Can a reduction of employer taxation lead to more equal income?

Question

In Argentina, it's usually said that the economic burden on employers is very heavy (just an example). So much so that, if it were lifted, more people would become employed, but I don't quite understand how this would work.

Employer taxation is one part of the problem, the other part is the "deductibles" from the employee's salary. As both are being paid by the employer, I would like to ask: Would a reduction in any of these allow for more equal distribution of income?

Employers pay, per employee: 17% as social security (retirement fund, union and insurance), 19.5% as contributions to different government entities.

There is also a 13th payment (the year has 12 months, but the employer pays 13, the extra is paid half in June and half in December). And the fact that firing employees has an added cost of more than a month's salary.

Answer 1

It is a complicated question. The answer to your title question is: no, it usually wouldn't lead to more equal income.

It was tried multiple times, the result is always the same: employers keep the same number of employees working for the same salary and increase their own profits instead. The previously taxed part which was going to the government previously doesn't "trickle down" to the employees and is kept by business owners/employers instead.

So much so that, if it were lifted, more people would become employed, but I don't quite understand how this would work.

They claim that if the employers had more money to throw around, then they would employ more people. That claim is (mostly) false, because businesses do not employ people just because they have extra money, they employ them because they have a demand for labor.

It MIGHT lead to higher employment or it might not. What it will lead to is wage stagnation, eg wages wouldn't change or will go lower in some cases. And the increased employment for lower salaries is generally not good for the economy.

"We have 100% employment." says government economist.

"Yes, but almost everyone is working for 1 dollar (or however little) a day and is on the edge of starvation or starving!"

Look at this from business perspective. Let's say you are a hardworking business owner. Your business has grown so much that you can't accomplish everything yourself. You decide that you need another worker to shoulder the burden. What you have here is demand for labor. So you look into the labor laws to see the minimum things you need to have to get another worker. As an example, if you are in construction business, that might mean that every worker must wear a hard hat. You don't like those conditions (or cant afford the hard hat for every worker, maybe), so you choose to go without and work more yourself.

More time passes, your business grows even more, there are more and more things that require your time and attention. Now you REALLY can't do everything alone, even if you worked 24 hours a day. You look into the labor laws and decide that your time and mental health is more worth than money, so you hire somebody to help you.

Labor laws are business laws like any other law. If you decided to open a restaurant, for example, you would need to keep it clean, not have cockroaches in it and not serve poison in your soup. You want to open a restaurant, there is a huge list of things you need to have and rules you have to follow. There is a minimum barrier of entry, of standards you have to follow if you want to do that job and be part of that market. The same goes with hiring an employee. Government decides what those minimum standards are going to be.

Now, say that you already have 5 employees. You are on the edge if you need 6th employee or not. If only you could pay all of them 100 dollars a month less, then you could afford 6th employee. But would they want to work for 100 dollars less or would they walk to a better place?

Supply and demand. Employer has demand for labor, workers supply the labor. It is a transaction that has its costs (taxes and "deductibles"). If you can't pay those costs, then you can't afford hiring extra labor.

You either have money for 6 cakes or you don't, as it were. If only there was smaller VAT, you could buy 6 cakes. As it is, you can afford 5. The problem with this is the underlying assumption that smaller taxes mean lower prices. And that is not true. The cake seller had already formed the price of cake. It has shown that there are buyers for that cake at that price and he is doing quite well with it. Lowering the VAT afterwards means the cake price for the customer remains the same, only more of it remains in seller's hands as profit.

And if we speak worldwide, Argentina has quite low labor taxes. There are countries which have 65% or more taxes and "deductibles".

In short, it is a good story to sell the population which doesn't have much basis in reality.

"If only the government wasn't taxing me as much, I would give more money to charity. I would give more money to you, my employee! I promise. I swear it on everything holy!" is the story told.

In reality, if the government lowers taxes, the guy who told the above tale most often doesn't give you/his workers more money but buys a new car instead. Or spends it on something else, like an apartment or summer holidays in Bahamas. Or spends the money on management bonuses. Or on hiring external marketing company. Or (rarely) he does do exactly what he promised.

It does lead to more economic activity but usually doesn't lead to more income equality.

Answer 2

The basic economic theory is that any form of taxation reduces the overall amount of economic activity because it introduces inefficencies. Simplest example: If I have a business idea where I can pay someone 10$ to make something that I can sell for 11$, I will employ someone and do it. If I have to pay another 2$ of taxes on top of the 10$ salary, I won't do it.

There is two problems with that. First, this assumes that the 2$ of taxes just disappear. If for example the taxes go into a health insurance for the worker, the worker might perceive them as salary equivalent. So he would get 10$ cash plus 2$ worth of health insurace, so 12$ salary total. Reducing the tax would not make the worker better off.

Second this implicitly assumes there are workers available that would want to work for the 10$. This may or may not be the case in real life situations.

Now even if you believe the initial argument, all it says is that reduced taxes lead to more overall ecomonic activity which usually means more overall wealth. This kind of argument says nothing about more equal income. At best you could try to argue that more overall wealth means more income for everyone but this may or may not be true in real life situation.

Answer 3

tl;dr:

Since the two other answers on this question do not provide science or evidence based answers from an economic perspective nor do they provide any citations to credible sources within economic literature let me provide a science based answer from an economist.

Actual research in economics would say that the answer depends on how exactly is the payroll tax organized. Is it set up in a way that it optimally ensures progressive redistribution? Or does it have some regressive components? Often even sub-optimal taxes will be generally progressive but one can't just assume that outright.

The short summary of the correct explanation is as follow:

1. First, social contributions and other taxes that depend on supply of labor are taxes on labor not on business. Government can only chose de jure incidence of the tax not de facto tax incidence - this is economics 101.
2. Income inequality exists because people's marginal product differs and because people provide varying supply of labor to the market. Hence paradoxically it is completely possible that the more we distort peoples incentive to supply labor the less income inequality we can have. In extreme nobody works because we confiscate immediately all income with 100% taxes everyone has equal income and income inequality as measured by GINI would be 0. Inequality is not about people having jobs in fact an optimal inequality smoothing tax rates heavily dis-incentivize people form working.
3. In economics taxes cannot be discussed without transfers. A transfer is just a negative tax.
4. Generally there is always some (marginal) tax rate function that can smooth out income inequality. This generally cannot be done without welfare cost as there is generally always tax-efficiency trade-off. However, it is a moral and ethical question what to choose - an existence of trade-off does not imply choice on either side is invalid.
5. In real life government does not necessarily sets taxes optimally with intent to reduce inequality. A sub-optimal tax schedule can very well increase income inequality. This has to be considered on case by case basis. In the case of Argentina it is hard to say (or at least I was unable to find research which would estimate all relevant parameters).

Full answer:

Part I: Social Contributions are Taxes on Labor not Business

Many non-economists hold common misconception that government can assign the incidence of tax burden or that government by declaration can decide what is taxed by a tax it declares. This is in economics known as flypaper theory of tax incidence (see Mankiw. Principles of Economics 8 ed. pp 239), and this theory was never actually even hold by any economist, it's a sort of inside joke to even call it a 'theory'.

Next, actually in economics taxes are considered to be levied on activity they depend on (Stiglitz, Economics of the Public Sector. 3ed). Social contributions do not depend on business activity per se but they depend on labor inputs of the company. That is a company that uses no labor would not pay them. Hence they are taxes on labor.

Furthermore, the tax incidence of this tax on labor will fall both on employee and the employer. Government can't choose how it distributes tax burden as this is determined by supply-demand interactions in the market (see examples in Stiglitz Economics of the Public Sector. 3ed pp 482-517.). In practice this means that just because government says social contributions are lets say 10% on employer and 10% on employee that does not mean that employer and employee actually share tax burden 50/50 (sure in rare cases government might choose ratio that accidentally corresponds to actual incidence - although odds of that happening are astronomically low given that de jure tax split can be any real number on interval [0-100] (where 0 would be all tax on employee for example) and anyone with high school mathematics should know there are infinite amount of numbers on that interval.

Unfortunately both theoretically (ibid. Principles of Economics, ibid Economics of Public Sector, Bradford (1978), Mirrlees & Adam (2010), Chamley (1986) and Judd (1985) - last two papers are technically about capital taxation but they show that any taxes levied on owners are borne by labor in long-run), and empirically (e.g. Roy-Cesar & Vaillancourt (2010), Gruber, 1997 etc.) most of the tax incidence of labor taxes happens to fall on labor supply (i.e. employment) or returns to labor (i.e. wages) and this is to the point that we can completely ignore for the rest of the discussion that some of it also falls on business owners (in most cases is simply too small so it's effect second order at best). The reason for this is that generally speaking demand for labor is elastic while supply of labor (especially at lower income levels) is relatively inelastic.

Part II: Income Inequality

The result from previous section however does not mean these taxes do not affect inequality. Quite the opposite actually! Why? Many people do not realize that large portion of income inequality is generated by inequality in labor incomes (capital incomes also contribute to income inequality but the contribution of labor incomes is enormous and not understood by non-economist - see Atkinson. Inequality: What Can Be Done?).

In fact the portion of income inequality generated by variation in labor income is so significant that basic optimal taxation models do not even include capital incomes in and just focus on labor incomes (see Mirrlees (1971), Diamond (1998) and Saez (2001)). To be clear the above does not mean that taxing other factors or sources of economic activity does not matter for income inequality (Mirrlees & Adam (2010), ibid Inequality: What can be done?) but income taxes are generally considered to be one of the most direct ways of addressing it.

How come? Well this is because many people do not realize that even people like Jeff Bezos, Elon Musk etc are not just capitalists (i.e. owners of capital through stocks etc) they are also workers. Most of CEO's are employed by the company (even in cases when they actually own the company). Furthermore, non-trivial portion of inequality is generated by superstars. When economist talk about superstars we do not mean just Hollywood but any individual who is able to generate substantial income from fame (famous scientist, politicians who get book deals etc. see Scheue & Werning, 2017; Rosen, 1981).

Part III: Optimal (Labor) Income Taxation

As our word becomes richer we get more sensitive to inequality regardless of the fact that even poor people nowadays live lives that would be envied by kings of the past. As a consequence literature on optimal income taxation and also redistribution (welfare transfers are often considered and modeled as 'negative taxes') is becoming increasingly wide and increasing amount of attention is paid to this issue.

Even though the question was about payroll tax from economic perspective there is really no meaningful difference between straight (labor) income tax, payroll tax/contributions, social contributions and other labor taxes. They are all labor taxes regardless of what ugly/nice name politicians give them.

These taxes can always affect income inequality. Let me give you one provocative example. A 100% confiscatory tax would remove any incentive for anyone to work - what would an effect on inequality be? It would literary solve most of the inequality. Many non-economist do not realize this but inequality has nothing to do with absolute poverty. If everyone lives at subsistence level there is no inequality. If we have economy of only two individuals Bezos (est. net worth 113 billion USD) and Pera (est. net wort 7 billion USD) inequality would be much much larger than in the example where everyone is starving.

However, in the above example I was purposefully being provocative given that many people don't realize what inequality actually is. It is certainly doable to also reduce inequality without reducing everyone to subsistent level. Optimal models of income taxation show that depending on what social welfare function we adopt (i.e. is our society Rawlsian, libertarian, utilitarian etc?) as redistribution cannot be divorced from some normative notions of morality, optimal marginal top income taxes can be as high as 50−75% in the US (Saez, 2011) and results from other places are generally in line with these.

The optimal taxation of labor incomes however has several distinct effects:

• By driving a wedge between social and private returns to labor taxes discourage employment and labor supply and encourage people to substitute leisure for labor (so-called substitution effect).
• By reducing real incomes it might create an income effect where because everyone is poorer (and assuming richer people like to consume more leisure) people will try to work more.

However, note save for special cases (not considered and even relevant here) taxes always lower aggregate welfare even with transfers - welfare in economics is the sum of individual utility not output/GDP etc. Labor income taxation generally implies efficiency-equity tradeoff where efficiency does not refer to maximization of GDP/output but welfare as measured by utility.

• On the receiving end if the taxes are used for welfare transfers they further discourage employment by providing people with other means of living.

An optimal income tax will try to balance all of the above (e.g. Saez, 2001). However, it turns out that as long as income distribution can be characterized as a log-normal distribution with pareto tail optimal marginal income tax schedule will always be such that it will be creating more unemployment (Atkinson, Piketty, and Saez, 2011) - poverty trap is optimal feature of taxation under either Rawlsian or Utilitarian social preferences (i.e. aggregated preferences of society as opposed to individuals). However, 'poverty trap' here does not mean people are actually poor - it refers to the fact that when we optimally smooth income distribution actually poor people will be relatively well of so that they will have little incentive to try climb the income ladder. This can be to some extent, but not fully, helped by some government policies but that discussion is beyond the scope of this answer (ibid. Economics of the Public Sector).

Consequently, regardless of the fact that payroll taxes fall mostly on workers and labor they can be used to reduce inequality. In fact this can be done regardless of the fact that they lead to higher unemployment. A world where there are two employed people with salaries $100 and $1000 is more unequal then a world with one unemployed person with welfare check (a post tax & transfer income) of $300 and one employed person with (post tax) income of $600.

Part IV: Optimal Taxation Meets Real World

Alas, to the sorrow of all public economists around the world governments often do not follow our prescriptions. Countries routinely engage in perverse sub-optimal tax schedules that redistribute money from lower income people to high income people. This unfortunately happens even in such modern and advanced countries like the Netherlands (where I live), or Denmark, Sweden etc. To be clear in the countries I mentioned on the whole tax system will be progressive but certain taxes might have regressive components.

For example, Jacob, Jongen, and Zoutman (2017) and (2016) show that in many advanced nations in general and in the Netherlands in particular redistribution is often set up in a way that while overall still progressive - it has redistributive components that actually create redistribution from poor to upper middle classes.

Furthermore, deductions set up for retirement funds are actually often (but not always) empirically regressive (when we talk about inequality in dynamic life-cycle setting) as richer people live longer but even there it is always important to actually empirically double check this and systems could in principle be 'tweaked' in a way to make them progressive (e.g. Brown, 1998).

Unfortunately, I was not able to find research that would offer enough relevant information for me to comment on the desirability of those high Argentinian payroll taxes in reference to reduction in income inequality. As a result the correct answer here would be that there is no correct answer - more research is needed (I hope this was not an disappointment after such long post and you at least learned something). Of course, I did not look at every single work there is on Argentina so I might have miss something, I hope this answer will at least give you tools and key words to do your own search.

Part V: Conclusion

So to sum up, one can always achieve lower income inequality through taxation of labor income and payroll taxes are just taxes on labor from economic perspective no matter what nice name we give them (you can call a cat horse - it wont turn it into one).

As a result unless some taxes are set up in regressive way one cannot reduce inequality by abolishing or lowering taxes and optimal income taxes will not be regressive even under charitable libertarian/conservative scenario (i.e. scenario where society is libertarian but puts at least some minimum value on redistribution).

However, reducing inequality is not necessarily the same as creating more employment for poor (again poverty traps are sadly optimal) nor should inequality be conflated with absolute poverty. Often (but not always) there is a trade-off between reducing absolute and relative (i.e. inequality) poverty.

Lastly, real world governments do not always set up their tax schedules in optimum way. It is possible that abolishing (or replacing those taxes with better ones) would reduce inequality.

Recommended further readings:

A general and well-rounded source on anything tax related is the so called Mirrlees Review (Mirrlees was a Nobel Prize winner in area of optimal taxation). Mirrlees Review is probably the best source for any non-economist to get (relatively) non-technical overview of most recent literature on optimal taxation of any sort.

---

PS: If you are interested in science based answers to economics questions consider directing economics question towards dedicated stack (see Economics.SE) where science/evidence based answers to economics questions can be provided by experts.