By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panelists agree that the 1.9% drop in the Dodge Momentum Index signals a pause rather than a collapse in nonresidential construction planning, but they disagree on whether the shift towards institutional projects is a healthy rotation or a sign of exhaustion in the construction cycle.
Risk: A financing hurdle or higher energy costs that throttles 2H starts, or an acceleration of the 'moderation' in data center activity into actual contraction.
Opportunity: AI-driven demand and data-center backlogs could reaccelerate if funding conditions ease.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
This story was originally published on Construction Dive. To receive daily news and insights, subscribe to our free daily Construction Dive newsletter.
Data center planning cooled off month over month from "extraordinary levels" to start the summer, while other areas of construction finally picked up a little steam, according to Dodge Construction Network.
The Dodge Momentum Index, a measure of nonresidential construction projects entering the planning stages, decreased 1.9% month over month in June, according to a Wednesday news release. The decline, which followed two consecutive monthly gains, stemmed largely from a tap on the brakes around data center planning.
"Despite June's pullback, nonresidential planning remains on solid ground," said Sarah Martin, director of economic research at Dodge Construction Network. "Data center activity continued to drive the index, but its pace moderated from the extraordinary levels seen in recent months and drove the DMI to pull back over the month."
Commercial planning, which includes data center work, ticked down 6.8% month over month in June, according to the report. Apart from data centers however, "nearly every other sector" in the commercial category improved over the course of the month, said Martin. Planning for traditional office buildings, warehouses, retail stores and hotels, for example, posted monthly gains in June.
Planning for institutional projects, which include healthcare and education construction, picked up as well, growing 10.9% month over month in June.
Year-over-year planning figures were still well ahead of 2025 levels, according to Dodge. The Dodge Momentum Index and the commercial and institutional segments all gained 21.8% when compared to June 2025, according to the report.
Dodge noted that without data center plans, the commercial segment would have only posted about a 7.6% jump from year-ago levels in June. A total of 59 projects valued at $100 million or more entered planning throughout June, according to Dodge.
Major commercial projects included:
The largest institutional projects to enter planning included:
Recommended Reading
Four leading AI models discuss this article
"The rotation from hyper-concentrated data center growth to broader institutional spending indicates a more sustainable, less volatile construction cycle."
The 1.9% dip in the Dodge Momentum Index is noise, not a trend. The market is hyper-fixated on data centers, but the real story is the 10.9% surge in institutional planning. This suggests a broader base for construction spending beyond the AI-hyped hyperscaler boom. While data center activity is 'moderating,' it is normalizing from unsustainable, parabolic levels, which is actually healthy for supply chain stability. If institutional sectors like healthcare and education are picking up the slack, we are looking at a more resilient, diversified nonresidential cycle. Investors should look past the headline 'cooling' and focus on the widening breadth of project starts, which signals a more durable economic expansion.
The cooling in data center planning could indicate that power grid constraints and site selection bottlenecks are finally hitting a hard ceiling, threatening the primary engine of nonresidential growth.
"Data center planning decelerated from unsustainable peaks, not from healthy baseline—distinguishing between cooling and crashing is critical to the outlook."
The headline screams 'data center slowdown,' but the real story is messier. Yes, commercial planning fell 6.8% MoM—but that's almost entirely data centers cooling from 'extraordinary' levels, not collapsing. Strip them out and commercial is up 7.6% YoY. Institutional jumped 10.9% MoM. Three data centers worth $1.4B still entered planning in June alone. The pullback looks like normalization after a spike, not demand destruction. The risk: if this 'moderation' accelerates into actual contraction, we're seeing early warning signs. But one month of slower growth after 'extraordinary' activity isn't a recession signal yet.
Data center planning volatility makes month-to-month comparisons noise; the real test is whether the YoY 21.8% gain holds through Q3. If July-August show continued sequential declines while comparables get tougher, the narrative flips from 'normalization' to 'demand rolling over.'
"N/A"
[Unavailable]
"The June dip is a timing pause, not a fundamental shift, but actual starts will hinge on financing and energy constraints; if rates remain elevated or power costs rise, the data-center capex cycle could stall longer than the article implies."
June’s 1.9% MoM drop in the Dodge Momentum Index flags a pause after an ‘extraordinary’ data-center surge, not a terminal top. While the headline says nonresidential planning remains solid, excluding data centers would leave a modest +7.6% YoY gain, underscoring that most strength is concentrated in hyperscale capex. Missing context includes project lead times, financing mix, and power/connectivity constraints that could delay starts even as plans advance. The biggest risk is a financing hurdle or higher energy costs that throttles 2H starts; the upside remains AI-driven demand and data-center backlogs could reaccelerate if funding conditions ease. The article glosses over timing risk and the depth of reliance on data-center planning.
Against my neutral stance: the data-center cycle is structurally supported by AI demand and long project backlogs; a temporary June pause could quickly resolve as funding lines open and interconnection capacity scales, keeping the uptrend intact.
"The institutional sector lacks the capital depth to offset a structural slowdown in hyperscale data center investment."
Gemini and Claude are overly optimistic about the institutional shift. They ignore that healthcare and education construction are notoriously interest-rate sensitive and plagued by labor shortages that erode margins. If data center growth 'normalizes'—a polite term for hitting a power-grid wall—the sector cannot simply pivot to institutional projects. Those projects lack the hyperscaler capital backing to absorb current high borrowing costs. We are likely seeing the exhaustion of the construction cycle, not a healthy rotation.
"Institutional strength means nothing if it's just mean reversion; Gemini hasn't proven the 10.9% MoM is durable demand, not a bounce."
Gemini conflates two separate risks. Yes, healthcare/education are rate-sensitive—but that's a *structural* headwind independent of data center normalization. The real question: does institutional's 10.9% MoM surge reflect genuine demand or just statistical reversion after weakness? If it's the latter, Gemini's 'healthy rotation' thesis collapses. We need YoY institutional growth rates, not just monthly swings, to know if this is real breadth or noise.
[Unavailable]
"A durable breadth rotation via institutional starts is unlikely if funding, power, and labor constraints bite, making a 2H slowdown more likely than a stable expansion."
Responding to Gemini: even if healthcare/education starts rise, breadth isn’t proven. MoM spikes can mask a shallow YoY base, and institutional demand is highly rate-sensitive with tight margins. If high borrowing costs persist and the data-center backlog crowds out financing for nonresi starts, institutional work won’t compensate. The risk: a 2H downshift if power, labor, and credit constraints bite; the ‘diversified’ story rests on fragile assumptions.
The panelists agree that the 1.9% drop in the Dodge Momentum Index signals a pause rather than a collapse in nonresidential construction planning, but they disagree on whether the shift towards institutional projects is a healthy rotation or a sign of exhaustion in the construction cycle.
AI-driven demand and data-center backlogs could reaccelerate if funding conditions ease.
A financing hurdle or higher energy costs that throttles 2H starts, or an acceleration of the 'moderation' in data center activity into actual contraction.