-cluster units since 2021, while overall housing production has recently weakened.

That does not establish that SB 1537 failed. Most of those outcomes reflect projects conceived or financed before its major provisions took effect. It does establish that procedural reform is operating against powerful economic headwinds.

The correct near-term test is not whether Oregon suddenly reaches 36,000 homes. It is whether comparable projects move through review faster, whether HAPO resolves recurring barriers, whether infrastructure awards unlock stalled sites, and whether the temporary adjustment process produces homes that would otherwise not pencil out.

The Strongest Critique Is That Oregon May Be Solving Yesterday’s Constraint

When financing was cheap and demand was strong, zoning and permitting delays were often the binding constraint. In a high-cost capital market, a project may be fully legal and promptly approved but still impossible to finance.

That is the strongest critique of SB 1537: Oregon may have built an elaborate system for speeding approvals at the moment when interest rates, construction costs, and weak project economics became more important.

Local governments raise a second objection. State mandates can force cities to process growth without enough staff or infrastructure revenue. Even with state funding, temporary grants may not cover the long-term operating and maintenance costs that new development creates.

Land-use advocates offer a third warning. Expedited urban growth boundary changes could weaken Oregon’s long-standing effort to limit sprawl, protect farmland, and direct development toward existing communities. The law includes limits, but implementation will determine whether those limits hold.

These critiques are not arguments for returning to discretionary delay. They are reasons to evaluate each part honestly. Faster permitting should be measured in time saved. Infrastructure funding should be measured in homes unlocked. Boundary expansions should be measured against affordability and environmental promises.

None should receive credit merely because the statute exists.

The Blueprint for Other States

Oregon’s model offers five lessons.

1. Create a housing referee with both assistance and enforcement power. State housing law is ineffective when no institution is responsible for resolving disputes quickly.

2. Treat permitting as an operating system. Deadlines, objective standards, appeals, staffing, and interagency coordination matter as much as the zoning map.

3. Pair local obligations with infrastructure support. States should not demand housing production while ignoring pipes, roads, schools, and utility capacity.

4. Make flexibility bounded and reviewable. Targeted adjustments can remove rules that unintentionally block housing without suspending health, safety, environmental, or fair-housing protections.

5. Separate policy performance from market conditions. A state should report approval times and resolved barriers even when interest rates suppress construction—and should not claim credit for units already in the pipeline.

Bottom Line

Oregon’s housing experiment is more sophisticated than the usual argument between local control and state preemption.

SB 1537 says the state can set expectations, help cities meet them, finance the infrastructure that growth requires, and enforce the law when cooperation fails. That is a credible governing model.

But it is not a housing-production machine. Oregon still has to prove that its new office can resolve real barriers, that infrastructure money reaches projects that will actually be built, and that regulatory flexibility produces homes without sacrificing the values embedded in the state’s land-use system.

Other states should copy Oregon’s willingness to own implementation. They should not copy its policies on faith.

The race to regulate artificial intelligence has officially left Washington.

While Congress continues debating how—or even whether—to establish a national framework for AI, Colorado has moved first. By passing the Colorado AI Act, state lawmakers have positioned Colorado as the nation’s first large-scale testing ground for comprehensive AI governance.

The law represents something larger than a single piece of technology legislation. It is an early glimpse into how Democratic policymakers may attempt to balance innovation, consumer protection, and economic competitiveness in the age of artificial intelligence.

If the law proves workable, don’t expect it to remain a Colorado experiment for long.

What’s happening

The Colorado AI Act establishes one of the country’s first broad regulatory frameworks governing “high-risk” artificial intelligence systems.

Rather than regulating every AI application equally, the law focuses on systems used to make consequential decisions affecting people’s lives, including:

• employment

• housing

• lending

• education

• healthcare

• insurance

Developers and deployers of those systems assume new responsibilities surrounding:

• risk management

• impact assessments

• transparency

• consumer notification

• discrimination mitigation

In short, Colorado is attempting to answer a question Washington has yet to resolve: how should the government oversee AI before significant harms occur?

Why it matters

Artificial intelligence increasingly touches nearly every aspect of modern life.

Companies now rely on AI to screen job applicants, approve loans, evaluate insurance claims, recommend medical treatment, and determine which advertisements consumers see. Those systems can create enormous efficiencies but they can also replicate bias, produce opaque decision-making, and generate outcomes that neither consumers nor regulators fully understand.

Colorado’s law reflects an emerging Democratic philosophy: innovation should continue—but not without guardrails.

Between the lines

The legislation also illustrates a broader governing trend.

For decades, states frequently waited for Congress to establish national regulatory standards. Increasingly, the opposite is occurring. States are becoming the first movers. California has long shaped national privacy debates. Colorado is now attempting to play a similar role in artificial intelligence.

If enough states adopt comparable frameworks, Congress may eventually face pressure to harmonize the rules nationally.

Can it scale?

Several Democratic-led states—including California, New York, Massachusetts, and Illinois—have all explored AI governance through legislation or executive action.

Each is likely watching Colorado closely.

The biggest implementation questions include:

• Can regulators enforce complex technical requirements?

• Will compliance costs discourage innovation?

• Can businesses operate efficiently under differing state standards?

• Will consumers actually receive greater protection?

Colorado is effectively becoming the country’s pilot program.

Bottom line

Every generation faces a technological revolution that outpaces existing law.

For previous generations, it was railroads, automobiles, telecommunications, or the internet. Today, it is artificial intelligence.

Colorado has become one of the first states willing to move from debating AI governance to actually building it.

Whether the Colorado AI Act ultimately succeeds or requires significant revision, one thing is already clear: the future of AI regulation is being shaped in state capitols before it reaches Capitol Hill.

Every September, politicians celebrate the workers who built America’s middle class. Michigan is attempting something more ambitious: rebuilding some of the institutions that helped workers build it.

For decades, that project appeared to be moving in the opposite direction but Michigan is leading the charge to change that momentum.

Michigan was once nearly synonymous with American organized labor. The sit-down strikes in Flint helped build the United Auto Workers. Detroit’s automakers helped turn collective bargaining into a vehicle for middle-class wages and benefits. At the end of the 1980s, more than one in four Michigan workers belonged to a union.

Then came decades of decline. By 2012, Michigan’s union membership rate had fallen to 16.6%. That same year, Republicans delivered another blow: making the birthplace of the modern American labor movement a right-to-work state.

Eleven years later, Democrats got the chance to reverse it.

After winning full control of state government for the first time in roughly four decades, Democrats and Governor Gretchen Whitmer repealed right-to-work. The change took effect in February 2024, making Michigan the first state in nearly six decades to repeal such a law. It was an extraordinary reversal of a conservative policy project that, for decades, had moved almost exclusively in one direction – But reversing a law is easier than reversing its effects.

Two years into Michigan’s experiment, union membership has not surged as some had hoped and even remains near historic lows. That doesn’t necessarily mean repeal failed, but it does suggest something more important about what government can—and cannot—do to rebuild worker power.

How Michigan Lost the Union State

It’s important to establish at the outset that Right-to-work didn’t destroy Michigan’s unions.

Michigan’s union decline began long before Republicans enacted the law in 2012. In 1989, 26% of Michigan wage-and-salary workers belonged to a union. By 2004, it was 21.6%. By 2012, it was 16.6%. Right-to-work arrived in the middle of that decline.

The policy itself is also frequently misunderstood.

Under federal labor law, a union chosen to represent a workplace generally has an obligation to represent workers covered by the bargaining unit. Right-to-work laws prevent unions and employers from negotiating agreements requiring represented workers to financially support that representation.

• Supporters call that worker choice: no employee should have to financially support a union he or she doesn’t want to support.

• Unions call it a free-rider problem: workers can benefit from collectively negotiated wages, benefits and representation without contributing toward the organization paying to negotiate and enforce them.

Whatever terminology you prefer, research suggests the policy matters.

One study examining five states that adopted right-to-work laws between 2011 and 2017—including Michigan—estimated that unionization was about four percentage points lower five years after adoption, while wages were about 1% lower. The effects were particularly pronounced in industries that began with relatively high unionization.

All that to say that right-to-work didn’t create Michigan’s labor problem; however, there is evidence that it made an existing problem worse.

Turning the Ratchet Backward

For decades, right-to-work offered a case study in how conservatives used state government.

The 1947 Taft-Hartley Act allowed states to enact the laws. Many Southern and Western states did. Then another wave reached traditionally union-heavy parts of the Midwest. Indiana enacted right-to-work in 2012. Michigan followed. Wisconsin adopted it in 2015. West Virginia followed in 2016.

The political ratchet largely moved one way – but Michigan eventually demonstrated that it didn’t have to.

Democrats won control of both chambers of the Michigan Legislature in 2022, giving the party a governing trifecta. In March 2023, Whitmer signed legislation repealing right-to-work. The repeal took effect on February 13, 2024.

For private-sector workplaces, unions and employers could once again negotiate union-security provisions requiring workers represented by the union to financially support that representation. The change did not recreate the same system for public employees, where the U.S. Supreme Court’s Janus decision prohibits compulsory agency fees.

The broader significance, however, went beyond the technical change. Michigan became the first state in 58 years to repeal right-to-work. That’s an important lesson for Democrats interested in state policy: policy regimes aren’t permanent – states can be laboratories for reversing policy experiments too!

But that leaves the harder question – did it work?

So, Did It Work?

The first year looked promising.

Michigan entered 2024 with its union membership rate at 12.8%—the lowest level recorded since comparable state data began in 1989. Then right-to-work repeal took effect. By the end of the year, Michigan had approximately 581,000 union members, up from 564,000. Union density climbed from 12.8% to 13.4%.

It would have made for a tidy story: Michigan repealed right-to-work. Union membership went up.

Then came 2025 where Michigan’s union membership fell back to approximately 566,000. Union density slipped to 13%. In other words, two years into the post-right-to-work experiment, Michigan has roughly as many union members as it did before repeal.

Those numbers require some caution. State-level union estimates come from relatively small survey samples, and federal statisticians warn against reading too much into small year-to-year changes. The 2025 figures are additionally based on only 11 months because October data weren’t collected during the federal government shutdown.

But inconvenient data shouldn’t simply be explained away. There is no evidence yet of the dramatic union revival that supporters might have hoped would follow repeal.

The more interesting question is whether we should have expected one.

What Government Can—and Can’t—Do

Repealing right-to-work doesn’t organize a single workplace.

It doesn’t convince workers at a nonunion factory that they should organize. It doesn’t win a representation election. It doesn’t force an employer to agree to a first contract. And it certainly doesn’t reverse decades of automation, deindustrialization, globalization and changes in the composition of the American workforce.

What repeal does is change the institutional environment in which unions operate.

A union that represents a bargaining unit must expend resources negotiating contracts, handling grievances and representing workers. Right-to-work allowed workers receiving those services to opt out of financially supporting the union.

Michigan has now removed that restriction for private-sector collective bargaining.

That can potentially give unions more reliable resources to service existing members and organize new ones. But there is an important distinction: strengthening the infrastructure of organized labor is not the same thing as expanding organized labor.

Michigan can make it easier for unions to build. Unions still have to build.

And because collective-bargaining agreements operate on multiyear cycles, the practical consequences of repeal won’t arrive everywhere simultaneously. Union-security provisions may have to be negotiated as contracts expire.

That makes Michigan’s experiment less like flipping a switch and more like resetting a machine that has been moving in the opposite direction for decades.

The real test isn’t simply whether Michigan’s union density moved from 12.8% to 13.4% and then 13%. It’s what happens next.

The Case Against

There is another reason not to judge Michigan solely by its union membership numbers. Opponents made a different prediction – they argued repeal would hurt Michigan’s economy.

The U.S. Chamber of Commerce warned lawmakers in 2023 that right-to-work helped Michigan compete for economic-development projects and jobs. If businesses were choosing between otherwise similar sites, the Chamber argued, right-to-work could tilt investment toward Michigan.

The philosophical argument was equally straightforward: workers should decide for themselves whether a union provides enough value to deserve their financial support. Those aren’t totally frivolous objections.

If stronger unions produce higher labor costs, businesses deciding where to build their next factory may take those costs into account. Michigan isn’t competing only against Ohio or Illinois. It is competing for manufacturing investment against states across the South where unionization is much lower.

That gives Michigan’s experiment two potential measures of failure.

1. Repeal could fail because it doesn’t meaningfully strengthen organized labor

or

2. It could strengthen labor while making Michigan substantially less competitive for investment and jobs.

So far, neither conclusion is obvious. There hasn’t been a dramatic union resurgence. But there also isn’t compelling evidence that ending right-to-work has produced the economic calamity opponents warned about.

That leaves Michigan in a much more interesting place than either side’s talking points suggest.

The Michigan Test

The next several years should tell us much more. Ultimately, Michigan’s right-to-work repeal experiment should be judged by three tests.

1. Institutional strength - Do Michigan unions become financially and organizationally stronger as collective-bargaining agreements turn over and the repeal works its way through unionized workplaces?

2. Organizing strength - Do those stronger institutions translate their resources into organizing campaigns, election victories and first contracts that actually bring more workers into organized labor?

3. Worker outcomes - Does stronger collective bargaining translate into higher compensation, better benefits and greater economic security without imposing an economic cost large enough to outweigh those gains?

These tests matter because Michigan’s experiment speaks to something larger than right-to-work.

Democrats have spent much of the past decade searching for ways to rebuild their economic relationship with working-class voters. Much of that agenda asks government to do more for workers: subsidize child care, lower health-care costs, provide paid leave, forgive debt or send tax credits.

Those policies can materially improve people’s lives.

But Labor policy operates differently. Its underlying proposition is that government doesn’t have to negotiate every economic gain on workers’ behalf. It can instead change the rules governing the workplace and give workers more power to negotiate those gains themselves.

Michigan is testing how far that proposition can go.

Bottom Line

There is an appealing simplicity to Michigan’s right-to-work repeal.

Republicans passed a law designed to limit unions. Democrats took power and repealed it. Organized labor gets stronger. Reality is proving more complicated.

Michigan’s labor movement was declining long before right-to-work arrived. Repealing one law was never going to reverse decades of economic and institutional change overnight. But that doesn’t make the repeal meaningless.

Michigan has removed a policy that research suggests contributed to lower unionization and wages. It has demonstrated that the conservative state-policy ratchet can run backward. And it has given organized labor a more favorable set of rules under which to rebuild.

What it hasn’t done is rebuild labor for them. That may ultimately be the most useful lesson from Michigan’s experiment.

Government can change the rules of the game. But Labor has to prove it can win under them.

When Vermont lawmakers enacted Act 76 in 2023, they did more than enlarge a subsidy program. They made a bet about what child-care policy should become.

Instead of treating assistance as a narrowly targeted benefit for poor families, Vermont extended help deep into the middle class. Instead of asking providers to absorb the difference between what parents could pay and what care cost to deliver, the state increased reimbursement rates. And instead of financing the expansion through temporary federal aid or an annual budget scramble, lawmakers imposed a dedicated 0.44% payroll contribution.

Three years later, the early result is neither the universal success story advocates sometimes describe nor the tax-heavy cautionary tale opponents predicted. Vermont has made child care substantially more affordable for thousands of families and has begun to stabilize a shrinking provider market. But it still cannot produce enough workers and slots—especially for infants, toddlers, and families in thin rural markets.

That mixed result is precisely what makes Vermont’s policy approach to child care particularly valuable for other states looking to address the child care crisis. Act 76 is a blueprint for how states can begin building a child-care system - it is also a warning that financing demand is faster than building supply.

Vermont’s Market was Broken

Child care has an unusual economic problem. It is too expensive for many families and still too cheap to support the wages necessary to retain workers. Providers cannot solve that contradiction by becoming more efficient: caring for infants and toddlers requires a high ratio of adults to children.

Vermont’s pre-reform data showed the damage. Between 2012 and 2023, the number of home-based providers fell nearly 60%. Statewide infant capacity declined 10% and toddler capacity 19%. By 2023, Vermont had fewer than 40 licensed slots for every 100 infants. Child-care workers earned roughly 55% of the average Vermont worker’s pay, and turnover exceeded the rate across the state economy.

The politics developed over more than a decade. Advocates framed child care as workforce infrastructure—something that allowed parents to work and employers to hire—while earlier legislation funded studies and incremental improvements. In 2023, the Democratic-led legislature passed H.217 over Republican Gov. Phil Scott’s veto. Scott agreed that child care needed investment but opposed the payroll tax and scale of spending.

That disagreement clarified the choice. A cheaper program could serve fewer families or pay providers less. Vermont chose a broad, durable system and attached a visible tax to it.

What Does Act 75 Do?

Act 76 expands eligibility for the Child Care Financial Assistance Program to 575% of the federal poverty level, bringing substantial portions of the middle class into the program. Families at or below 175% of poverty have no family payment; assistance then phases down as income rises.

On the supply side, Vermont raised provider reimbursements by 35%, stopped tying state payments to a provider’s lower private rate, and later narrowed the gap between payments to home-based and center-based programs. It also financed readiness, quality, and capacity grants.

The sequencing mattered. Stabilization grants and higher provider rates arrived before the final eligibility expansion in October 2024. In theory, that gave programs time to raise pay, hire, or reopen classrooms before subsidized demand peaked. The dedicated Child Care Contribution took effect in July 2024. The assessment equals 0.44% of covered payroll; employers can pass no more than one-quarter of it—0.11%—to employees. Through November 2025, it had collected about $99.6 million.

Ironically, the tax is both the policy’s political vulnerability and one of its chief strengths. Earmarked revenue is harder to quietly repurpose and gives providers more confidence that higher rates will last. But payroll taxes can modestly reduce compensation or hiring over time, even when employers formally pay them. States should acknowledge that cost rather than pretending the funding is free.

Key Policy Takeaway - Affordability Moved First

The clearest outcome is participation.

Child Trends found that families receiving subsidies increased 48%, from 5,389 in July 2023 to 7,983 in April 2025. State monitoring shows children enrolled in CCFAP rising from 7,533 in June 2023 to 12,234 in September 2025.

That does not prove every new enrollee obtained a new slot; some families were already paying privately and became eligible for help. But it does show that Vermont successfully reached beyond the traditional low-income subsidy population. For those families, the program converts a large, volatile household bill into a smaller payment tied to income.

The provider response is real but slower. Child Trends found a 2.4% increase in the number of providers between November 2023 and December 2024, including a 3% rise in home-based providers. The share of providers accepting subsidies increased by 4.8 percentage points. State data show licensed programs rising from 1,045 in September 2023 to 1,072 two years later.

Those are meaningful reversals after a long contraction, but they are not yet a supply transformation. Vermont’s 2026 monitoring report still identifies workforce shortages as the largest barrier to expansion. Affordability improved more rapidly than availability, and the gaps remain greatest where care is most labor-intensive.

The Strongest Critique is About Execution, Not Intent

Act 76’s early fiscal record also calls for humility.

State economists now expect the payroll contribution to raise about $11 million less per year than originally projected because the covered tax base was smaller than estimated. The program was not facing an immediate funding crisis: Vermont projected that roughly 99% of its approximately $172 million FY2026 CCFAP budget would be spent. Still, the miss is material and should shape other states’ planning.

Administrative complexity is another constraint. Providers have reported friction around enrollment-based payments and the interaction of child-care subsidies with Vermont’s prekindergarten system. Meanwhile, contemplated pre-K restructuring did not proceed on its original July 2026 timetable.

There is also an evidence problem. Vermont can document enrollment, payments, providers, and licensed capacity. It cannot yet confidently attribute changes in parental employment, statewide labor supply, child outcomes, or long-term provider wages to Act 76. Future evaluations should resist converting plausible benefits into measured ones.

The Blueprint for Other States

The lesson is not simply “adopt a payroll tax.” The model has five linked parts.

1. Finance families and providers together. Expanding subsidies without raising provider rates can create longer queues for the same scarce slots. Raising rates without family assistance leaves care unaffordable.

2. Stabilize supply before expanding eligibility. Vermont’s sequencing was sensible, even if the runway was too short to eliminate shortages.

3. Use permanent revenue—but forecast it conservatively. A replicating state should model collections at least 10% to 15% below the central estimate and create an automatic response if revenue or enrollment diverges from projections.

4. Pay explicitly for hard-to-provide care. Infant care, nonstandard hours, rural service, and care for children with specialized needs will not emerge evenly from a general rate increase. Enhanced rates and targeted capacity grants should be designed from the beginning.

5. Measure functional capacity, not just licenses. The relevant questions are whether classrooms are staffed, whether families can find hours that match work, and whether providers can retain employees with wages and benefits—not how many slots exist on paper.

Bottom Line

Vermont has not solved the child care crisis entirely but it has done something instructive: it has shown that a state can rapidly remove price barriers and interrupt provider decline, while revealing where cash alone stops working.

For Democrats looking for a governing model, Act 76 is worth adapting.

But the honest promise is not instant universality – it is a durable public system built in the right order—and judged by whether a parent can actually find care.