Inflation

What is inflation, where does it come from, and where does it go?

Nov 01, 2021

Children play around with banknotes in Weimar Germany. From Imgur.

Hyperinflation is going to change everything. It’s happening.

Across the world, inflation is higher than it has been for decades, and discussing it is hotter than at any point since the 70s. But the discussion isn’t particularly good, about what inflation is (besides “prices increasing”) and about where it comes from, or even about how to end it. Plus extremely muddled conversations about items like “hyperinflation” or “high inflation”.

What is inflation?

Inflation is always and everywhere a monetary phenomenon in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output.
Milton Friedman

Inflation is “a generalized and sustained increase in the price level”. Most countries measure inflation by tabulating the prices of products that the average person consumes, and averaging them out given how much the usual person spends on each of them.

Exactly how much weight each product has is a matter of debate, and how often to shift them can result in various different kinds of price indeces. This last item is particularly, because for example different groups of people consume more or less of certain things, which means that inflation has different effects on them. For instance, if poorer people spend more on food and less on, say, clothing, then inflation that has stronger increases in food than clothing will be worse for poor people than the converse. The Wall Street Journal actually has a fun calculator where you can play around with this, because Covid and changes spending became a big deal last year.

There’s, generally speaking, two lines of thought regarding inflation. The first is the traditional, more mainstream story: there is too much money chasing too few goods. Originally, this referred to the amount of money (physical money) which was assumed to be completely arbitrary, but it turns out most money isn’t actually created by the Central Banks - it’s created by private (or government) banks instead. Thus, the role of the central bank isn’t just to create money directly, but to boss around the banks into making more or less money on their own. Regardless, inflation is what happens when there’s too much spending in the economy, and the amount of things that spending goes towards is fixed, at least temporarily.

On the topic of banks and money, the issue here is that there’s two big misconceptions on how banks create money. The first one is that banks loan out their deposits - this is wrong because banks loan out money, and also give people the chance to deposit money, and back those deposits with Central Bank reserves. You could easily imagine a bank that borrowed money and loaned it out at higher rates, no deposits involved. The second misconception is that the Central Bank tells the banks how much money to create via reserve requirements. This view, common in macroeconomics textbook, misses that reserve requirements are usually stable across time, and also gets the causality backwards - the Central Bank doesn’t set rates and then banks figure out how many loans make sense at a given rate, the banks loan out however much money they think is profitable and then the Central Bank wonders whether the banks loaned too much or too little, and adjusts rates accordingly.

The other view of inflation, which is far less common in academia, is that inflation is mostly about the prices of individual products, which are set by supply and demand in their respective markets. Basically, inflation happens when there’s too little supply of things that people spend a lot of money on. The analysis is the same (too much money, too little stuff) but the policy implications are different - mainly, a sort of “wait out inflation and/or spend money in certain ways”, which I’m not going to get into but it’s not been a successful strategy in the not particularly good Latin American 70s.

Where does inflation come from?

So far, everyone has that inflation is caused by, broadly speaking, too much spending on too few goods (and services, but normally goods). The problem here is which comes first: the excessive spending, or the shortage of goods. A similar question has been posed about the Great Inflation of the 1970s, which had a little bit of both but was generally mostly too much spending (plus some shortages, of things like oil, wheat and minerals), caused by policymakers with wrong beliefs about the economy and afraid that cutting back on stimulus would repeat the Great Depression.

So where does the extra spending come from? Usually, because of the magnitudes involved, from the government. When the economy is running below capacity, stimulating it through monetary and/or fiscal policy is necessary to get it to full employment. But at that level, at least in the short term, stimulus can’t increase output: because investments take time to come to fruition, increasing the level of spending in the economy through expansionary policies results in, quite simply, people bidding up the same number of goods. Consequently, at least in the short term, stimulus policies can’t actually grow the economy when it’s in equilibrium.

You could also have problems involving other types of things. Firstly, spending could be normal, but problems in production might make aggregate supply dip down. At that level, it’s likely that contractionary policy won’t help with inflation, and might actually cause future deflation by locking the economy below potential. The policy solution to this is frequently to just do nothing and wait it out, although in some cases regulatory barriers (such as trade protections or things like zoning laws) could be a problem. The “heterodox” policy recommendation is something called incomes policy, and normally it’s a complicated mess that comes dangerously close to central planning (and to making all the errors central planners make).

Exactly how the economy being at or below potential influences inflation is a big can of worms among economists. The main channel is the labor market. When the economy is below potential, there’s a lot of unemployed people going around, so employers can get away with paying their workers less (because of supply and demand). Conversely, a stronger economy means that most people who can get jobs have them, so workers can increase their earnings by asking for raises or, for example, by switching to better paying jobs. Since wages are part of companies’ costs, then they will be reflected on inflation, which will be below or above target depending on how strong the economy is. It’s possible that stimulating the economy in equilibrium will cause wages to go up more than is sustainable, resulting in higher prices, resulting in more wage increases, leading to even more price hikes, etc. This is called a wage-price spiral and it doesn’t normally happen outside of highly inflationary contexts.

There’s, roughly, two final sources of inflation that are intertwined in ways that are fairly straightforward: inertia and expectations. Inertia is the role of past inflation in future inflation; for example, if inflation goes from 5% a year to 2%, then a lot of contracts will still assume 5%, so some parts of the economy will be “out of sync” with the rest and push inflation upwards. Additionally, expectations have a role to play too: if companies expect prices (or costs, which are also prices) to increase in the future, they will increase prices in the present to not be left behind. Consumers act similarly, by increasing present spending and demanding higher wages (or switching jobs to get them) so their purchasing power isn’t reduced.

Finally, big picture aspects of the economy, such as the degree of competition, or productivity, that might increase or decrease inflation even at the equilibrium. For economies with more competition, equilibrium inflation will be closer to the target rate, because workers will have more power to translate a tighter labor market into higher wages and thus push up inflation. A more productive economy will also result in higher inflation, because productivity tends to be reflected in wages (to a degree that depends on competition, to be fair) and, over time, in more potential output. Lastly, if interest rates were to naturally go down, then inflation would be lower as well, since the Central Bank would have a more narrow margin to act.

Is inflation actually bad?

Yes, but. It depends, as everything, but we’re taking about inflation that is higher than some society considers acceptable. Imagine a country with 60% annual inflation, but that inflation only can vary up to 2% a year - so anywhere between 58% and 62%. Because that’s a very narrow range, then it’s not a very important part of life, and people can just do their daily business by assuming inflation will be 60% every year (or 5% a month) and being more right than wrong, on average.

So the problem with inflation isn’t whichever specific number it takes, but the fact that it should be stable and predictable. The reason most places have inflation goals (called targets) that aren’t very high is that higher inflation also tends to be much more volatile - so if you have 30% inflation, people do actually start to incorporate inflation into their decisions, which they don’t unless you add in this explanation.

The traditional reason for inflation being bad is that it makes prices noisy, i.e. makes them go up and down in ways that don’t actually reflect anything from the real economy. This happens because companies can’t tell if the price level is higher because all other prices are higher, or because consumers are buying more of their product, and those two have distinct responses (raise prices, produce more). Plus, there’s other more niche costs, like the cost of walking around looking for the best prices (“shoe leather costs”) and the costs of updating prices on things like labels and menus (“menu costs”). Looking at the evidence, most of these costs hardly ever materialize or aren’t very large.

To editorialize a bit, then, the main cost of inflation is lower purchasing power and not really hard to parse abstractions. This is relevant because our understanding of inflation comes from the idea that nearly any amount of it is painful - when, in reality, the most pressing issues (purchasing power, disincentives to lend or save, higher inflation being really volatile) don’t materialize at every level.

However, there might be another problem, a much bigger one, coming from inflation. To keep the economy stable, fundamentally, you need allocations of rights to economic resources, and rules for transferring those rights. In places that use money, this is frequently done by making money switch hands, or by lending and borrowing it. But here comes the problem: contracts that define these switchovers are set in nominal terms, without adjusting for inflation. Even if you include adjustment clauses, they will never be perfect, because you can’t rewrite contracts ex post.

There’s, basically, three “bad” types of inflation: high inflation, hyperinflation (which isn’t the same as high inflation), and deflation (negative inflation). We’ll go into some detail about each in the following paragraphs.

Whoever controls the money supply…

High inflation is, fundamentally, a different beast to regular above-target inflation. Exactly how high is high is a matter of debate, but let’s leave it at “somewhere above 10%” for now. So far we’ve talked about high-ish inflation as a result of excessive spending on the economy, and pointed the blame at either insufficient supply or bad government policy. But when inflation goes even higher (exactly how high is a matter of debate), it becomes “high inflation” - i.e. inflation that is really high (for reference, think like 20% or higher) and very persistent. There’s, broadly speaking, two big reasons for high inflations. The first is really irresponsible monetary policy, coupled with plain old bad fiscal policy. The second is that ending a high inflation on the low end is really costly, so everyone tries to prevent it from ending if they’re going to have to shoulder the costs.

When we talk about irresponsible monetary policy, we don’t talk about “interest rates were too low in the 2000s” type irresponsibility (a talking point that is also exactly wrong). It’s a different kind of badness. Normally, governments spend money by either raising revenue through various taxes and duties, or borrowing from the private sector. But as many TikToks (and also a whole school of thought) claim, they could just print the money they spend out of thin air - aka seigniorage. It’s patently clear that a country with independent central banks cannot have inflationary problems stemming from this - the Central Bank just decides how much seigniorage it’s willing to provide, and the Treasury has to suck it up and raise money otherwise. But if the Central Bank isn’t independent, then it has to finance whichever deficit is chosen ahead of time. If the amount of debt the government can take out isn’t infinite (though it is unknown), then at some point very close to that level, the authorities lose control of inflation and hyperinflation ensues.

Is this bad? It depends, but generally yes. The main reason why doing it is bad is that, if inflation is understood as too much money chasing too few goods, then you’re just adding money to the pile - and how bad the problem gets depends on how many goods are added as a result, i.e. how effective fiscal policy is at growing the economy. Funding the government through seigniorage isn’t the problem per se, but it is the main symptom of the problem - political weakness. If the government were strong, it would be able to impose costs on private agents (aka “taxes”) and redistribute those amounts to other agents (aka “spending”). A weak government, however, can’t do the first part, or not as much, because upsetting private agents could result in, say, a coup. Consequently, the money printing doesn’t happen because the government doesn’t want to raise taxes - it happens because it can’t, and is trying to have its cake and eat it too. For moderate rates of inflation (say, teens to high twenties) there’s not really much reason to worry, because the government is usually plenty powerful and just irresponsible, and most countries don’t tend to transition into higher inflations because the costs of inflation tend to be much starker and the costs of exiting are much lower.

The government can raise so much revenue through printing money - a Laffer Curve for inflation, so to speak, since at a higher level than that the costs of inflation become more pronounced and the amount earned is therefore smaller. Because of this, for any given amount to be funded, it’s generally agreed upon that there’s multiple equilibria for the economy - you can have a higher and a lower level of inflation, and the lower one is generally very unstable. In practice, this means that high levels of inflation, even ones that seem stable, are extremely volatile, which is caused by a lack of “anchors” for expectations; people know that betting on the wrong level of inflation is really costly, so they tend to overreact and push actual inflation higher. Anchoring some important variable, or promising something like “no deficit spending”, could at least eliminate the higher inflation equilibrium and make the other one stable.

What about the costs of ending it? When inflation is low, the anchors required aren’t particularly meaningful - for instance, pegging the exchange rate for a year or two. However, higher and higher inflations are also associated with more and more disfunctional governments, so they have to make a bigger show for everyone to buy into the stabilization scheme. But if high inflation is fundamentally about a weak government spending too much money, it means it’s spending on something: state-owned companies, subsidized credit, welfare, the military - whatever it is, it costs a lot of money, and that means someone has to lose out for inflation to end. If stabilization has large implications for distribution, then everyone will use their influence with the government to make sure they particularly don’t pay any of the costs. If the costs are spread around enough (usually, because the high spending isn’t concentrated in just one place), then everyone finds it best to just wait out someone else - and stabilization never occurs. If nobody expects inflation to either end or get out of control anytime soon, nobody budges - until whoever is either least politically powerful or least tolerant of inflation gives up, and has to lose out by some amount.

There’s a tight, important correlation between economic growth and high inflation. If the economy is stuck and unable to grow, for any of a variety of reasons, then the “too much money chasing too few goods” story isn’t just true - it’s self reinforcing. Everyone notices the new spending is unable to boost their income, so they just race to spend the gains before they’re wiped out by higher prices. This makes the predictability of inflation lower as it goes up, because the amount of time before prices are raised gets lower and lower. Consequently, it’s possible for a moderate inflation to become a high inflation to become a hyperinflation simply by not ending it soon enough - it al goes into autopilot.

To infinity, and beyond

The economy is a tangle of contracts and agreements; some written, like labor contracts, and some unwritten, like prices on a shelf. Their duration is influenced by how predictable future economic conditions are. If a price might be too low to replace the item it represents, then the contract gets “rewritten” and the price is increased. Similarly, if wages are found to be too low halfway through a contract, the workers ask for it to be rewritten - or just take a different job where they aren’t. The difference between levels of inflation - high, low, hyper - isn’t the specific figures (50%? 2%? 1000%?) but the durations: normal inflation has long contracts (mortgages, for instance), high inflation has month(s), hyperinflation has less than that - weeks, days, maybe even hours. Hyperinflation, simply put, is what happens when the value of currency drops so fast it starts tending towards zero.

Traditional images of hyperinflation tend to be from Europe in the 1920s - Weimar Germany, Hungary, Austria, Poland. Saddled by debts and war reparations, the governments turned to seigniorage at higher and higher levels to pay them off while the economy languished. Eventually, this reached a breaking point, and the inflation rate spiralled out of control. A more recent image, and more useful to understand the mysteries of hyperinflation, is 1980s Latin America, which features the high inflation duo of large government deficits funded by printing money, and really weak states.

Why do hyperinflations happen? For example, many countries had gigantic deficits funded by printing money in 1920s Europe, but only a handful actually had hyperinflations and not just (very) high inflations. What’s the difference? Normally, expectations. High and very high inflations are extremely volatile because of the mechanisms outlined above: any disturbance, no matter how small, sets off a race to the top where nobody wants their price to be left behind. If people expect the government to step in, this could be kept under control while purchasing power stumbles along. However, if future policies aren’t clear, then this race might not be stopped - it could be accelerated. For instance, a devaluation of the currency might make future costs unpredictable for companies, setting off price increases, setting off wage demands, setting off more price increases - and ever accelerating inflation.

Hyperinflation isn’t just a matter of prices growing exponentially faster - it signals a complete breakdown in the government’s ability to control anything about the economy. For example, take Branko Milanovic’s read on post-Soviet shock therapy:

The most important thing to know is that by the time reforms took place, the government no longer controlled the economy. This was therefore not a standard type of macro adjustment, like in (say) Egypt or Argentina, where the economy is behaving “normally” but there are fiscal or external-sector imbalances. In the Polish case, the economy was in a free fall, several exchange rates existed, hyperinflation was already present … Neither coercion nor incentives were any longer operative.
Under such conditions, macro stabilization had no alternative; its role was simply to validate what was already happening but in a most pernicious and chaotic way. Instead of closing eyes before multiple exchange rates, to unify them in one; instead of promising subsidies which the government could no longer deliver because its taxation ability collapsed, to eliminate them formally; instead of pretending to impose price controls (or the price freeze like in Yugoslavia) which the government had no ability to enforce, to legally allow all prices to be market-determined.
So the Big Bang was less a standard program of macro adjustment than a simple validation (or legalization) of what was there. It was acceptance of the fact that government’s writ was no longer.

Hyperinflation is just a very gruesome consequence of a much bigger problem: the government doesn’t control the economy anymore. The good thing about this, and about timespans getting so short in hyperinflationary conditions, is that getting out of one is “easy” - both the original Sargent paper and more recent studies seem to point out that by credibly committing to shake up the economy (usually in a pro-market manner, since hyperinflations tend to happen in state-dominated economies), hyperinflation can at least be reined in back into a regular high inflation.

It’s baaack

But what about negative inflation - deflation? If inflation is bad, then isn’t deflation good - because it means things are getting cheaper. This claim, and its mirror “higher inflation is good because it wipes out debt”, simply confuses prices with incomes, and misses out that the economy must not be especially good if either of those are happening.

Deflation (or very low inflation) occurs when aggregate demand doesn’t grow enough. In this scenario, the economy is in what Keynes called a liquidity trap: because interest rates are so low (what’s called the Zero Lower Bound), they can’t be cut anymore, and monetary policy isn’t enough (or at least not conventional policy) - but fiscal policy might not be either, because a liquidity trap is held together with pessimism. A clear example of this dynamic comes from Japan, at least in the 90s: the Bank of Japan decided it didn’t want to do unorthodox things (such as buying unusual assets), so it didn’t, and the economy got stuck in a sort of deflationary trap out of stubbornness. Paul Krugman, noting that the liquidity trap could come to other advanced economies, drew parallels between inflation and deflation: inflation happens when central bankers can’t promise to behave responsibly, and deflation when it can’t “credibly promise to be irresponsible”.

Ben Bernanke referred to this in 1999 as “self induced paralysis”, and demanded “Roosveltian resolve” - the kind that ended the Great Depression by promising to end it. This did come, in the mid and late 2000s, when Prime Minister Shinzo Abe and Central Banker Haruhiko Kuroda promised to do whatever it took until the economy got back on track - a package of monetary and fiscal stimulus known as Abenomics. This worked, in fact, and stimulated the economy significantly.

The success of Abenomics points to both monetary and fiscal policy being able to stimulate the economy even at this sort of low-level equilibrium: mostly by using the credibility that leads to this sort of scenario ever arising in the first place and promising "irresponsibility”, that is, big deficits and stimulus until the economy improves. Committing to a higher price level, with clear targets and no exit ramps, is the best way to get out of a liquidity trap and escape the Zero Lower Bound. Big, clear promises and even shifting to more controversial forms of policy - like nominal GDP targeting - could provide the necessary oomph missing.

Conclusion

Inflation is, after all, a monetary phenomenon - too much money chasing too few goods. High inflation is when the government gets involved, and hyperinflation is when it gets out of hand; deflation, when there’s not enough money. Stabilizing moderate, high, and hyper inflation isn’t easy, and requires a big array of tools, but generally involves reforming the economy, freeing the central bank from political pressures, and reducing the deficit. Getting out of a deflationary rut does the opposite, blowing up the deficit and pumping the economy with cash and big promises of high growth and maximum employment.

For a supply-shock inflation, there isn’t really a lot to do, unless there’s some clear policy to alleviate it. Given the current inflation (which I believe to be driven by supply-side issues and not policy), part of it will wind itself down, and part could be helped - for instance by building more houses to decrease rent inflation, investing in renewable energy to insulate the power grid from oil and gas prices, and repealing unnecessary tariffs and regulations that make ports and transportation systems less efficient - all while providing the necessary investments for them.