Hacker News Reader

Comment by stormbrew | Hacker News Reader

These things always really need a giant flashing neon note that "productivity" doesn't mean how much workers get done but how much money is made off of what workers get done. They're only loosely connected, and most productivity gains have come from workers having to "do less" to "make more".

View on HN

Weird story: From the mid-1970s to the mid-1990s, as-measured productivity significantly declined and then stayed at a lower level. This was during the initial few generations of technological impact on industry, including "just-in-time" inventory which kind of requires computerization. Yet, at this same time, "bosses and economists" were seen in public wondering if computers weren't a net negative on industrial production.

In addition to being weirdly defined, productivity is, as the graph demonstrates, very unstable over the short term.

If you want a longer version of the graph in the article, see "The 1990s Acceleration in Labor Productivity: Causes and Measurement" from 2006 (https://files.stlouisfed.org/files/htdocs/publications/revie...), page 190 (10 of 22).

> "just-in-time" inventory

That really was a net negative eventually. Covid managed to completely wreck the worldwide supply chains because of that idiotic approach. God forbid anyone keep any buffer in case anything happens.

It's definitely not an idiotic approach, and companies do keep buffers. In the most advanced cases, probabilistic models are devised to estimate how big these buffers should be. Asking for companies to keep buffers for unpredictable "once in a century"-type events is unrealistic.

Yeah that's the problem, those estimates are usually arounds zero it seems. I mean sure on paper it checks out to be most profitable and most of the time it also works in real life. But you end up with this rube goldberg supply chain machine that can't be stopped or you apparently end up with a cyclic dependency problem and you cannot restart your production.

We build structures to take a one in a ten thousand year flood or earthquake, but it's too much to expect corporations to keep more than 2 weeks of stock? Sure.

Efficiency argues for keeping those buffers as small as possible. In a series of "normal years", a business that keeps their inventory down to what they need in the "normal case" will be more efficient than one which keeps a larger inventory to handle more rare events. As a result, the former will out-compete the latter in the market. (Inventory, in this example, is just one "buffered" resource that needs to be managed correctly under differing circumstances.)

Efficiency, past a point, is therefore the enemy of resiliency.

Until that day that something bad happens.

And then you have an issue where one business may be prepared for the bad event, but something downstream of it is not; they can produce all the widgets, but can't ship them anywhere for example.

Also, to some degree, in events like these what inventory is needed is not predictable, because humans are unpredictable and irrational at times.

Panic buying of paper towels early on in the pandemic, as an example, was not predictable.

> Panic buying of paper towels early on in the pandemic, as an example, was not predictable.

What? It was entirely predictable, except maybe for those high-end executives who live in hotels and/or with full-time house service, and are thus completely detached from life of ordinary people.

The panic buying was limited to the most obvious category of goods: basic consumables with some degree of shelf life, prioritized by survival and then comfort. This means food (particularly canned, shelf-stable, or freezable - plus baby food and pet food - and ingredients, including flour, yeast and baking soda), hygiene (soap, toilet paper!, and - perhaps specific to pandemic - hand sanitizers and masks), fuel, household cleaning (including cleaning agents and, surprise surprise, paper towels), comfort consumables (coffee).

It's not hindsight on my part - think of how people individually decided what to buy. They didn't buy whatever they see everyone else buying. They just asked themselves: what do I eat? How do I keep myself and the household clean? What else do I buy on a regular basis? What do I need to maintain a semblance of my current lifestyle? The answers, prioritized, were what everyone then went to stock up on.

That there are downsides is not in dispute, but you’ve not shown that it’s a net negative. Perhaps it is, perhaps the shortages we experienced because of a very unusual event like COVID outweigh all the day to day advantages of having little slack in the system. But you haven’t shown that.

>From the mid-1970s to the mid-1990s, as-measured productivity significantly declined and then stayed at a lower level*

*in the USA.

The post-bretton woods era is one of globalization, with American jobs being sent overseas (to more productive labor forces)

> "productivity" doesn't mean how much workers get done but how much money is made off of what workers get done

This is so true. The amount of bureaucracy has actually increased. This makes every worker work more. But this bureaucracy is unproductive work, thus does not lead to a rise in income (for the company).

E.g. my healthcare provider uses fax machines (yes that FAX) to communicate with insurance providers. Fax is asynchronous and without confirmation/tracking of work done. Often, the fax is sent but the other side simply files it in a random place or forgets to process the work. So, I (the patient) now needs to follow up for weeks with insurance and healthcare provider to check on the status of that FAX.

This is unproductive work and yet, it is taking a toll on every individual involved in this process.

But it increases GDP! Yay!

It doesn't increase GDP if they get less productive.

Many things increase GDP without being good. For instance Oil spills increase GDP as suddenly a bunch more (cleanup) work is being done and paid for.

There's an old joke about an "economic hero" being a wealthy man going through an ugly divorce while dying of cancer.

see https://en.wikipedia.org/wiki/Parable_of_the_broken_window

It's even stupider than that: it divides this figure by a largely fabricated estimate of how many hours people actually worked. This is a SWAG metric that is largely made up. The commentary is most likely irrelevant.

This is not how productivity is defined or measured.

I'm really curious how you think it's defined or measured then. I'm obviously abstracting a bit, but a lot of people in the replies here seem to think it's related to how much time you spend watching cat videos on company time and it's definitely not that.

It has nothing to do with what I think. Another commenter has already posted the definition from the Fed (FRED).

Productivity is not a measure of prices, profitability or anything similar. It’s a measure of output per unit of input.

Classic HN downvote fest because people incorrectly disagree with a factual post.

You were probably downvoted because you made a curt statement with no backing arguments like in this follow-up post.

from FRED: "The efficiency at which labor hours are utilized in producing output of goods and services, measured as output per hour of labor."

The solow residual is technically total factor productivity but is generally accepted as labor productivity. it's just an accounting identity that is estimated along with GDP and other vaguely useful but not very accurate measurements like the unemployement numbers.

Then perhaps you'd like to enlighten us?

Workforce productivity at the national level is typically defined by some measure of output, the amount of goods or services produced (typically GDP), over some measure of input, the number of hours worked/workforce participation.

So, how much money is made off of how much workers work?

No, not at all: for example, you can be a non-profit and still contribute to GDP, since you're still creating economic activity. Heck, even what the government does contributes to GDP, and that's not making money for anyone.

I think this is exactly what parent meant. "How much is made" doesn't strictly mean "profit" in terms of a for-profit institution. The net output of a non-profit is directed somewhere, either internal or external to the entity, and that can loosely be considered "making money", or at least in the sense I believe parent meant.

GDP is denominated in dollars, so this seems to be a somewhat vacuous position - yes, that's how we measure economic activity, but it doesn't have to involve money changing hands.

Productivity is based on the value of the work done, not any profitability assessment. The original post which set off this chain asserted it was about not how much workers get done but how much money is made off of what workers get done ... which is unambiguously wrong.

> The original post which set off this chain asserted it was about not how much workers get done but how much money is made off of what workers get done... which is unambiguously wrong

I don't believe the difference is consequential here, since the originating point still holds even using your definition. I wouldn't say it's "wrong" so much as imprecise, as the way I interpreted the statement would encompass your more detailed description.

It's like when I ask people "how much money" they make, I intend them to include non-cash compensation in the number (in dollar equivalent), and pretty much all do without additional prompting.

Yes. My point was that when people read these articles they think of a much more casual definition of productivity that has more to do with a sense of "getting things done," but the word is jargon for something that has little to do with that.

I was playing loose with the jargon meaning for sure, but I'm pulling out to what articles in the Washington Post or other economics-focused media really care about: the impact to corporate bottom line.

So, how much money changes hands off of how much workers work?

No, it's a measure of a weighted average quantity of output of goods and services (not of money) compared with the quantity of labor input.

"Quantity of output goods" is measured in dollars, as mediated by the current price, no?

Not in general, no. They do count the dollars, but they also measure the dollar-to-quantity ratios of various goods and services. The final productivity measure is based on these adjustments.

So if the amount of money that is exchanging hands goes up but the amount of goods and services produced stays the same, then the measured productivity does not go up.

You may be thinking about how GDP is calculated, specifically regarding government employees. For this category of spending, the "quantity" measured for the dollars-to-quantity ratio is simply the number of government employees. So as long as the government is hiring more people, the money they spend on those people counts towards real GDP, regardless of what those people are doing.

However, government spending is not used in calculating productivity, which measures only certain parts of the private sector where it is possible to also measure output of goods and services instead of relying on measures like 'employee counting'.

If this is a good faith question, you can answer it by reference to any good economics text; or even, you know, Wikipedia. But it feels like maybe it's not.

Here's another example: you volunteer at a homeless shelter, where you serve food on a soup kitchen line. You have contributed to the GDP of the United States. By all means, feel free to fit this into your preferred framework.

What I'm doing is putting how you describe productivity into the form from the original comment. And I'll stand by that last version: How much money changes hands off of how much workers work?

The input to labor productivity is how many hours are worked, correct? And no one is measuring output in terms of the number of bowls of soup produced by homeless shelters; those are converted to dollars based on a current index price.

So you have economic activity, how much money changes hands, compared to labor inputs, how much workers work. Simple?

> And no one is measuring output in terms of the number of bowls of soup produced by homeless shelters; those are converted to dollars based on a current index price.

The Bureau of Labor Statistics does have multiple teams dedicated to documenting how price is related to quantity of output. They don't literally count bowls of soup at every homeless shelter, but they do document millions of price vs quantity measurements on a regular basis. This data is then used in the calculation of productivity.

Your fundamental confusion is that you keep equating economic activity with "money changing hands", which is wrong.

Ok, here's a question for you: back in ancient days, most women worked at home. Is taking care of your own children and your own household an economic activity as would count in GDP, for example? Is someone making a bowl of soup for their spouse different from someone making a bowl of soup at a homeless shelter?

"GDP measures the market value of the goods and services a nation produces. Unpaid work that people do for themselves and their families isn't traded in the marketplace, so there are no transactions to track. ... The lack of reliable data influenced the decision to leave household production out of GDP in the internationally accepted guidelines for national accounting." (https://www.bea.gov/help/faq/1297)

"Economic activity" that does not equate to "money changing hands" in some form, isn't an "economic activity" that counts for GDP or productivity, right?

Yeah I don't think this is true. GDP is an economic measurement, not some kind of intrinsic thing.

So you can make a beautiful thing for your home - not GDP. Make it and give it to someone - not GDP either.

Pretty sure money has to change hands, or in the case of government, we measure it as $ spent.

Indeed, the value of volunteering is not counted in US GDP.

This is just factually incorrect. Volunteering at a homeless shelter does not contribute to US GDP.

Reply on news.ycombinator.com