Endless Measuring
Measuring everything is inevitable, starting with billing purposes. Weighing scales are used to measure out food by weight. The Bundy clock at work measures how many hours you’ve worked between clocking in and clocking out. Electric meters and water meters sit at the inlets to our homes and businesses, measuring our consumption. Physical things must be measured in order to ration them or sell them – perfectly natural.
This drive is easily extended to things that are less physical and far more subjective. Likes, comments, engagements, and impressions on social media are a measure of how much people liked seeing what you put out. Incomes are evaluated based on salaries and revenues, ignoring non-taxable benefits and individual circumstances affecting the cost to earn said incomes. Most notoriously, governments are measured on economic growth – James Carville’s “it’s the economy, stupid” is a nugget of realist dogma, no matter how many confounding factors there are in the relationship between the ability of the government to the statistical construct of Gross Domestic Product[1].
This tendency is stretched further into even totally subjective worlds. Getting to know others has come to resemble a sexualized job application – measurable characteristics such as height, weight, incomes, educational and professional standing, and plenty of other criteria being used as filters to evaluate potential friends and love interests, to quickly sort through a large pile of candidates in a world with limited time[2].
Questions governing one’s whole life – where to live, whether to marry or not, what to do with one’s life, what to seek – all of these should ideally be answered as soon as possible, so that time can be put towards achieving these tasks. Reflection, repose, or thinking too hard about such things eats up valuable time – time that could be spent working, not dreaming. After all, unlike thinking, you can count time.
You may ask at this point how we can decide what is better if we do not have a clear vision of what is good. The answer predominant today seems to be to break it down into measurable, actionable steps in order to provide you the resources to do what you need to do. The ability to change, called option value or optionality – in the forms of money, time, mobility, and freedom from responsibility – is considered the ultimate good in absence of a goal.
Optionality neatly substitutes itself - a measurable good for a good life - an immeasurable goal. Optionality seems to be the default metric used in order to measure the goodness of lives, particularly the lives of others in absence of context or thought. You may recognize subsidiary measures such as wealth (which allows you to afford options) in the form of cash (spending power) or multiple homes and passports (options in location). Extensive travels and unique hobbies showcase options in place and an abundance of time, further reinforcing the high option value that one chooses to exercise. Even in careers, the mobile, work-from-home coder who can hop from corporation to corporation makes more and gets a better reputation than a tradesman or lawyer, tied down as they are to physical presence, building codes, and jurisdictions. Having options is just plain cool, and we love it.
This endless jostling for options and rejection of commitment is the foundation of modern progress and scalability. Labor and management are trained in standard processes to be interchangeable. Professionals in accounting, engineering, and law are bound by financial reporting standards, national and international codes, and jurisprudence, all pushed towards a predictable and standardized output with minimal room for judgment. In the same way that factories apply standard inputs and standard processes to produce standard outputs, we are trained to be mental sweatshop workers – applying standard metrics to evaluative standards to produce standard evaluations[3]. Management consultants with their best practices, lobbyists with their think tanks, and thought leaders with their non-profits are the bosses of these mental factories, making sure we’re all reading from the same song sheet.
The Industrial Revolution produced the physical factory. The Managerial Revolution produced the mental factory. As Peter Drucker’s famous saying goes[4], “what gets measured, gets managed” – meaning two things; managing something requires measuring it, and once it’s getting measured, that metric is getting managed. The former, we have already covered – we bring life under control by making metrics and managing to them. The latter, we will go into.
The World of Managed Metrics
Let’s start with that most managed of economic metrics – inflation. Inflation is the general rise in prices of goods, or the depreciation of money against other goods. Inflation is used widely as a barometer of the economy – low and stable inflation is considered healthy, while too much (hyperinflation) and too little (deflation) are bad in their own ways, which makes it a perfect example of a managed metric.
Inflation is most commonly measured with the consumer price index (CPI). The CPI is calculated by taking a representative basket of goods and services, noting down their prices, and tracking them over time. The monthly changes in these prices, taken together, are reported as headline inflation. For a quick example, let’s imagine you spend one third of your money on rent, one third on food, and one third on fuel and utilities. If rent were to go up by 10%, you would experience 3.33% inflation (since rent is 1/3rd of your budget).
The CPI has two main uses – it is used to gauge the health of the economy by tracking the general movement in prices. Low and stable inflation is preferred because it implies economic growth without price instability – the supply and demand for money and goods and services are moving together, upwards, in harmony. Inflation being too high or too low means the markets for money and goods are desynchronized, while unstable inflation means that these markets are not coordinating.
This metric is also incredibly managed because of its importance. As an example, let us take the interest rate – the rate set for borrowing by a country’s central bank, which is the basis for the rates of interest paid by banks on deposits. Raising this interest rate raises the interest rates banks offer, increasing the demand for money and reigning in inflation. Lowering this rate works in reverse. This is why the Interest rate and inflation are always together in the news – the interest rate is usually considered the primary policy tool for fighting inflation.
This management extends to definitions and replacements. Over time, products and services are added to and subtracted from the representative basket. Replacing things like videotape rentals with CD’s or Netflix subscriptions might be a no-brainer, but adjusting for different cuts of meat or different brands of frozen pizza might skew things one way or the other. How much house should the average person rent – some might prefer 300 square feet of condominium per head, while others would like 3000 square feet of home and backyard per family (750 square feet per head on an average family of 4). Choosing replacements and definitions should be a matter of best estimate, but even the best estimates are overgeneralizations, and that opens the metric up to even more manipulations.
Eventually, the CPI was made not to track a fixed basket of goods, but a fixed level of purchases – after all, with so many different goods available in so many different places, it just isn’t practical to do it any other way. This created a lot of missing data – both due to the sheer number of data points needed and the peculiar specifications of different products. In the face of missing data, the Bureau of Labor Statistics (BLS) sometimes imputes the prices of these missing items. Hedonic regression[5] is used to estimate prices for products captured by the CPI, adjusting prices for the quality of the objects on offer. Clothes made of different blends of expensive cotton and cheap polyester, for example, would have different prices in the CPI, and hedonic regression allows the BLS to consistently estimate the different prices of different blends of fabric.
It also lets them cook the books, reporting a lot lower inflation than what the consumer may actually be feeling, all because the new products and replacements are “better” somehow than the previous ones.
Shadowstats is a website that critiques government statistics and publishes their own versions based on older, more conservative methods of data gathering. Traditionally, the CPI was based on a fixed basket of goods and services, with the same components and weighting, year after year[6]. This would maintain your standard of living, since you consume the same goods over time, no matter what. Real life, however, is not so stable, and the fixed basket was becoming less and less relevant. In the 1990s, politicians and economists moved to adjust the CPI by adopting a constant level of satisfaction – substituting items in the representative basket between periods. This is a looser standard allowing more substitutions and adjustments in the CPI, lowering the level of inflation it measures by adjusting the quantity and quality of goods that are relatively too expensive downwards.
The potential for manipulation here is clear. This approach allows a lower reported inflation, boosting the health of the economy, justifying slower wage/salary growth, and lowering inflation-indexed payments like Social Security. By reducing these expenses, government can push to balance the budget on the sly while proclaiming economic health and strength even as the whole thing rots from within.
The metric has been managed away, leaving only message.
The Statistical Lie
This is the power of metricization and the statistical lie. These “objective” measures allow people, institutions, and governments to push messages and ideas on an industrial scale, with all the objectivity implied by the industrial-technical complex that produced the definitions. This mass-produced lie gains credibility from the complexity of its definition and the scale of its production, while using those very things to conceal the fact that it’s not true. Entire bureaus of government, departments of corporations, think tanks and lobby groups alike, function as mental sweatshops, where highly educated and credentialed people spend incredible amounts of time and money designing the latest models and ideas – the additions and expansions to the apparatus of statistical lies by omission, managing and massaging metrics to maximize advantage.
I don’t want to be too harsh on lies. Lies exist because things are too complicated to explain in their entirety - the same way that The Lord of the Rings doesn’t include every meal, every night watch schedule, and every bathroom break Frodo took on the way to Mordor, and is a much better and less truthful story for it. In addition, everybody lies – from white, polite lies at social events, statistical lies as we’ve covered, or just lies, like kids caught with their hands in the cookie jar. Sometimes, lies can even be helpful – the statistical lie is a quick explanation of complex phenomena that gets us all on the same page and moving in a particular way. It is the scale and quality of these lies that is off the charts - allowing us to live in a completely different world as measured by these metrics. Of course, real life asserts itself eventually, but it is far more tolerant and less defined than simple numbers.
It’s a lied, lied world out there.
[1] This is evidence of a straightforward bargain between people and their government. The government assumes responsibility for people’s economic and physical security and stability (represented by GDP and peace and order), and we agree to their taxes and laws. Government is a benevolent vampire. https://argomend.substack.com/p/vampire-society
[2] Everyone knows about six feet, six figures, and six inches, but let’s have some less common examples. Here’s the Chinese marriage market. https://www.aljazeera.com/gallery/2013/4/6/finding-a-spouse-in-a-chinese-marriage-market
[3] Longtime readers will recognize this idea as Global Paradox 2023 – too much information makes us want to use it less, rather than more. https://argomend.substack.com/p/global-paradox-2023
[4] Which he apparently never said. That’s the Harvard Business Review for you. https://medium.com/centre-for-public-impact/what-gets-measured-gets-managed-its-wrong-and-drucker-never-said-it-fe95886d3df6
[5] See page 25 of the following PDF: https://www.bls.gov/opub/hom/pdf/cpi-20180214.pdf
[6] This and the next sections are sourced from Shadowstats’ critique of the CPI. https://www.shadowstats.com/article/no-438-public-comment-on-inflation-measurement
Argo-Thanks for sharing this. I particularly love the journey you took for the reader on things like Industrial Revolution, Goodhart's, and on initiating a review on standards of measurement. Great read. Hope you're well this week. Cheers, -Thalia
This was just fantastic Argo - a really deep and super well written piece on the perils of metricization. Of course, my favourite part was your discussion on optionality :) I never considered how wide-reaching Goodhart's Law actually penetrated until you laid out your piece like this. I particularly also enjoyed your conclusion - never thought of metricization adding layers of lies to what is actually real and true - but you're so right.
What do you think are solutions to pathological metricization, if any?