Do People Really Have Inflation Expectations?

A common feature of macroeconomics is to run models assuming that individuals in the economy have inflation expectations. This flows naturally from assuming that they have rational expectations (can accurately predict the future and fully understand how the economy works). The standard reason for why we need inflation expectations is that during the 70s the government tried to trade off low unemployment in exchange for higher inflation. However, (so the story goes) people got wise to this trick and began to expect higher inflation and demand higher wages to compensate. As a result, government policy lost its effect and unemployment rose. Since then, economic modelling and discussions of unemployment and inflation must contain some provision for what consumers expect inflation to be.

But do people really have inflation expectations? And how accurate are they? Do consumers really pay much attention to such economic data, make predictions for the future and constantly revise their estimates in line with the results? Or is this yet another example of economists ascribing superhuman abilities to ordinary people who in reality have little interest in topics that fascinate economists?

To find out a survey entitled “What Do The Irish Know About Economics” will be a good guide. Ordinary people were asked questions about important economic data and provided multiple choice answers containing bands. So people were asked “What is the current inflation rate in Ireland?” and asked to choose between a) 0-2%, b) 3-5%, c) 6-10%, d) 11-15%, e) 16-20%. So precise knowledge wasn’t required, merely a rough idea. This should be straight forward as Irish inflation rate has been pretty stable and low since the early 80s. However only 42% correctly knew that the level was between 0-2%. However can inflation expectations play such a large role in macroeconomics if the majority of people do not even roughly know what the inflation rate is? Consumers with rational expectations would have no such problems, but consumers in the real world do.

Worryingly, 13% thought it was above 11%. In other words a sizeable minority believed we were in a position of high inflation last seen in the oil crisis. People polled poorly across all indicators. Only half knew that unemployment was between 11-15% and that unemployment benefit was €150-200. Only half knew what the minimum wage was, the rate of GDP growth or that the size of the Irish economy was somewhere between €100-250bn (an extraordinary broad range that baffles the mind that anyone could get wrong). A majority of people were also wrong in where they thought most government revenue came from and where most was spent. Only a minority knew the correct size of the public service and these are only the major data I’m mentioning. Answers on more obscure data such as trade balances and financial matters were even lower still.

Most surprisingly of all, despite dominating the headlines constantly for the last five years, only 38% knew what the budget deficit was. This was despite enormously broad levels were given (the correct answer was between €1-20bn) most people thought the deficit was at least double (if not triple or quadruple) its actual level. The same is true of the national debt as well with only 40% knowing it is between €100-250bn.

All of which poses problems for assuming inflation expectations and rational expectations in general. It is assumed that consumers know all the relevant economic data or are at least on average correct. However, in reality most people have deep misconceptions and struggle to answer even the broadest questions about important economic data. How they can they have accurate inflation expectations? Perhaps more people would know about inflation if it was more of a problem and in the news more, but seeing the poor knowledge of budget deficits (which are constantly in the news) makes this unlikely.

The fundamental problem with the assumption of inflation expectations is that it presumes people understand statistics. Unfortunately, there is plenty of evidence that people overly rely on anecdotes and unrepresentative examples. An example is a survey recently conducted by the Irish Times entitled “Public Opinion Is One Thing, Facts Are Another” with sub headline “Poll shows views held by the public about Irish life have little grounding in reality”. It found that “the public has some profoundly mistaken beliefs about the key economic and social facts of Irish life”.

When asked about the level of foreign-born people in the country, the average estimate was that they were 25% of the population, double the correct figure of 12%. Two-thirds thought the level of crime was rising, when in actual fact it is declining. The vast majority thought the unemployed were the largest recipients of welfare, when in reality it is pensioners. However, by the far the grossest misunderstanding came with perceptions of where the government spent most of its money. Half of people thought politicians were the number one recipients, with one quarter saying welfare. In Ireland, €102 million is spent on parliament while €20 billion is spent on welfare.

Similar disconnects have been found in Britain too. The British public estimated that the teen pregnancy rate was 15% whereas in reality it was 0.6%. It was estimated that 24% of the welfare budget was wasted on fraud, whereas the rate is only 0.7%. People thought one third of the population were immigrants (the real figure was only 13%) and that 24% of the population were Muslim, when only 5% are. And so on.

There are plenty of examples of American misunderstanding of statistics too provided in comedic style by Bill Maher. (Statistics begin after one minute).



It was found that these misconceptions were often due to sensationalist headlines and political spin. The reality of low welfare fraud does not sell papers, create provocative headlines or grab readers attention. They are also due in large part to representative bias where people think events they can easily call to mind are more likely to occur, but dramatic unlikely events are easiest to remember. As a result people’s perception can be greatly different from reality.

It is hard to expect consumers to know what the inflation level is when even economists can’t agree. There are two ways of measuring inflation, the Laspeyres index and the Paasche index. The Laspeyres used to be the most common method and is calculated by taking a bundle of goods and comparing their prices in two different years. However, it assumes that people don’t change their spending patterns in response to price changes. The Paasche index takes this into account and results in a lower reported inflation level. There is no consensus among economists as to which level is preferable and economists usually split the differences between the two. So if the professionals aren’t even sure as to what the true level of inflation is, what chance does the average person who has never studied economics have?

Then there are those who do not trust government sources and believe that they are deliberately understating the level of inflation so as to improve their chance of getting re-elected. The most well known example of this is ShadowStats which argues that the true inflation rate is treble the official government level. I personally am not much of a fan of ShadowStats but it is possible in theory for government data to be incorrect. Even if economists do not take ShadowStats seriously (and some do) ordinary people may and Americans in particular are very suspicious of their government.

It is hard to expect consumers to have an accurate and unified view of inflation when there is no one inflation rate, but rather hundreds, each one for a different good. So part of the reason Irish inflation rate was so high was that property prices were massively increasing. However if you weren’t buying a house, this wouldn’t affect you and your inflation expectation would be correspondingly lower. So it is hard to speak of one inflation expectation in the economy considering its hugely diverse nature. Unless wages were somehow individualised to meet each person’s individual view of inflation levels, they will never match (the problem is normally gotten around by assuming there is only one representative consumer in the economy, which doesn’t make things any better).

But even if people did have accurate inflation expectations, they may not be able to do much about it. The reason expectations are included is that it is believed that workers would increase their wage demands in the face of higher inflation. However, since the 70s there has been a large drop in the size and strength of unions, particularly in the private sector. American unions have shrunk so much that they are rarely mentioned anymore. So how do non-union workers ensure their wages keep up with inflation? The unfortunate reality is that workers have very little power over their wages. They could quit and get a different job, but this isn’t a reality for most workers. Most skills, contacts and relations are firm specific and would be lost in the event of leaving the job. There are also information problems that prevent workers from being fully aware of the labour market and making appropriate decisions. For these reasons and move, most workers are unlikely to take the drastic step of moving job. As a result there is little pressure on wages to adjust even to keep up with inflation.

Businesses may not be able to grant wages to keep up with inflation if inflation is rising due to costs outside their control. For example, if the price of oil increases by a large amount, workers will want higher wages to compensate. However, businesses will be unable to provide them as their costs have gone up but not their margins. So we could easily have a position where workers demand higher wages but are refused.

So while inflation expectations are a common macroeconomic tool, I am dubious of their usefulness. It is asking too much to expect ordinary people to understand economics as much as economists and to be able to fully act on their beliefs. Most people don’t pay enough attention to the future and take little interest in economic data (talking about inflation expectations to ordinary people is guaranteed boredom on their behalf). People don’t know or care enough for inflation expectations to have a large effect on the economy.

13 thoughts on “Do People Really Have Inflation Expectations?”

  1. And if you think it is bad in Ireland, here in the USA where 12% of Americans think Joan of Arc was Noah’s wife, it is worse. Most can’t ditinguish between the budget deficit and the national debt.

  2. Interesting. RE is a mathematical assumption slotted into DSGE macro models, it doesn’t mean a whole lot in reality. Even prominent macroeconomists consider it nothing more than a built in assumption to make the models work. Sure, back in the day, it was this intrinsic belief that individuals really had inflation expectations. But I might note, that individuals – you and me, are not assumed to have expectations. It is assumed that trade unions, thinking on behalf of workers, are monitoring the CPI and analysing real wages. They than bargain on behalf of workers. What if you’re not in a union – yeah…? Also, financial markets – bond investors in particular, rating agencies, have expectations based on their modelling of the markets – when the Fed says it will “keep rates below X% until employment picks up” – these market players build this into their models and form forecasts/expectations.

  3. “The most well known example of this is ShadowStats which argues that the true inflation rate is treble the official government level. “

    Regarding Shadowstats, its “alternate” inflation estimates are unreliable. They do not even recalculate inflation data at all: what they do is add a questionable arbitrary constant to the government inflation rate:

    Moreover, the “Billion Prices Project” at MIT data provides very strong confirmation that the official CPI estimates are a reasonable measure of inflation, as I point out here:


    1. Oh, and I forgot to add: once Shadowstats’ “alternate” inflation data is rejected as false, all of its “alternate” real GDP measures are rendered worthless, because the real GDP calculations are based on the worthless inflation figures.

      1. I decided to reserve judgement on Shadowstats and instead focus on how some people (rightly or wrongly) can believe government figures are biased. I’m personally pretty sceptical of Shadowstats myself but (and I could be wrong) I think Steve Keen recommended them in Debunking Economics. Not that proves anything but that it is an issue for debate.

    1. It’s called the “consumer” price index, not the “investor” price index for a reason. But, you make a good point, which is that low consumer price inflation may mask rampant asset price inflation. This was so during the 1990s and 1920s in the U.S.

  4. Even if consumers only have the foggiest ideas of inflation, they may still express their expectations of accelerating inflation through their purchases. Manufacturers’ expectations of inflation are also important, because if they’re wrong, expect the business cycle to be worsened. An example of this is the 1937 recession in the U.S.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s