Employing Correlation

The relationship between construction employment, flatbed pricing, and truckload markets is a tale of both correlation and caution. Let’s dive into the dynamics behind these trends and understand their limits, with a nod to expert insights from Ed Zarenski. He offers a ton of great content on his site. Please check him out.

One chart I love to follow of his is construction spending to volume. I’ve added it below. It’s particularly helpful when trying to make sense of things in our latest inflation lapse as price (spending in green) has become a more unreliable narrator for volume given their widening gap.

More helpful then, are projections he has for the volume tied to spending projections. In the supply chain, any leading indicator is a prize. Tender and post data get you closer to the spigot, but any portfolio is best diversified, so many of us look for corroborators too.

Why construction?

Well, everything you see in this room came on a truck at some point, right? Then looking at how many rooms are being built, and what to degree helps figure how many of those trucks may be needed. Housing starts start the truck if you want a pneumonic.

On that note, a recent Bluesky post of his caught my eye.

Ed Zarenski’s construction data shows that the top hiring months for construction since 2011 are February, March, September, May, and July. What does this mean for trucking? Construction drives flatbed demand, with project cycles heavily influencing freight capacity needs. For example, flatbed spot rates tend to peak between March and June, mirroring the annual rush of housing starts just as the ground thaws.

My rendition does not exclude the wild pandemic months, and stops at 2014, so I don’t get Ed’s precision. It’s nonetheless visible.

The chart below, showcasing DAT Flatbed Load-to-Truck Ratios, demonstrates this demand compression. Early Q1 pressures align with mulch, lumber, and building supply seasonality. For anyone keeping tabs on freight, this validates how construction cycles and flatbed performance are tightly woven.

Flatbed Freight’s Role in the Freight Cycle

Building on Ed’s data, I explored correlations between construction employment (BLS data) and flatbed rates while associating Dry Van trends closer to overall Truck Transportation employment. Specialized trucking forms a small part of overall truck transportation, so construction employment trends emerge as a strong proxy. Over time:

• Periods of Stability: During stable market cycles, such as the post-lockdown recovery of 2020, construction employment and flatbed rates moved in near-unison, with correlations exceeding 0.8. This alignment suggests that in “normal” demand cycles, flatbed freight closely tracks construction labor demand.

  

• Periods of Shock and Contribution Mix: Memory is good enough to tell how disjointed the world was, and possibly has been since the pandemic first struck. Their watermarks break the scales of many visualizations comparing the extremes. It is hopefully an edge case, but it is one of myriad types of shock.

   

  Fuel and resource extraction activities (see paper) are also big determinants of flatbed demand. Although multiple sectors were down in 15-16, construction employment losses lagged as shown in a later graph. The precursor here was the oil glut and price collapse mid decade.

• Dry Van Dynamics: Running correlations between dry van spot rates and truck transportation employment were lower overall (.31 vs .34) but still showed alignment during stable demand phases. Manufacturing is the largest factor for this segment. Therefore, green shoots in this sector are leading to brighter spirits in 2025. Higher freight costs usually come with higher sales revenues, the volume throughput varies as we’ve seen in construction spend.

• Trucking as a Barometer: Specialized trucking (like flatbed) offers a sharper lens into specific market activities—such as construction and industrial projects—while broader truck transportation trends are more reflective of broader set of national economic conditions.

This leads to a key takeaway: Correlations can be context-sensitive. Use them cautiously, particularly when the market experiences a demand-side shock or a structural supply chain bottleneck.

Utilizing a static coefficient, meant to enhance any model, could turn into its downfall. For some, incorporating anything like this at all would be overwhelming. The point isn’t added sophistication, but building strategy with a variable set of information.

When to Use, When to Lose

Blending these insights with Phares, Miller, and Burks' recent findings, in a paper titled Shedding light on truck driver supply and demand: Heterogeneous state-level recovery of trucking employment following the COVID-19 employment shock, it’s clear that trucking employment and pricing cycles offer a nuanced barometer of freight trends:

1. Use These Correlations During Normalcy:

   • For example, post-pandemic recoveries (e.g., 2020-2021) saw trucking employment and flatbed rates move in tandem, signaling demand strength across sectors like construction and logistics.

     

   • Actionable Insight: Use YoY changes in construction employment alongside flatbed rates to project regional freight demands.

   

2. Corroborate and Extrapolate: As noted in the paper, there are strong ties between infrastructure projects and local demand. Many goods first start at sea, either ping-ponging across a network or as an intermediary step in value creation. TEU and commodity flows then signal to ports and accompanying warehouses solid expectations with decent lead times to prepare.

3. Look at all three phases: Operations and insights are coming from past, present, and projecting sources. It’s important to know what’s being collected or presented and what it represents for leading indication or validation. A mix of timeframe can be as illuminating as source diversity.

4. Lose Faith During Disruption:

   • During periods of volatility (e.g., 2020's early pandemic or 2022's post-boom slowdown), these correlations falter.

   • Actionable Insight: In these times, supplement spot rate and employment trends with a mosaic of indicators to ensure early flashing lights can come on, but they’re not leading the show.

Closing Thoughts

Understanding freight is like solving a puzzle—each piece fits into a larger picture. By layering flatbed-specific insights with broader employment data, you can navigate freight cycles more effectively if exposed to this mode. The key is knowing when to trust these correlations and when to look beyond them.

No one set of data is gospel, nor does one need high levels of sophistication to incorporate. Data is multifaceted. It can come from oneself, third parties, the government, sharing graphs and talking to stakeholders, and on.

Learning when to employ it is just as useful as what you’re using.