Why Measuring Inflation Is Like Baking Grandma's Chocolate Cake
CPI Series: Part 2 — Feeling the Problem Before We Solve It
If you made it through Part 1, you now have a more accurate picture of CPI than most people ever develop, including many of the people paid to have opinions about it. You know what the basket is, why things get weighted differently, and why the BLS tracks price changes rather than prices themselves. That’s no small thing.
But there is a difference between understanding what CPI does and understanding why getting it exactly right may be impossible. The first is a matter of reading carefully. The second requires sitting with the complexity long enough for it to stop feeling abstract. So before we go any further, we’re going to bake a cake.
Specifically, Grandma’s chocolate cake. The one she’s been making since 1970. The one you always remember tasting amazing.
The question we are going to sit with is this: how would you even know if your measurement of inflation was wrong? Not wrong in an obvious way, but quietly, systematically, consistently wrong in a direction that nothing in the system was designed to detect. That is the conceptual problem this article is building toward. Because if you cannot answer that question satisfactorily, then everything built on top of that measurement, every interest rate decision, every wage negotiation, every inflation-adjusted figure in the economy, is built on a foundation that has never been fully examined.
The Measurements Keep Shifting
In this section of the analogy, the recipe represents the CPI basket itself, the collection of goods and services the BLS uses to track the cost of living. The ingredients are the individual items in that basket. The ratios between them are the weights. Keep that in mind as we go.
The first obstacle for CPI is that the economy it measures never holds still, and neither does grandma’s recipe.
Grandma’s original recipe called for two cups of flour and half a cup of cocoa. But over time, as tastes shifted and ingredient costs changed, the recipe had to be adjusted. Maybe it’s now a cup and three-quarters of flour and three-quarters of a cup of cocoa. The individual ingredients are the same, but the balance among them has been quietly redrawn to reflect people’s changing demands.
This is exactly what the BLS does when it updates the CPI weights. Housing used to carry a different share of the average household budget than it does today. So did transportation. So did food. As spending patterns shift, the ratios between categories are rebalanced to ensure the index continues to reflect reality. Starting in 2023, the BLS began updating these weights annually rather than every two years, precisely to keep the measurement aligned with what consumers are spending money on.
And that updating must happen. If the BLS stopped adjusting the weights, it would start measuring the cost of things people no longer buy in the proportions they once did. Demand for outdated goods drops, and their prices can behave oddly as production shrinks or markets thin out. If CPI kept using its old weights, the index could start to drift away from how people live, effectively measuring yesterday’s shopping cart. The updates are not a flaw. They are what keep CPI credible.
But now take the problem one step further. What if it’s not just the ratios that change, but the entire cake? What if your town stops wanting chocolate cake altogether and switches to vanilla? You can either continue measuring the cost of a chocolate cake that no one wants anymore, or update the recipe to reflect what people are buying.
The answer must be to switch to baking a vanilla cake. Because the whole point of CPI is to measure the cost of living for a typical American household, and if that household has moved on to vanilla, then chocolate is no longer the right thing to measure. You change the recipe not to lose the thread, but to keep it. Every adjustment, every rebalancing, every basket update exists for the same reason: to make sure that tomorrow you are still measuring the same concept you were measuring yesterday, even if the specific goods that define that concept have changed.
That’s a genuinely difficult thing to do well. And for the most part, the BLS does it well.
Which makes what comes next more unsettling. Because the problem we’re about to look at has nothing to do with the ingredients, or the ratios, or even which cake you’re making. It has to do with the measuring cups themselves.
The Measuring Cups Are Shrinking
In this part of the analogy, the measuring cups represent the dollar itself, the unit we use to measure everything else. The ingredients represent the goods that productivity makes cheaper.
Imagine you’ve been baking Grandma’s cake for decades. You know the recipe by heart. You’ve made all the right adjustments over the years, updated the ratios, switched to vanilla when the town moved on from chocolate, kept everything honest and current. And the cake keeps coming out tasting pretty good. Maybe even better than it used to.
But something has been happening in the background that nobody noticed. The measuring cups have been slowly shrinking.
Not dramatically. Not all at once. Just a little, every year. What used to be a full cup is now seven-eighths of a cup. Then three-quarters. The recipe looks identical. You’re following the same instructions. But the actual amount of each ingredient going into the cake has been quietly decreasing for decades.
And here’s what makes it so disorienting: you start to notice that you are using more cup flour than you used to, but getting less cake. Something feels off. You could swear the recipe used to go further. But the cake still tastes good, better than it ever did, so you talk yourself out of it. Maybe you’re misremembering. Maybe this is just how it works now.
And here is what makes this so difficult to detect. The cake keeps tasting better. Not just fine, genuinely better. The flour is more refined, the cocoa richer, the eggs more consistent. Every year, the ingredients improve, and every year, the cake comes out a little more impressive than the year before. So even as you find yourself using more flour than you used to, even as something in the back of your mind says the recipe used to go further than it does now, the cake on the plate in front of you looks better than anything Grandma ever made. How do you argue with that? How do you tell someone the measuring cups are shrinking when the cake keeps getting better?
This is the dynamic that has hidden the error for so long. Living standards kept improving. The things people could buy with their wages kept getting better and more abundant. So even as the dollar was quietly losing purchasing power, the productivity gains flowing through the economy were delivering real improvements in what that dollar could buy. The cake tasted better. And that made it almost impossible to notice that the measuring cups were shrinking.
And it is exactly what has been happening with the dollar and productivity.
Over the past century, productivity gains have been quietly making everything cheaper to produce. Better machinery, more efficient supply chains, improved agricultural yields, and, more recently, technology and automation have steadily driven down the cost of producing the goods that make up the CPI basket. Those lower production costs have flowed through to lower consumer prices, or at least lower prices than would have prevailed without the productivity gains.
In standard economics, inflation is defined as the rate of change of a price index, such as the CPI. By that definition, CPI has done its job: it tracks those observed price changes. What NETs argues is that those changes already include a powerful, invisible offset from productivity, so the inflation we see is a net number, after the productivity dividend has quietly lowered costs.
Which means CPI has been measuring net inflation all along, and reporting it as if it were gross inflation.
The measuring cups were shrinking. But the ingredients kept getting better. And that was enough to keep everyone from asking the question that needed to be asked.
The Last Person Who Remembered
There is one more thing worth understanding about how we got here.
When Grandma was alive, the recipe had a guardian. Someone who had made the cake a thousand times, who knew instinctively what it was supposed to taste like. Who could tell immediately when something was off, not because she was following a formula, but because she carried the reference point inside her. The recipe existed on paper, but Grandma was the living standard against which everything else was measured.
When Grandma left, we kept the recipe. We updated it carefully, adjusted the ratios, and switched from chocolate to vanilla when the town moved on. We did everything right. But we lost the one thing the recipe could never capture, the ability to taste whether we were still making the right cake.
In the real economy, Grandma was the combination of two things: the gold standard, which gave the dollar a physical anchor independent of the economy it was measuring, and the market itself, which carried the collective knowledge of millions of transactions and could self-correct in ways no central measurement framework ever could. When both of those left together around 1970, one removed by policy choice, the other gradually displaced by the growing belief that economists and policymakers could actively manage the economy, we were left with the recipe and nothing else.”
And the further we get from when Grandma left, the harder it becomes to remember what the cake was supposed to taste like. The people updating the recipe today never tasted the original. They are adjusting based on what they can measure, calibrating against a standard they inherited rather than one they ever experienced directly. The recipe keeps getting updated. The measuring cups keep shrinking. And the cake kept coming out tasting pretty good, until it didn’t. Over time, the cracks started to show up in the data: wage series that don’t match productivity, housing that is unaffordable even on paper, food that never seems to get as cheap as farm efficiency would suggest.
Pretty good is not the same as right. And the further we get from when Grandma left, the more the difference begins to show, not because anyone went looking for it, but because, after long enough, an error that compounds every year eventually becomes too large to explain away.
So What Are We Actually Measuring?
Step back and look at what we’ve established.
The ingredients keep changing. The ratios keep shifting. The entire recipe needs to be updated periodically to ensure we’re still measuring the right thing. The BLS handles all of that with genuine rigor and intellectual honesty. Those are hard problems, and they do a serious job of managing them.
But underneath all of that, running silently in the background for over a century, is a different kind of problem, one that has nothing to do with how carefully the recipe is followed. The measuring cups have been shrinking, and the improving quality of the ingredients has masked it so well that we’ve only ever had faint hints that something was off, nothing screamingly obvious. To be clear, no inflation measure can ever be perfect; money is an abstract measuring stick, not a physical rod. But some errors are random noise, and some are systematic drifts, and NETs is arguing we’ve had the second kind.
The result is that what we call inflation, the number that moves markets, sets interest rates, adjusts Social Security checks, and anchors almost every major economic decision in the American economy, is a net figure. It is monetary inflation minus the productivity dividend. And because productivity has been a persistent, compounding feature of the American economy for over a hundred years, that dividend has been silently subtracted from the measured inflation rate every single year.
That gap, whatever its precise magnitude, has been compounding silently for over a century. And the nature of compounding means that even a small consistent error, sustained long enough, produces an enormous divergence between what we think has happened to the dollar’s purchasing power and what has. It cascades outward into every inflation-adjusted figure in the economy, real wages, real GDP, real investment returns, and real Social Security benefits. Every conclusion we’ve drawn from those figures carries this error embedded within it.
This is not an indictment of the BLS. They have been measuring what their framework was designed to measure, and they have done so carefully and honestly. The error is not in the execution. It is in the foundation, a framework that was never built to account for a sustained, century-long productivity boom of this magnitude.
The cake kept tasting good. So nobody checked the measuring cups.
That is what Part 3 is for. And the conclusion might surprise you, because before we can say CPI is wrong, we must reckon with something that almost nobody in this debate is willing to admit. That CPI is working as designed. The problem runs deeper than the measurement itself.
Sources
U.S. Bureau of Labor Statistics. (2025, January 30). Consumer price index: Overview (Handbook of Methods). U.S. Department of Labor. https://www.bls.gov/opub/hom/cpi/home.htm
Author: Kyle Novack
May 8, 2026
A Monumental Venture, LLC: research project (Novack Equilibrium Theory – NETs)
Attribution Required: © 2025–2026 Kyle Novack / Monumental Venture, LLC. For educational use with credit; commercial use requires permission. Full details in linked PDFs.