Why Dark Energy Cannot Stay Constant Forever

Most people assume the rate at which the universe expands is a settled number, something measured once and filed away. It isn't. Two completely independent methods of measuring that rate keep returning two different answers, and the gap between them has survived every attempt to explain it away as error. At five sigma statistical confidence, the same threshold required to confirm the Higgs boson, physicists are no longer asking which measurement is wrong. They're asking what the disagreement is telling us about the nature of dark energy itself.

The Number That Won't Resolve

The Hubble constant, written as H₀, describes how fast two points in space are moving apart per unit of distance between them. It sounds like the kind of thing that should have one answer. It doesn't.

The first measurement method, called the distance ladder, uses Cepheid variable stars and Type Ia supernovae as cosmic mile markers, building outward from our own galaxy with each step calibrated against the last. This approach consistently returns a value near 73 kilometers per second per megaparsec.

The second method works backward from the cosmic microwave background, the afterglow of the early universe emitted 380,000 years after the Big Bang. Run through the standard cosmological model, this data consistently predicts a present-day expansion rate closer to 67 kilometers per second per megaparsec.

Neither team is making an obvious mistake. Both methods have been refined for decades, cross-checked against independent techniques, and stress-tested by some of the most careful researchers in the field. The gap between 67 and 73 simply will not close.

What JWST Was Supposed to Fix

Before the James Webb Space Telescope launched, there was real hope it would resolve this. The leading theory was that older Hubble Space Telescope images suffered from stellar crowding, with nearby stars bleeding light into Cepheid brightness measurements and inflating the local expansion rate. JWST's sharper resolution was supposed to correct this and pull the local value down toward 67.

It didn't happen. When the corrected JWST results came back, the contamination was real but small. After accounting for it, the local measurement still sat comfortably above 70. The instrument that was supposed to dissolve the tension confirmed it instead.

This matters because it closes off the most comfortable explanation. If a better telescope can't make the numbers agree, the standard cosmological model may be missing something structural rather than something accidental.

The Assumption Hiding Inside the Model

The standard model of cosmology, called Lambda-CDM, treats dark energy as a cosmological constant. Lambda is fixed and unchanging, the same energy density in empty space at every point in cosmic history. This assumption isn't incidental. It's baked directly into how early-universe data gets translated into a present-day expansion rate.

If Lambda isn't actually constant, every value of H₀ inferred through the model carries a hidden distortion. Not from faulty data, but from an assumption that quietly doesn't hold.

The equation that governs how dark energy density changes with the universe's expansion is:

$$\rho \propto a^{-3(1+w)}$$

Here, rho is the energy density, a is the scale factor describing the universe's relative size, and w is the equation of state parameter, the ratio of dark energy's pressure to its density. For a true cosmological constant, w equals exactly negative one, and the exponent becomes zero. Density stays flat no matter how much the universe expands.

For anything other than negative one, density changes with expansion. And the direction of that change determines almost everything about the universe's long-term fate.

What the DESI Survey Found

The Dark Energy Spectroscopic Instrument measures the clustering of galaxies across enormous volumes of space, using a pattern called baryon acoustic oscillations as a cosmic ruler^[1]. By measuring this pattern at different cosmic epochs, researchers can reconstruct the universe's expansion history in detail.

When DESI's 2024 and 2025 data releases were fit against models allowing dark energy's equation of state to vary over time, the fit was measurably better than the fixed cosmological constant. The preferred values suggested dark energy that had been more intense in the recent cosmic past, drifting toward less extreme behavior closer to today.

This is not confirmation. The statistical significance sits around 2.5 to 3.9 sigma depending on which datasets are combined, well short of the five sigma threshold for discovery. But it's a consistent lean, not a single anomaly, and it points in a specific direction: dark energy might not be sitting still.

The Line at Negative One

There's a particular threshold in the equation of state that changes everything. Above negative one, dark energy density holds steady or dilutes as the universe expands. Expansion accelerates, but within limits. Below negative one is a different regime entirely, one that physicists call phantom energy.

In a phantom scenario, dark energy density doesn't dilute with expansion. It grows. More space generates more dark energy, which drives faster expansion, which generates still more space and still more energy. There's no equilibrium point in that loop. The feedback simply compounds.

This possibility was first formalized in 2003 by physicist Robert Caldwell and colleagues, who modeled a universe with w equal to negative 1.5 and calculated total structural destruction in approximately 22 billion years^[2]. They called the endpoint the Big Rip.

Current measurements from the Planck satellite place w at negative 1.028, with an uncertainty of plus or minus 0.031. The phantom threshold sits inside that uncertainty range. It cannot currently be ruled out.

How Sensitive the Timeline Actually Is

The relationship between w and the time remaining before a hypothetical Big Rip is not linear. It's exponential, and the sensitivity near negative one is extreme.

$$t_{rip} - t_0 = \frac{2}{3H_0|1+w|\sqrt{1-\Omega_m}}$$

This equation shows that as w approaches negative one from below, the denominator shrinks toward zero, and the time to the singularity stretches toward infinity. A value of w at negative 1.01, barely phantom, pushes the Big Rip out to roughly two trillion years. A value of negative 1.5 collapses that timeline to 22 billion years.

Current measurement precision simply isn't tight enough to distinguish between these outcomes. The same uncertainty that allows for a cosmological constant also allows for a phantom universe with a timeline measured in trillions of years rather than billions.

Why This Isn't Settled, and Why That's the Point

It would be easy to read all of this as confirmation that the universe is doomed to tear itself apart. That isn't what the evidence shows. The cosmological constant remains entirely consistent with current data. The Hubble tension could resolve through early-universe physics that has nothing to do with phantom energy. The DESI signal could weaken as more data accumulates over the next several years.

What the evidence does show is that the comfortable assumption, dark energy as a fixed, unchanging property of empty space, is no longer the obvious default. Multiple independent lines of evidence, using different instruments and different physical principles, are leaning in a consistent direction. That consistency is what makes this worth taking seriously rather than dismissing as noise.

What We Actually Know

The Hubble constant has two measured values that disagree at a statistically confirmed level. JWST closed off the most likely instrumental explanation. The standard cosmological model assumes dark energy is constant, and that assumption may not hold. DESI's baryon acoustic oscillation data shows a moderate but real preference for dark energy that varies over cosmic time. If dark energy's equation of state sits even slightly below negative one, the mathematics describes a universe whose expansion accelerates without limit, ending in a finite-time singularity.

None of this is proven. All of it is measurable, and the measurements are getting more precise every year. The next decade of data from DESI, Euclid, and the Vera Rubin Observatory will likely tighten these constraints enough to settle the question one way or the other. For now, the two numbers sit side by side, neither one retreating, and the gap between them remains the most interesting unanswered question in modern cosmology.