Do you believe that countries should borrow more now? Many
of you will say yes, and many will also say no, but few will say maybe, that is
until you read to the end of this post 😊. Your answer will partly reflect where you
live. In the U.S. your answer will likely reflect your political affiliation.
In previous communist states your answer might be a resounding no on the basis
of bad experiences. There seems to be no clear-cut answer on whether countries,
and in particular governments, should borrow more.
It sounds like we need a framework. Boring, yes! But
exciting in the sense that it will help us talk through the pros and cons of
our decisions.
Let us first try to understand why borrowing is on the minds
of everyone. Let us evaluate the intentions of our actors.
Spending can be justified
Sensible people agree that emergency spending is justified.
Think of wars, of famines, and pestilence. Societies do not like these. It wants to minimize
the impact of those things on our lives. Society reduces the lost utility from
minimizing the impact of wars, famine, and pestilence on everyone. It makes
sense to spend since the cost of not spending seems higher than the cost of
spending. Since our governments supposedly act out the wishes of the people,
spending ought to reflect those wishes. Those wishes are often tied to our
utility of things (e.g., here we want to avoid further damage).
As a general rule we want government spending to be well targeted,
timely and temporary (or the 3 T’s of fiscal policy). Targeted in
the sense that citizens know what is spent on (and that it reflects voter
wishes) and that the returns to that expenditure achieves its objectives (e.g.,
growth or income distribution) in the most efficient manner. It should be
timely – i.e., what is the purpose of infrastructure spending during a pandemic?
Infrastructure spending takes time to have an economic impact (it depends on
how long construction projects become usable). It should be temporary in the
sense fiscal policy ought to be also sustainable and not crowd out private consumption
and investment.
How well do you think your government is doing with respect
to the 3 T’s? Do fiscal operations reflect your views? Do they reflect society’s
views?
But make sure you understand the consequences of a larger
government
If governments do not reverse spending once emergencies
subside, then it implies a larger government. No one is talking about this,
which for me is baffling.
Put aside, just momentarily, your philosophical views about
whether you like big government or not. A bigger government will mean that society
will face a reduction in personal incomes via higher taxes. Taxes in this case
come in the form of inflation, if debts are monetized, or taxes in the ordinary
sense of the word – the decrease in take-home pay at the end of every month.
If governments reflect the views of its citizenry, then most
people will not mind that more of their incomes are given to governments.
To tax the economy fully,
partially or not at all, depends on what society values.
What we do know is that
expenditures, on average, should not exceed revenues to a point where debt
becomes explosive. Communism would want all assets to belong to the state, or
at least exercise significant control of what is purchased and sold, while competitive
markets would want as little state intervention as possible (not zero – since we
also don’t want markets to be sneaky). Both ideologies critically depend on
what society values. Most countries value some form of government and the tax
rates should reflect that.
So, it seems that there is some
minimum level of government that most people would be satisfied with.
The goods and services that the
government provides tend to be non-exclusive and non-rival goods. This simply
means that everyone has access to the public good and that consumption of the
public good does not imply less consumption of that good for another person.
Properly structured, the size of government would adhere to some principles of
equity. Examples may include a free and good education system, which provide
opportunities to narrow the gap between rich and poor, and the provision of
critical healthcare as basic human good.
However, the notion of a
“free” good is slightly misleading.
Someone has to pay for the school
or park or doctor. It is not really the government that provides these goods
and services. Rather, the government is the administrator of tax funds that are
used for paying the service or good. It still adheres to the principles of
non-exclusive and non-rival; however, it should not be taken for granted that -
it is still a paid service.
The cost of education and
healthcare would be priced according to market fundamentals if there was no
government. I.e., the price of education would reflect demand and supply. However,
if the market for education and healthcare was somehow controlled by a few
schools and healthcare providers (i.e. a lack of competition) then large
markups would be charged over what a typical person would want to pay.
Unfortunately, the demand for healthcare and education may be price inelastic –
meaning that schools and medical-care providers can charge above the market clearing
price making it perhaps too expensive to send children to school and providing
healthcare for those who need it. If society values healthcare and education,
and society is willing to pay for those that cannot afford it, then the
government can serve as the middleperson to administer the process. Unfortunately,
governments are tied to a political party, which may lead to a wedge between
what society values and what the politician values.
If society agrees that
taxpayers should pay for certain non-rival and non-exclusive goods and services,
we need to ensure that taxpayers are not cheated.
This entails that every $1 tax
generates equivalent spending of $1 that goes the intended targets. This implies
zero leakages (such as corruption) and a fast rollout of activities (i.e. no bureaucratic
delays). However, governments often cannot meet this requirement. Taxes are
often used to pay for other things, which includes government salaries and
wages, transfers and public infrastructure spending. In cases where
governments are inefficient or corrupt, the $1 spent on education and
healthcare is much less.
In many countries, governments
are elected representatives. If they are voted into power, and the people like
what the elected government is doing, then they will continue to receive
support from the people. If they do not like what the government is doing, then
politicians run the risk of being replaced. The incentives of corruption and
inefficiency rise when the penalty of shirking is not strong enough, or the
rule of law not respected.
(Grossman 1987)
suggests a way to measure the optimal size of the government (measured either
in terms of GDP or employment). Government is either a benevolent force seeking
to correct for excesses of an unrestrained marketplace (the Pigovian view)
or as a source of distortions from being used by special interest groups (the
Public choice view).
Grossman’s interesting paper
attempts to reconcile the two views by making government an input into the
private sector’s production function. The theoretical model (which is backed up
by an empirical analysis for the U.S.) shows that early stages of government
development benefits the public and can be productive by limiting externalities
or failures in the market. However, as the government grows it becomes beholden
to special interest groups, which exerts a negative influence on welfare and
productivity. It is thus optimal to reduce the size of the government in order
in increase output. If government enters the utility function of households,
then an increase in government spending (financed by taxes) reduces growth and
savings in the long run. However, under the assumption of a Cobb-Douglas
production technology[1]
the optimizing government can still satisfy the conditions for productivity
efficiency (Barro 1990).
Ultimately, the size of the
government should be determined by the utility that society derives from a
government. There is also a cost.
We have a mental framework for
this value function. Welfare analysis builds on an egalitarian framework (where
welfare is equal to the utility of the worst-off member [Rawlsian] or
where each individual is endowed with the same amounts of resources) or a
utilitarian [Benthamite] framework (where welfare is equal to the aggregate
utility of society). The two welfare functions are often at odds with each
other since the latter cares not necessarily about equality, while the former
does. Can larger governments achieve greater welfare then – in any of the
definitions? And if it can, can it do so better than the private sector?
Part of the answer has to do
with an individual’s willingness to lay down certain freedoms at the altar of
the government in exchange for free or cheaper education, government as an employer
and provider of health care. These sacrifices include transferring earnings and
wealth to the government, relinquishing control over privacy – location, tax
numbers, financial details, health indicators and what gets taught at schools
to be managed by government. Apart from individual sacrifices, one would have
to manage and reward hard work and innovation differently, since monetary
compensation used to be the encouraging medium.
The flip side of how much value
society should attach to the government needs to be weighed against the
efficiency and competency of the private sector. Can the private sector, however,
not also deliver education, healthcare and provide employment and construct
roads, building and ports? It can and does, but it needs funding. Taxpayers in
communities, or representatives in communities can hire the services directly from
the private sector if the law allows for it. Instead of just allocating funds
to the government, societies should be free to determine their needs and how
to acquire it. If the right regulations are set up (i.e. ones that enforce
fair play and allows for competition), then prices of goods and services should
reflect market fundamentals. This is the free market. Unfortunately, not all
players in the market play by the rules. Community representatives can be
bribed, big companies can completely kill competition by undercharging and
then later hiking prices, or first mover firms can limit competition early in
the development process. The literature identifies that curbing excessive
mergers and acquisitions and promoting firm entry is important in ensuring the
correct incentives for private sector innovation (Aghion et al. 2009).
Controversial topic: Taxpayers
should have a say on government determination
A good question that should be
asked more regularly in the public space is whether taxpayers should have
anything to say about the expenditure envelope of the government. Some might
rightly argue that they do have a say because most governments are elected in a
democratic fashion. Unfortunately, not all governments necessarily represent
the views of the taxpayer – especially if the tax incidence falls on a much
smaller group of individuals and firms relative to the larger voting
population.
Fiscal lessons from the past
should be internalized much better than what we currently observe. Mistakes are
repeated and this will come at a large cost (i.e., you don’t have an infinite
government budget – governments can default, and these defaults can be costly!).
To avoid persistent fiscal policy errors and to close the gap between voter
wishes and fiscal outcomes a better system of monitoring and voting could be
set up.
In this section I argue that two
parallel voting systems should be set up. One where the general population votes
an individual or party to represent their needs (which is currently the case
for many countries) and a second where taxpayers decide on the execution and
possibly the allocation of expenditures. Furthermore, this system should
utilize innovations in electronic voting platforms to increase voting frequency
and the range of topics.
Why do we need an outdated
mode of voting when electronic advancements provide the platform for real-time
votes on all matters pertaining to society?
Can society not change its
utility function when it becomes more informed or when it learns? If society
pays taxes, then surely it is society, not the elected members from a four to
five-year voting cycle, that should determine where and how funds are allocated.
Society should also determine the size of the government. The government should
not be arrogant to think that it knows better than society. And even if society
makes poor choices, then at least the responsibility of fixing the failures should
be on all members of society. Elected officials who do not do their work
adequately should be replaced at a click of a button. This also eliminates the traditional
view of parties. A two-party system, as an example, can never meet the needs of
the general public fully. Having representative candidates, where
representation is on a policy or project, not party, might get closer to
servicing the needs of communities.
A couple of serious questions
come to mind. (1) Will an increased voting frequency not be destabilizing? (2)
Do we not need time to see through reforms rather than just replacing officials
for under-delivery? (3) What stops people from “gaming” the system – i.e. can
any person come up with any policy and then how does the pass or fail of the
policy proceed?
In case (1), I argue that an increase
in the frequency of voting will lead to a stationary distribution very quickly.
I.e., learning to vote occurs at a fast pace. The ability to vote in real time
will allow for error correction and a reversal of bad policies. Bad policies tend
to punish all of us. What hinders us is to correct the mistake quickly. The
problem with four, or five-year voting cycles is that policy errors can be
locked in and that policies do not necessarily reflect the views of the voting
public at all points in time. There is
also the question of path determinacy. If a country should reach an inevitable
point in the future, then four-year voting cycles might delay that outcome. An
increase in voting frequency could speed up the pace of reaching that goal.
We know that we can hardly
express society’s welfare function (due to Arrow’s impossibility theorem – see Arrow
(1950)).[2]
We thus retain the idea of a majority voting scheme as a second-best pareto
optimal allocation and that through trial and error (tâtonnement) society will
move towards maximum social welfare.[3]
The majority voting scheme adheres to the pareto and nondictatorial axioms (a
single agent does not determine society’s preferences), but fails in terms of
the transitivity axiom (rankings are not transitive in the sense that if A > B and B > C then A > C does not necessarily hold – called the Condorcet
Paradox) and the axiom of pairwise independence (i.e. that A and B are independent of C).
However, with real-time frequent
voting platforms, agent preferences and errors can be corrected. As an example,
preferences may change and voting schemes may entail stages where the optimal
choice is elected through means of elimination.
Question (2) is addressed by allocating
some public funds to research and setting up the contractual agreements that
determine the duration of policies. A state auditor as an example will verify
if the public officials comply with administering the policy. The official agrees
to terms with the public that voted for her and the terms of reference include
certain quality markers on projects and policies. Failure to deliver results in
a change in representative.
A four-step process could be
followed to overcome gaming. I illustrate this with an example. Assume that a
person is interested in removing labor brokers. In step 1 that person should
put the policy up for discussion. This is done on the voting platform. This
requires as many upvotes to be deigned of public interest. If the interest breaches
a threshold number of the population (e.g., 50%), the policy is escalated to a
team of diversified researchers who study the costs and benefits of the policy.
The research group needs to include skilled, educated, and nonpartisan individuals
on a topic. Once the research is done it is released for public comment – that
is step 2. Once enough information about the policy is obtained then it goes to
the electronic voting platform where it is put to a vote – this is step 3. In
the final step the policy, if enacted, should be rolled out and administered by
the community or country representative that was voted into power. These
representatives earn their salaries from the taxpayer.
However, the application of
expenditures, once society has voted on the policies, lies with those who pay
taxes. Taxpayers are shareholders who can appoint a government
representative. If the government representative fails in her duties, then the
taxpayers may quickly judge the efficiency of the representative. This avoids party
systems that protect corrupt officials. It limits state capture since the power
is in the hands of a voting public who can make real-time decisions, instead of
behemoth governments who slow down processes. This should ensure that expenditures are much
more efficient and entails fewer leakages. It improves the mapping of
expenditures to the intended targets.
Error-correction is contingent on
the terms of references. If the policy was unsuccessful after a specified
period, then the voting system should allow for a reversal of the policy. There
should be no corner solutions or “sticky” outcomes within this framework.
Figure 5:
Voting framework
Since voting is now done on a wide scale of subjects one
might be inclined to see this system as a burden on citizens’ time. Many of the
proposal will take energy and time to set up in the beginning. But once
policies work (i.e. no more need for error-correction), then the process could
be monitored and only reviewed when needed. This implies significant noise in
the beginning of this voting methodology. However, due to urgent reforms and
general public interest, the outcomes should reflect voter needs much faster
than in current systems.
In many ways this system represents a modernized version of
referendums. Referendums are traditionally only held infrequently and on
important subjects. However, who decides what is important? Politicians may
mask agendas by holding referendums or society might make a mistake by voting
in policies during a referendum with unfortunate and unforeseen consequences
(regret). Utilizing frequent votes will represent the views of the society and
not stakeholders and will allow for corrections. The cost of voting is also
significantly reduced when using electronic systems. These systems need not be
insecure. We use electronic banking regularly as an example. Perhaps the
biggest obstacle is the roll-out of internet and access to electronic devises
in developing countries. I venture a guess that if we were able to reduce
inefficient expenditures, we might be able to provide internet access to
everyone…
[1] A mathematical formulation that describes the bundling
of factor inputs to produce outputs.
[2] Arrow was against plurality voting systems that are
done on the basis of ranking. An alternative voting system is based on approvals.
For more details see https://electionscience.org/commentary-analysis/voting-theory-remembering-kenneth-arrow-and-his-impossibility-theorem/.
[3] The Pareto principle states that one cannot improve
the utility of some members of society while reducing the utility of another
member. Social welfare can thus be improved when policies improve the welfare
of some individuals while not decreasing the welfare of others. Sometimes
equity is of greater societal concern than efficiency (who determines this?),
which suggests an alternative welfare function specification.
