Hauke Hillebrandt

Nov 2025 · Inkhaven

Increasing public sector salaries to increase state capacity and the 'Efficient Public Sector Hypothesis’

Open in Google Docs

People often say that public sector workers, civil servants, or politicians are incompetent and overpaid. But if the public sector job market is efficient, we have a trilemma, where only two of the following can be true:

  1. Incompetent i.e. not working hard or are incapable. In other words, their work has little benefit. I get that vibe even from smart people criticizing the UK civil service, the US FDA, CDC, WHO.
  2. Civil servants earn too much i.e. they earn an above market rate return and at their level of competence wouldn't make as much in the private sector. In other words, their wages have too high a cost.
  3. The public sector job market is efficient: Civil servants are directly or indirectly democratically appointed i.e. anyone can in principle get these jobs, and they are hired based on competence and not nepotism, even if the market does not set wages (as we disincentivize politicians' parties to hire/appoint poorly by not voting for them). This might be true because if public sector jobs were much easier i.e. you could get away with working less hard and/or being less competent for above-market pay, then word would get out and demand for these jobs would increase.

Public sector pay should be roughly cost-effective and public servants are not overpaid AND incompetent. Nor are they underpaid: if they'd be paid more for their skill set in the private sector, then generally they'd switch (except for idealists who sacrifice salary by working in the public sector and increase public sector efficiency).
Simply put, as taxpayers we get what we pay for. I call this the 'Efficient Public Sector Hypothesis'.

Evidence for this is for instance that many were surprised that Rex Tillerson, who had a very impressive corporate career and worked himself up to become the CEO of ExxonMobil, but then underperformed as US Secretary of State.

Why might we be biased to think the public sector is less competent than the private sector?

  1. In contrast to the private sector, which has a very well-specified aim (maximize profits), governments optimize for less well-specified aims (increase welfare of its citizens).
  2. Taxpayers see themselves as involuntary, unhappy customers of seemingly 'overpriced' government 'services' they can't cancel, with intransparent and complex 'billing'. They might believe that most of their taxes go towards politicians salaries. Tax to GDP ratios in rich countries ~40%. This might seem high and so taxpayers might think they're 'paying too much' and feel like they 'get their money's worth'. But few know that most taxes are spent on:

But even if the 'Efficient Government Hypothesis' is true, public sector competence can be miscalibrated over the long-term. But this does not happen in a vacuum and the government cannot improve itself i.e. pull itself up by its own bootstraps—without exogenous input and so we should not blame it.

Instead, we can only increase public sector efficiency by:

  1. Hiring fewer people (e.g. some economists have argued that spending on health or education is not cost-effective)
  2. Hiring more  (e.g. biosecurity): Take, for instance, the Covid response. Many have argued that governments have consistently failed. But consider that the the WHO's budget is only ~$2B / year (cf New York's biggest hospital's budget is $8bn / year) and the FDA has a ~$3bn annual budget and 15k employees ($200k / employee) in many different states and countries and a massive mandate. A governmental organization’s budget is generally not the organization's fault, except if the budget requested was too low or if their heads don’t tell us that they are too poorly funded to fulfill their aims. Ultimately, even if the budget requested is consistently too low to get the job done, then that just means that we are not paying the politicians in charge of budgets enough to prioritize correctly.
  3. Decrease pay (e.g. hire more cheap nurse practitioners (high $/QALY) instead of expensive brain surgeons (low $/QALY)
  4. Increase pay (increase civil service pay): Again, we might see lots of incompetence but, as per efficient sector hypothesis, it's not because civil servants are overpaid on an individual level relative to their competence (i.e. not worth their salt). Rather it's because we set their wages too low to attract qualified candidates (or haven't hired enough people).
    In the US, senators and governors make $175k. The president and the highest paid civil servant (chief medical officer) make ~$400k/y, the vice makes $260k. In the UK, MPs currently earn ~$100K/y. In contrast, median executive pay at the 300 of the biggest U.S. companies is ~$14M. Imagine increasing politicians' pay to $1M/y- 650 MPs would cost $0.65B. If more competent MPs were to counterfactually raise UK GDP by just 0.2%, then the cost-benefit ratio is ~10x. In the US, it might be ~10x higher.

So maybe we should: