Why Doesn't The Government Impose Taxes In Chickens?

One of the arguments of Modern Monetary Theory (MMT) is that governments develop a monetary system in order to make provisioning itself easier. If we put aside pointless debates about the ultimate origins of money, this view creates some subtle differences with neoclassical modelling assumptions, and this shows up -- sometimes -- in how we analyse the economy. I would not characterise the neoclassical assumptions as "wrong," rather that they create political blinkers on the role of government in the economy.

Why Chickens?

The origin of this article was the following tweet by David Andolfatto.*

I'm sorry, but this is just wrong. Govt wants your chicken and takes your chicken--you're down a chicken. Govt says it wants your $, makes you sell your chicken to govt for $. You're still down a chicken. What's the difference? https://t.co/Ndt2ubcRI7

— David Andolfatto (@dandolfa) April 24, 2019

Since he is working within a 280 character limit, the risk is that we read too much into this. Although I disagree with the sentiments (for reasons to be discussed), I think the point can be understood using mainstream economic assumptions.

If we go beyond what is embedded in the 280 characters, we can see how the mainstream assumptions differ from MMT.

Basic Analysis

Andolfatto is correct when he points out that there is no difference from the perspective of the household. But even then, he has embedded a new version of a barter-based double coincidence: the government wants your chicken, and you coincidentally have one.

Since even American citizens are unlikely to possess aircraft carriers (yet), the government cannot rely on this double coincidence to occur. Furthermore, how can it get people to enforce laws and so forth? Force community service on lawyers so that they take turns to enforce the law? (There is one case where governments have successfully provisioned themselves in kind -- drafting young men into the military. This works, since they just need bodies to act as cannon fodder.)

How many chickens is this guy worth?

Furthermore, there is an institutional issue. We live in an industrial capitalist society that relies on the rule of law. Tax collectors are not allowed to just seize real goods that they feel are convenient for the government. (In agrarian societies, this sort-of worked, since most peasants had agricultural commodities on hand.) The government needs to impose a tax law, specifying the obligations. This allows economic planning to occur.

Imagine a Canadian tax collector showed up at my door and demanded a live chicken**. The only live animal I could offer is one of my cats (pictured on the left). The question is: what is the exchange rate between cute fuzzy cats and chickens? Since we do not live in a world where there is a magical general equilibrium exchange ratio between all goods, there is no obvious way to determine whether the tax obligation has been discharged. The use of a monetary unit of account ends such uncertainty.

(This demonstrates a key difference between voluntary transactions, and legal obligations -- taxes, debts. For a voluntary transaction, if your counterparty does not use a reasonable valuation -- "off market" -- you just find a different counterparty. The need for market makers to be "on market" means that we can typically infer an arbitrage-free pricing "curves" in fixed income and derivatives (except when markets are totally disjointed). However, the need to discharge an obligation forces both parties to agree on the valuation (or acceptability of deliverability of goods into a contract). This explains why financial analysis can get away with equilibrium concepts, whereas equilibrium is dangerous in economic analysis.)

Thus, the superficial analysis shows that there are massive practical differences between taxes imposed in the monetary unit of account versus payment-in-kind. However, there are deeper theoretical issues, which are discussed next.

Theoretical Points

The real question of interest: is it safe to argue that neoclassical models -- particularly those that are given in terms of real goods -- already cover the MMT argument that real resources matter for the analysis of fiscal policy?

The Andolfatto quote matches some fairly standard assumptions of neoclassical analysis.

• Household-centric. The discussion of taxes is in terms of how it effects a household, in particular, it's utility. Is this really the correct frame when discussing government policy? If the objective is to defeat some menace, stomping on some households' utility is the least of the government's concerns.
• Composite good.  The assumption that a household would have a live chicken on hand is what you would expect if you always work with composite goods. That is, a household has a standard consumption basket, and so taxes are in essence reducing the size of the basket that can be bought. However, this is not helpful if the government wants to achieve particular goals -- e.g., replace carbon-based energy sources in the economy (a.k.a., the Green New Deal). The availability of particular inputs cannot be taken for granted.
• Taxes almost never are matched to the producers of desired goods. Many specialised goods and services the government wants to purchase cannot be bought on existing markets: artillery, bridges, road repair, etc. If the government does not produce the desired output itself, it needs to induce the private sector to develop the specialised skills to produce the output. That is, if the government just seized tanks, nobody would go into the business of producing tanks -- since they would be just seized, and cannot legally be sold into the market.

This means that standard neoclassical theory is not particularly helpful. If the economy just had a single good (composite or not), the economic calculation problem would be rather trivial. That is, the economy could easily be centrally planned, and optimal outcomes obtained. The reality is the multiplicity of goods makes central planning difficult (although the abandonment of anti-trust laws means that some conglomerates are entering the central planning business).

However, there are circumstances where these assumptions are not too far off. This is the case when the government attempts to have no distinctive economic footprint in the economy. Essentially, this is what neoliberalism is all about.

If the government expenditures are just used to buy standard goods in the markets, hire staff, and make transfers, there are no specific goods the government worries about. They may purchase office equipment, but that would just be a standard consumption basket. And government wages and transfers are just going back to the household sector, which is then presumably used to purchase the household consumption basket.

In which case, the effect of fiscal policy is just to take some workers out of the labour market, and provide demand for goods in standard consumption baskets. In which case, an aggregated approach should provide a decent approximation of reality.

Concluding Remarks

We can probably get away with ignoring the real effects of fiscal policy if we constrain governments from doing anything other than creating demand for standard consumption baskets. Moving in that direction has certainly been the thrust of post-1980 policy. However, that is a political choice, and there is no reason to believe that governments will always have that luxury. 

Footnotes:

* In case the tweet is not available, the text is: "I'm sorry, but this is just wrong. Govt wants your chicken and takes your chicken--you're down a chicken. Govt says it wants your $, makes you sell your chicken to govt for $. You're still down a chicken. What's the difference?"

** David Andolfatto verified that the "chicken" in the tweet was in fact a live chicken.

(c) Brian Romanchuk 2019