Australia has Baumol’s disease

Why does the wage of a musician in a string quartet rise over time at roughly the same pace as wages in other areas of the economy, despite the lack of productivity gains in the performance of music?

William J Baumol solved this riddle in the 1960s.  His insight, known as Baumol’s cost disease, is fundamental to understanding changes in the economy over time.  If we are going to debate the shift towards a service economy, productivity, unemployment, health and education costs and government intervention in markets, we need to fully appreciate his insight.  Unfortunately you won’t find his ideas in many introductory economics textbooks.

Baumol’s disease can be explained simply.  Only some areas of production will see strong improvements in labour productivity, typically through the substitution of capital for labour (such as in manufacturing and agriculture).  Workers in these industries use their bargaining power to capture part of the increased profitability arising from productivity gains in their wages.

As a wage differential between these newly productive jobs and other existing jobs emerges, workers are faced with a choice of seeking employment in the high paying productive sectors or the lower paying unproductive sectors.  Employers in the unproductive sectors still need workers and the wages need to remain competitive with the productive sectors to attract them.  Bargaining between workers and businesses in each sector (including government sectors) equalises wages over time.  Thus productivity growth in just one area of production is shared amongst workers across the economy.

This means the cost of labour intensive services with low productivity growth will increase relative to other goods and services in the economy over time.  In addition, the amount of labour employed will shift towards service sectors with low productivity growth.  Ever wondered why all developed countries seem to have relatively large service sectors? This is part of the story (the other part is displacement of tradable industries to cheaper locations).

Baumol’s original article and subsequent works highlighted the importance of his cost disease to the provision of fundamental health and education services.  The crux of his argument follows from key the finding that increases in the cost of providing these fundamental services in the economy are unavoidable (the disease has no cure). Therefore tough decisions need to be made about whether these services are provided through an expansion of government provision, and the increasing share of the economy the government will necessarily have to take, or whether these social services can be provided through private enterprise, and the associated problems of gaming and moral hazard this may entail.

We can expand on these points to explore the relevance of Baumol’s observation to contemporary political debates.

One key area of public debate is the decline in manufacturing and the apparent loss of productive industries where workers ‘actually make something’.  While clearly one fundamental driver of this change is the shift of manufacturing offshore, another key reason is that the economy simply needs fewer workers in these areas because each worker is becoming more productive.

The recurring debate about the rise in public healthcare costs also needs to catch Baumol’s disease.  To make any progress in the provision of health care all sides of politics need to acknowledge that the cost of health care MUST rise over time relative to other goods and services if the economy is experiencing any form of productivity growth (noting the current temporary decline in productivity growth in Australia).  Budgeting for this cost growth in health care over time is absolutely necessary, particularly in light of the demographic shifts occurring.

Education is the other area where governments need to acknowledge that cost escalation is unavoidable.  While I don’t want to give governments an excuse to waste money operating schools inefficiently, but there are limits to productivity growth in this area.  Teachers will still teach classes of roughly the same size, irrespective of the computing and associated technology which supports them.

We can see the long term trends in the graph below. Healthcare, arts and recreation, and professional services have increased their share of the labour force substantially over the past 27 years.  Meanwhile, areas of the economy where productivity growth is high, such as agriculture, manufacturing, wholesale trade, and utilities, have required a much smaller proportion of the workforce over this period. Of course Baumol’s disease describes long term trends which act in parallel to patterns of regional displacement of tradable industries to cheaper foreign markets.  In the short run many other factors are also at play.

Unfortunately most economists fail to acknowledge Baumol’s disease when analysing changes in the economy.  You won’t find a mention of it in the media (just two results when searching Google news), and you will find plenty of confused ideological articles about modern poverty and flatscreen television.  How can someone live in poverty with a flatscreen television, a coffee maker and a DVD player? Simple.  These goods are becoming relatively cheaper over time, while other fundamental goods and services – health, education, food and housing – are becoming more expensive.  In fact, I could buy a flatscreen, a DVD player and a coffee maker for about the cost of four days rent or a week’s groceries for my small family.  This is an important consideration in analysis of the changing cost of living.

Sensible economic discussion needs to catch Baumol’s disease.  Without it we cannot properly analyse the influence of short term political and financial factors against a backdrop of structural changes in labour markets and prices as a result of productivity growth in some areas of the economy.


  1. Does this then mean that the wages growth in mining will cause an increase in unemployment elsewhere in the economy, as other sectors cannot meet the upward pressure on wages?

    • Not exactly. Baumol’s disease describes long term trends that wash through the economy through the cycle. If there is short term unemployment in non-mining areas there won’t be wage pressures in those sectors despite the wage differential.

  2. Elective surgery has become much more productive. Whether it is cataracts, hip replacements, hemicolectomies, or prostate resections, we can do about twice as many per day as we could do 10 years ago – a productivity growth rate of about 7% per annum (using the rule of 72).

    Better techniques and equipment also mean people go home earlier and return to work earlier, but probably not twice as early.

    However, the big expanding part of health care is aged care and convalescence, and productivity gains are very hard to make on this front.

    Just thought I’d throw this in, but this is a brilliant article – thanks.

  3. Macrobusiness is great. An article on Baumol’s disease, well done.

    Note that public transport is also something that will suffer from rising costs.

    It is also important to not hobble the “productive” sector, seeing it is what does “pay the way” for all the rest. The McKinsey Institute Report on Productivity in the British Economy is very helpful in this respect. Strict urban planning; costly, lengthy and prohibitive permission processes; and inflated urban land costs; reduce productivity in several ways.

    It is also important to realise that the source of all this wealth, is the utilisation of natural resources. Because there are fewer and fewer people actually involved with the utilisation of resources, most people end up taking wealth for granted, and mistakenly attribute lawyers, hairdressers and coffee-makers with being part of the wealth creation process.

    There is actually a wealth creation process and a wealth consumption process, both going on simultaneously. Some people can “create wealth” for themselves by offering a great service – but they are not adding to the total. The only reason people can create wealth for themselves by offering services, is that the fewer and fewer people who create wealth for the whole, by utilising resources, are doing it more and more efficiently.

    The madness that has seized NZ today, of leaving resources in the ground, regarding themselves as morally superior to Australia (“100% pure NZ” – a lie anyway), and then wondering why the brightest and hardestworking people all bunk off to Australia; is a classic case study in this effect.

  4. PhilBest, it is nice to finally agree on something.

    “It is also important to not hobble the “productive” sector, seeing it is what does “pay the way” for all the rest.”


    Although you lost me with the link to urban planning 🙂

    I don’t want to keep stoking the supply side debate but I will have to lay out my arguments more clearly again some time soon.

  5. Fascinating. Interested in developed economies move to services and reassignment of manufacturing capability offshore, something nearly all developed economies have experienced through globalization. I recently saw Dalia Marin (Chair International Economics Munich U) talk where she describes this shift in both the US and European countries. Discusses changes in skill sets that evolve but also notes there has been a growing shift offshore of ‘educated’ service jobs, not top end, rather the more mid-tier level, particularly in technical, engineering, design, management that goes with along with it, and increasingly legal work.

    In essence she was arguing there will be a continued shift of jobs, and increasingly in the ‘services’, offshore. She joked that maybe the only secure jobs would be those like cleaning ladies, caring for the elderly etc and high end managerial skills (whatever sector).

    Taking that further, it could mean that in the future many functions of government and corporations would be undertaken offshore, at substantially reduced cost. I wonder how the Baumol’s cost disease would operate then?

  6. Rising health and education costs (and falling standards) are only inevitable if run by a monopoly. However, this is not inevitable if run by a competitive private sector.

    • How do you figure annual cost increases of about 6% or more for the past decade in the competitive, private education sector, without any demonstrable improvement in productivity or education outcomes ?

    • If you don’t have symmetric (at least to some point) information for both sides on the market, you have a market failure. Health services are one of those services which have market failure. If you know what the services provider (and adviser – the doctor) knows then you can have efficient health service market. Unfortunately this is the same like the financial services. It is different from the market of other consumer goods. That is why it needs strong regulations, otherwise there will be more than one Dr Patel.

        • Yes, but if we had only private unregulated health sector, we would never know about Patel and all other patients who are under ground zero. because the doctors’ mistakes are always under ground zero.

  7. so i ask the question on areas that have been under govt control in recent times (health, education) how much of the growth in these areas is due to an innate lack of ability for productivity growth in these areas, as discussed above, and a fundamental trend towards this, or from bad, wasteful policies of govt at all levels?

    • That is the important question for these government departments to answer themselves.

      Remember, governments have always been quite wasteful. Perhaps more so now. So the change over time would be interesting.

  8. Thanks for this wonderful analysis and explanation of modern economic theory. Explains a topic fraught with ideological claptrap in a clear and concise way.

  9. “Education is the other area where governments need to acknowledge that cost escalation is unavoidable. While I don’t want to give governments an excuse to waste money operating schools inefficiently, but there are limits to productivity growth in this area. Teachers will still teach classes of roughly the same size, irrespective of the computing and associated technology which supports them.”

    I have home educated my children so I guess my non-participation in the “money economy” counts as a productivity loss however I can’t quite fit the inevitability of cost escalation in education with the success of web delivered teaching such as the Khan Academy and the recent “Stanford Engineering Everywhere” iniative both of which are offered at no cost to students.

    I have also purchased small class online teaching with daily individual feedback over a six week period at a cost of USD200 which gave my child training sufficient to enable her to be competent in writing first year university research papers. I am certain a school would have wanted a lot more time, at a lot higher cost for a far inferior service.

    Maybe we need to rethink the whole teachers/school model in order to make productivity gains. The creator of the Khan academy has suggested online demonstrations worked through as homework and classroom assistance for what is currently classed as homework.

    • Julienz, my son was also homeschooled for many years (now happily at secondary college). Thanks for the Stanford U links – he has used Khan Academy for some time at a backup to classwork or to explain more clearly concepts poorly communicated in class. It is a remarkable project.

      Education in the future must change. We are teaching kids similar methods to those used 100 years ago (move in herds, sit at desks, little open exchange of thought etc). I can clearly see secondary schools of the future offering most courses online with attendance required once or twice a week (tutorial sytle or to discuss assessments). My son says he can learn anything he wants online or via the library. I don’t doubt it – but a guiding hand may still be needed!

      This may mean a dramatic reduction in the number of secondary teachers or a reassignment of existing roles, a complete rejuvenation of the way school campuses are utilised (eg for broader community activities). Anything is possible.

    • “Maybe we need to rethink the whole teachers/school model in order to make productivity gains”

      Brilliant. That is very true, but classroom schooling as it is today is very habitual and cultural. Older generations think that it was good for them (mostly) so it must be good for their kids. Even though there may be far better ways to teach academic and social skills (/compliance?).

      If I recall correctly, the benefits from more flexible schooling was cited as justification for government funding for the NBN. Given time we will make better productivity gains, but perhaps still less than other sectors of the economy.

      “my non-participation in the “money economy” counts as a productivity loss”

      Not so. It is just that your above average productivity is not counted.

  10. I’m finding it really hard to correlate and distinguish the legend with the chart …. maybe just my laptop screen or am I becoming colourblind?

  11. Some asian countries avoid this disease by importing endless cheap labour. Imagine the cost advantage Singapore has in building its infrastructure vs Australia (and yet Australia could also enjoy this benefit !).

  12. I cant quite agree re the suggestion that education costs will see constant rises. Maybe in Primary and Secondary, but Tertiary should see (if the vested interests dont prevent this) a steady decrease for some time as much of what is taught in undergraduate courses can be provided via the internet, with lectures taped one time and all resources delivered for essentially nothing. The costs will still be in developing resources, and periodically updating them, and providing tutorial asistance via one one one video call interactions (this used to be done by standing outside a tutors office and waiting to catch him – often quite difficult). Other costs may be where lab work is required, using specialized equipment at the UNI. The gee wiz “super broadband fibre to the home” is exactly what is supposed to drive this outcome.
    If the same was applied to upper secondary (ie as these students are supposed to be motivated to work), we could have a situation where students only attend classes at school for 50% of the time, the rest via the web. In this way a physical school building could get double usage, a productivity gain of 100%.

    • I agree Ray that tertiary studies should begin to see quite good productivity gains thank to high speed internet.

      Primary and secondary schooling has an element of ‘baby-sitting’ / supervision which is why teacher student ratios can’t be dramatically improved.

      But remember that all productivity gains are relative. While we might see a short term boost in education, we may very well see similar gains in other areas for similar reasons.