Hacker News Reader

Comment by llamaimperative | Hacker News Reader

Fun fact that a lot of people find surprising: the US federal government was already at or near its smallest size (headcount per capita) in American history.

It’s honestly shocking that so many public services work under such resource-constrained conditions at all.

View on HN

This is surprising to me, but I also wonder how contractors factor into the equation.

To naively soapbox for a second, I don't know much about public policy, but I have a persistent hunch that a lot of the "government inefficiency" comes down to the obsession with hiring contractors for everything. While I know some people in gov't contracting and they do good work (and definitely faster than I've seen get done in federal agencies proper), there has to be so much inefficiency in the whole process, so many unnecessary layers added on.

It's one explanation (albeit obviously an incomplete one) I've heard for how we used to, say, build infrastructure so quickly, even for all the flaws that had - in Robert Moses' time, when the government wanted to build something, they would go down to their office full of engineers and draftsmen and get it sketched out within the week. Now they have to go through a whole process of hiring a contractor for design, then construction, etc, etc...

Anecdata: here in California, when CalTrans needs road work done in a hurry, e.g. disaster recovery, they bring on contractors that are paid bonuses for completing their jobs speedily (hooray, incentives). When time is less of a factor, they do the work themselves.

This is the key point, most of the waste comes from contractors who can navigate the insane contracting process and the fact that we’ve brain drained the federal government to the point that the Feds just take whatever contractors say as gospel.

Cutting federal employees further isn’t touching the primary problem at all.

That's an interesting metric. Do you have a citation?

It's just the bare facts from the Bureau of Labor Statistics. This WP article has a graph -- basically, at its height (the New Deal/WWII) the federal workforce was about 7% of the workforce and have been going downward since then to about 2% today.

https://www.washingtonpost.com/business/2025/02/12/federal-w...

Thanks, interesting. I wonder how much of that is the progression of women entering the workforce. It's quite a specific metric to use.

I suspect a fair bit of it is the elimination of Federal make-work programs established during the New Deal.

IMO that's a poor measurement of "government size". Not that I have a better one though. But it doesn't account for technology or scaling.

I also like expenditures as % GDP. It was generally far below 5% federally during non-war periods between civil war and ww1, and for most that time no income tax for regular workers. We also had pretty solid growth in that period, pretty close to that of the post WW2 boom.

https://pbs.twimg.com/media/GjNDwxtXwAAYvL5?format=png&name=...

It indicates that reducing headcount is almost certainly a poor way to "increase efficiency."

Im curious in which ways is that a valuable way to look at it, opposed to headcount per road mile, land managed, package delivered, ect? I would expect some reduction in headcount required as we dont need to hand sort every letter, ect.

In a different way of looking at it, federal taxes have outgrown inflation by 10x if you benchmark 1950. In what area are 10X more services provided, and in what areas are 1X services provided.

I've always been perplexed at the assumption that federal spending should track GDP.

Lots of things. Neither Medicare (1965) nor Medicaid (1965), which provide healthcare to millions of impoverished, disabled, and elderly people, didn't exist in 1950. Nor did NASA (1958) and the US space program. The NSF (1950) which funds most non-biomedical research in the US, was founded in that year, and while the NIH technically existed in 1950, it was mostly just the research institute in Maryland, not the funding agency it also comprises today. DARPA (1957), the defense research and funding agency that among other things created the ancestor of the very network you are using to post here, also didn't exist. You could go on and on.

We don't need to sort every letter but we need to deliver them by hand. And packages, etc.

How many people are required to run the government compared to the size of the population they serve is absolutely a valid metric and makes tons of sense. Not only comparable historically but also across countries.

Per Capita is per person, not GDP btw.

I obviously know what per capita means.

My point is interrogating the domains where that metric is useful and isn't. Are we delivering 10x more letters per capita? Does US mail sent use scale with GDP per capita?

In many areas I would expect a 1:1 relation with population, but not GDP. For example, I would expect primary education cost to scale at with population (eg teachers and schools per student). If cost per student has gone up 10x instead, why is that?

Reply on news.ycombinator.com