Here in the UK, a lot of activities that were traditionally the responsibility of the state (health, education, policing) are increasingly being outsourced. Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this. It always seems to taken as axiomatic. Can anyone point me to unbiased work that has been done on outcomes in outsourcing of state functions?
As someone who works in the private sector, I find the idea that it is just intrinsically more efficient than modern state organisations to be pretty hard to believe. But thats just anecdotal, and I'd like to know more and maybe have my opinion modified.
Stated more correctly, smaller, more focused, organizations outperform larger, more bureaucratic(more management) organizations for particular tasks.
It's not that governments are outperformed by the private sector every time for everything (some things require size and longevity). Conceptually, it has nothing to do with whether an organization is private vs public. It's about incentive, regulation, and consequence horizons.
Incentives
Generally, private organization's incentives (if tied appropriately to the desired outcome) are more aligned and offer higher rewards. Government organizations, if completely transparent, have a hard time justifying high rewards to their constituents. All other things being equals, people prefer higher rewards. So, the talent tends to accumulate in the private organizations. There are some exceptions to this (political influence), but I'm generalizing. :)
Regulation
The smaller the organization, the fewer rules. Not only do laws tend to work this way, but internal organizational rules work this way as well.
As organizations grow, managers try to scale their ability to influence direction by legislating organizational rules. As more and more of these rules come into play, the overhead and unintended consequences of these rules grow. Since these managers are unable to directly participate in the execution of the tasks (the reason they created the rules at the start), they are slow in detecting changes that should require rule changes.
The overhead and lack of agility created by both size and expanded regulation(rules), makes large organizations far less efficient than smaller organizations. Unfortunately, governments have disproportionally larger sets of rules. Not only is their ruleset larger, and their employment not aligned to performance, but they also have unionization that has a differing set of goals. This creates even more operational friction making it very difficult to allow people to do what they do best. Tragically, this makes efficiency and productivity in large government organizations nearly impossible.
Consequence Horizons
Smaller organizations tend to have better alignment between their existence and their ability to execute. This is because they are paid for execution and are typically unable to structure longer term contracts. As a result, if the people paying the bills are not happy, they cease to exist quite quickly.
As organizations grow in size, they start to manage for risk avoidance and not for execution. The larger they get, the more they are able to influence the market to avoid direct competition, to change the criteria by which they are judged, and to structure contracts such that they are far more stable over longer periods. This 'cushion' allows these organization to ignore or tolerate consequences to a higher degree. As a result, they are less sensitive to market judgements about their efficiency, productivity, or ability to execute. They loose sight of that as a goal and start to engage in higher order goals (happiness, employee well being, brand, reputation, influence, etc). These are not bad necessarily, it just makes them less efficient at particular tasks.
In particular, large government organizations are disproportionally disassociated from consequence. This is directly related to election cycles, politics (spinning the message to create success out of failures), and constantly getting inexperienced leadership. The first two are points are obvious, but let me explain the last. New politicians are elected and they appoint the people that they feel are best to run these large organizations. Typically, neither the politician or the person appointed has never spent a day working in that organization or executing the tasks for which they are responsible. With large organizations, it takes truly special managers to make an organization efficient at something you have never done yourself. This is even more difficult when you are not answering to your boss for your ability to execute, but instead are answering based on the ebb and flow of the political landscape.
In general, my point is, that you are right. Large private organizations do not have a significant advantage over large public organizations. That advantage can be widened based on political systems. However, the real advantage that people reference when making this argument is the enormous advantage small to medium private organizations "enjoy".
Understand that I am only talking about efficiency and execution and not societal value (differing goals). I also recognize that there are many tasks/goals that require larger organizations and more longevity.
> Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this.
This is hard to measure empirically because of selection bias - you can only study things which have already been privatized, and the first things to be privatized are the things you think will privatize successfully. For example, there are quite a few countries that have successfully privatized their telephone networks. But that doesn't mean you could privatize the police or military with the same success!
I think this is a case of the usual issues of in-sourcing vs outsourcing. I know a large insurance company that flip-flops on their decisions every 3-5 years. When in-sourced, staff are "lazy, cost too much, can't be fired due to HR rules, etc, etc" so they outsource, at which point the outsourcer becomes "they are not responsive, too hard to work with, every little items needs to be tallied against the contract, every little change requires too much time and requirements specification, etc, etc" so they turn around and in-source again, to begin the cycle over again.
In theory, in-sourced allows for better flexibility on what work is being performed (you are already paying the dollars for hours, so you can allocate as you desire) but you may lack expertise, less flexibility to expand then reduce/contract the workforce (which is especially true with governments and their unions), too hard to get rid of people if they do not grow skills and adapt as the organization changes, etc. Outsourcing works well when there are well defined boundaries, and requirements to expand/contract the workforce (within reasonable limits, the problem here is that the Olympics are too big for an outsourcer to deal with because there are not enough staff in reserve).
None-outsourced government IT projects aren't any better.
We decide to install system X so we need to hire some experts in system X. But system X experts are expensive and if we hire them as staff they are paid according to fixed internal pay scales - or there is a hiring freeze - or we have to hire 'internal candidates" from other government branches first.
So we hire consultants. These are a lot more expensive. Then part way through the project there is an investigation into how much "external consultants' are costing. So there is a ban on external consultants which the government shows has saved us $$$ million.
The result of course is that we have a half-implemented system X that costs us $$$$million and doesn't work.
> Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this.
The train of logic is that if coprorations become efficient to maximize returns on investments and products sold, then since government programs have neither a need for a return on investment and exist outside of market economics because all the associated costs are a forced payment by the "customers" that they have no reason for themselves to be efficient. The government as a whole might want them to be efficient, but there is no intrinsic benefit for a government program to maximize efficiency because that has no influence on its short run continued existence.
Not saying it couldn't be as efficient, but large programs end up with too many people in positions of power which means there's probably at least one bad apple in there trying to milk the lack of demand economics or operating costs.
It's the same as buying any other sort of goods and services. Do you think the NHS makes its own paperclips and grows its own potatoes? No, the private sector really is massively more efficient, for two absolutely crucial reasons: 1) there are actually many sellers with interchangeable products and 2) paperwork and catering aren't actually core to the NHS mission.
If 1 and 2 aren't true, if there are a limited number of sellers of the service, and the service is core to the mission, then outsourcing will certainly fail.
The majority of government outsourcing deals are ideology or to hide the numbers.
Take PFI, if I offered to take over the rest of your mortgage on your house. But you had to pay me equivalent of the mortgage payment for 100years and you had to buy my services for any cleaning or maintenance of your house for the 100year term - then even though you were now mortgage free (ie. you had reduced your deficit) - you would be mad to take the deal.
If however your only interest was being able to claim that you had reduced your deficit going into the next election - and it wasn't your money anyway - you might look at it differently
During my time in the USA, I saw that most people think the private sector is inherently more efficient than the government. As you said it seems like an axiom.
However, I disagree it's people+procedures+technology that make a great organization and none of those concepts are exclusive of either private or public sector.
The situation is not as black-and-white. The private sector is more efficient in some cases, but that doesn't mean _everything_ should be privatized.
A trip to the DMV (Department of Motor Vehicles) will probably leave anyone with the impression that government offices are dens of inefficiency and bureaucracy. And that may as well be.
I think the sweet spot lies in the middle. There is no reason, for example, for the government to put a (say) landscaper on its payroll to maintain a public park. The overhead of a FT government employee is just too much; the job can be done well by a myriad of private companies. Same goes with bus drivers, janitors, painters, etc.
If the job requires domain knowledge and other such intangibles, then it usually can't be outsourced. This is why many IT projects designed to "upgrade" or replace existing IT systems at government agencies suffer: there are a lot of intangibles that just can't be all laid out on the contract form.
Comments
Here in the UK, a lot of activities that were traditionally the responsibility of the state (health, education, policing) are increasingly being outsourced. Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this. It always seems to taken as axiomatic. Can anyone point me to unbiased work that has been done on outcomes in outsourcing of state functions?
As someone who works in the private sector, I find the idea that it is just intrinsically more efficient than modern state organisations to be pretty hard to believe. But thats just anecdotal, and I'd like to know more and maybe have my opinion modified.
I think people often confuse the axiom...
Stated more correctly, smaller, more focused, organizations outperform larger, more bureaucratic(more management) organizations for particular tasks.
It's not that governments are outperformed by the private sector every time for everything (some things require size and longevity). Conceptually, it has nothing to do with whether an organization is private vs public. It's about incentive, regulation, and consequence horizons.
Incentives Generally, private organization's incentives (if tied appropriately to the desired outcome) are more aligned and offer higher rewards. Government organizations, if completely transparent, have a hard time justifying high rewards to their constituents. All other things being equals, people prefer higher rewards. So, the talent tends to accumulate in the private organizations. There are some exceptions to this (political influence), but I'm generalizing. :)
Regulation The smaller the organization, the fewer rules. Not only do laws tend to work this way, but internal organizational rules work this way as well.
As organizations grow, managers try to scale their ability to influence direction by legislating organizational rules. As more and more of these rules come into play, the overhead and unintended consequences of these rules grow. Since these managers are unable to directly participate in the execution of the tasks (the reason they created the rules at the start), they are slow in detecting changes that should require rule changes.
The overhead and lack of agility created by both size and expanded regulation(rules), makes large organizations far less efficient than smaller organizations. Unfortunately, governments have disproportionally larger sets of rules. Not only is their ruleset larger, and their employment not aligned to performance, but they also have unionization that has a differing set of goals. This creates even more operational friction making it very difficult to allow people to do what they do best. Tragically, this makes efficiency and productivity in large government organizations nearly impossible.
Consequence Horizons Smaller organizations tend to have better alignment between their existence and their ability to execute. This is because they are paid for execution and are typically unable to structure longer term contracts. As a result, if the people paying the bills are not happy, they cease to exist quite quickly.
As organizations grow in size, they start to manage for risk avoidance and not for execution. The larger they get, the more they are able to influence the market to avoid direct competition, to change the criteria by which they are judged, and to structure contracts such that they are far more stable over longer periods. This 'cushion' allows these organization to ignore or tolerate consequences to a higher degree. As a result, they are less sensitive to market judgements about their efficiency, productivity, or ability to execute. They loose sight of that as a goal and start to engage in higher order goals (happiness, employee well being, brand, reputation, influence, etc). These are not bad necessarily, it just makes them less efficient at particular tasks.
In particular, large government organizations are disproportionally disassociated from consequence. This is directly related to election cycles, politics (spinning the message to create success out of failures), and constantly getting inexperienced leadership. The first two are points are obvious, but let me explain the last. New politicians are elected and they appoint the people that they feel are best to run these large organizations. Typically, neither the politician or the person appointed has never spent a day working in that organization or executing the tasks for which they are responsible. With large organizations, it takes truly special managers to make an organization efficient at something you have never done yourself. This is even more difficult when you are not answering to your boss for your ability to execute, but instead are answering based on the ebb and flow of the political landscape.
In general, my point is, that you are right. Large private organizations do not have a significant advantage over large public organizations. That advantage can be widened based on political systems. However, the real advantage that people reference when making this argument is the enormous advantage small to medium private organizations "enjoy".
Understand that I am only talking about efficiency and execution and not societal value (differing goals). I also recognize that there are many tasks/goals that require larger organizations and more longevity.
Very informative post. Thank you!
> Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this.
This is hard to measure empirically because of selection bias - you can only study things which have already been privatized, and the first things to be privatized are the things you think will privatize successfully. For example, there are quite a few countries that have successfully privatized their telephone networks. But that doesn't mean you could privatize the police or military with the same success!
And because you asked for evidence-based citations, here you go: http://www.quora.com/Is-there-empirical-evidence-to-support-...
I think this is a case of the usual issues of in-sourcing vs outsourcing. I know a large insurance company that flip-flops on their decisions every 3-5 years. When in-sourced, staff are "lazy, cost too much, can't be fired due to HR rules, etc, etc" so they outsource, at which point the outsourcer becomes "they are not responsive, too hard to work with, every little items needs to be tallied against the contract, every little change requires too much time and requirements specification, etc, etc" so they turn around and in-source again, to begin the cycle over again.
In theory, in-sourced allows for better flexibility on what work is being performed (you are already paying the dollars for hours, so you can allocate as you desire) but you may lack expertise, less flexibility to expand then reduce/contract the workforce (which is especially true with governments and their unions), too hard to get rid of people if they do not grow skills and adapt as the organization changes, etc. Outsourcing works well when there are well defined boundaries, and requirements to expand/contract the workforce (within reasonable limits, the problem here is that the Olympics are too big for an outsourcer to deal with because there are not enough staff in reserve).
None-outsourced government IT projects aren't any better.
We decide to install system X so we need to hire some experts in system X. But system X experts are expensive and if we hire them as staff they are paid according to fixed internal pay scales - or there is a hiring freeze - or we have to hire 'internal candidates" from other government branches first.
So we hire consultants. These are a lot more expensive. Then part way through the project there is an investigation into how much "external consultants' are costing. So there is a ban on external consultants which the government shows has saved us $$$ million.
The result of course is that we have a half-implemented system X that costs us $$$$million and doesn't work.
> Conventional wisdom is that the private sector is just much more efficient than the state can ever be, but I have never been able to find evidence-based studies for or against this.
The train of logic is that if coprorations become efficient to maximize returns on investments and products sold, then since government programs have neither a need for a return on investment and exist outside of market economics because all the associated costs are a forced payment by the "customers" that they have no reason for themselves to be efficient. The government as a whole might want them to be efficient, but there is no intrinsic benefit for a government program to maximize efficiency because that has no influence on its short run continued existence.
Not saying it couldn't be as efficient, but large programs end up with too many people in positions of power which means there's probably at least one bad apple in there trying to milk the lack of demand economics or operating costs.
It's the same as buying any other sort of goods and services. Do you think the NHS makes its own paperclips and grows its own potatoes? No, the private sector really is massively more efficient, for two absolutely crucial reasons: 1) there are actually many sellers with interchangeable products and 2) paperwork and catering aren't actually core to the NHS mission.
If 1 and 2 aren't true, if there are a limited number of sellers of the service, and the service is core to the mission, then outsourcing will certainly fail.
The majority of government outsourcing deals are ideology or to hide the numbers.
Take PFI, if I offered to take over the rest of your mortgage on your house. But you had to pay me equivalent of the mortgage payment for 100years and you had to buy my services for any cleaning or maintenance of your house for the 100year term - then even though you were now mortgage free (ie. you had reduced your deficit) - you would be mad to take the deal.
If however your only interest was being able to claim that you had reduced your deficit going into the next election - and it wasn't your money anyway - you might look at it differently
During my time in the USA, I saw that most people think the private sector is inherently more efficient than the government. As you said it seems like an axiom.
However, I disagree it's people+procedures+technology that make a great organization and none of those concepts are exclusive of either private or public sector.
The situation is not as black-and-white. The private sector is more efficient in some cases, but that doesn't mean _everything_ should be privatized.
A trip to the DMV (Department of Motor Vehicles) will probably leave anyone with the impression that government offices are dens of inefficiency and bureaucracy. And that may as well be.
I think the sweet spot lies in the middle. There is no reason, for example, for the government to put a (say) landscaper on its payroll to maintain a public park. The overhead of a FT government employee is just too much; the job can be done well by a myriad of private companies. Same goes with bus drivers, janitors, painters, etc.
If the job requires domain knowledge and other such intangibles, then it usually can't be outsourced. This is why many IT projects designed to "upgrade" or replace existing IT systems at government agencies suffer: there are a lot of intangibles that just can't be all laid out on the contract form.