Author: Samuel Williamson

In Charles Dickens A Christmas Carol, Ebenezer Scrooge paid his clerk Bob Cratchit 15 shillings a week. What is the present-day equivalent of those 15 shillings? The MeasuringWorth comparator gives two answers.

In 2020, the real wage value of 15 shillings from 1843 is £75.28.
In 2020, the relative labour earnings value of 15 shillings from 1843 is £611.30

Why are they so different and what do they mean? The reason is that they are answers to two quite different questions. While they are both measured in British pounds, they should NOT be compared to each other.

The real wage is computed using a retail consumer price index of a basket of goods and services. In principle, by inflating the collection of goods (e.g., a goose) the average family could buy with 15 shillings, the index says that would be similar to a collection of goods an average family in today’s London could buy with £75.28. This is a totally unrealistic number!

Today a family of six could not afford the groceries to make a Christmas dinner for £75.28, let alone live for week. If we assume that Cratchit worked 10 hours a day for six days each week, that £75.28 rate would be £1.26 an hour. The current national minimum or living wage rate in the UK is £8.91 per hour. Even for workers under the age of 18 it is £4.60.

Price indexes do not reflect the actual increases in the cost of living over long periods of time because the market baskets purchased by representative consumers have changed so much. For example, food purchases constituted roughly 50% of the index in the 1840s; food is less than 10% today.* The Cratchits could not have purchased most of the goods in today’s basket because they didn’t exist in 1843.

The relative labour earnings is computed using an index of average earnings paid to workers. Thus, the “relative” wage of a worker who earned 15 shillings a week in 1843 is £611.30 today. At 52 weeks of work, that would correspond to about £32,000 a year. Bob Cratchit was responsible for maintaining Scrooge’s account books. The average UK accountant today earns a salary of £62,042 per year; the average bookkeeper, £26,848 per year. While Cratchit’s job may not have required a CPA, it involved much more than the tasks required of the average bookkeeper today. The fact Scrooge was paying Cratchit an amount slightly above that of an average bookkeeper speaks to Scrooge’s miserliness.

*The CPI and RPI are better designed to compute inflation rates than the “cost of living” over long periods on time.

* All prices in this article are of a gallon of regular gasoline as reported by U.S. Energy Information Administration who have published annual prices from 1929 to today.

The New York Times (May 25, 2021)  has published an article titled “What the Tulsa Race Massacre Destroyed.”* The article is a masterful presentation of graphics that give a wonderful feeling of what Greenwood, the prosperous black neighborhood in Tulsa, was like in May of 1921.  The article started out saying “The Tulsa Race Massacre of 1921 killed hundreds of residents, burned more than 1,250 homes…” Three paragraphs later it states that on one destroyed block “there were four hotels, two newspapers, eight doctors, seven barbers, nine restaurants and a half-dozen professional offices of real estate agents, dentists and lawyers.”  That was just one block among many destroyed.  The story of the destruction of life and property in the first two days of June one hundred years ago is tragic and makes one want to think “if only this did not happen, what would the back community of Tulsa accomplished?”

The article has nine authors and three pages explaining their methodology.  There are 20 different sources and 14 other people listed as assisting in the production. It was a big undertaking and very impressive.

So why can’t the New York Times learn how to measure relative worth?  In the fifth paragraph of the article it says: “The financial toll of the massacre is evident in the $1.8 million in property loss claims — $27 million in today’s dollars —”

This number come from one many cost of living or purchase power calculators that one finds on the internet.  Except for ours, they all use the CPI to inflate a value from the past and in most cases the answers are one dimensional and misleading.  In this case, the error is spectacularly bad.  A quick consideration of replacement costs shows this.

For example, 27 million divided by the 1,250 homes comes out to $20,000 a home and that does not take into account all the business lost. A simple search of google asking “what is the cost of building a new hotel?” gives the following answer:  “The national average range is $13 million to $32 million with most people spending around $22. 1 million on a 3-star hotel with 100 rooms.”  Restaurants and office buildings might cost from $250 to $500 a square foot to build. The article says that some half a dozen churches were burned, one of the was large brick Mount Zion Baptist Church, which one can imagine would take tens of $millions to re-build today.  Remember that Greenwood not only lost buildings, but also water, sewer and power utilities.

The MeasuringWorth relative worth comparator, Purchase Power Today of US Dollars, gives seven choices for the relative worth of a financial amount in the past.  They range from $21 million to $534 million and are measured using price, income, household expenditures and output indexes.

There is no doubt that the output measure should be used to measure the relative value of this $1.8 million in 1921.  What the comparator shows is that in 1921 the share that $1.8 million was of the GDP is the same as the share $534.06 million is of GDP today.  Another way to put it is to say if we spent over $0.5 billion on restoring the Greenwood neighborhood today, it would represent the same percent of the economy’s output that was destroyed a hundred years ago.

A second best choice for the relative worth would be to use the wage measure and that would say $1.8 million in 1921 has a relative wage of $149.14 million in wages today.  This can be interpreted as saying it would take this much to hire as many workers today as the $1.8 million would have hired then.

This is not the first time the New York Times has used inflation calculators to come up with very bad estimates of relative worth.  Part of the blame may be search engines that lead to calculators that “adjust for inflation” simply by using the CPI and lead users to believe there is one definitive answer.  Unfortunately, lots of economists do the same.

Please do not use the comment box below.  You can send me an email at sam@mswth.org if you wish.

• The New York Times may not allow you to connect from the link here.  The address of the article is: https://www.nytimes.com/interactive/2021/05/24/us/tulsa-race-massacre.html

The March 11th column by David Brooks in the New York Times stated that:

“As Michael Hendrix of the Manhattan Institute notes, America spent $4.8 trillion in today’s dollars fighting World War II. Over the past year, America has spent over $5.5 trillion fighting the pandemic.”

This is the worst possible use of using the CPI to compare relative worth (See the essay on this issue on our website.)

It makes far more sense to use the GDP to compare these expenditures. This table shows the GDP and armed forces expenditures for the war years in current year values. The third column reports those expenditures as a percent of GDP. The last column is that percent times the today’s GDP of $21 trillion.

For three years the war expenses are over 30% of GDP . If we take the relative GDP share of these expenditures during the four years from 1942 to 1945, the total of those numbers is $24 trillion, four times the $6 trillion to be spent on relief today.

WWII was a period in our history where this country expended more of its resources and suffered more sacrifices to defeat an enemy than any time since the Civil War. It is ridiculous to compare it to the outlays today.

As an economic historian and a political junky (I have buttons I wore during the 1952 Presidential campaign) I am always a bit frustrated when people are blaming or giving credit to an incumbent president for the state of the economy.  While sometimes it is true, the causes of the movement of the economy are more complicate than the activates of one agent.  It is the same as measuring relative worth.

I have written an essay titled  The Relative Performance of the Economy under the Presidents of the United States from 1900 to 2020.  What I do is compare the performance of several aggregate economic variables over the different administrations. The goal of this new essay on our site is to help the reader better understand how the economy performed during each of the presidential terms from 1900 to 2020.

QUESTION: In the MeasuringWorth website, in what category would land prices fall? Just doing the Commodity real value exchange doesn’t seem to factor in the skyrocketing New York City real estate prices in the 19th century, which is my concern.

ANSWER: As we try to explain, questions of relative worth depend on the context. If you were talking about land prices today, what would you be comparing them to? Would it be the cost of a bundle of consumer goods?  Probably not, so the same would apply to the 19th century as you suggest.

Those who I know that own or are looking to own in NYC these days think about two things, what the total monthly mortgage, tax, etc. payment will be (very dependent on interest rates), and the share that payment is of their income.

There are other ways of looking at the question that might use other indexes.  You can check out our tutorials.

QUESTION: In 1896 the McKinley presidential campaign spent at least $3.5 million. In 2016, Trump and Clinton spent about 2.4 billion. What calculation would you use–purchasing power, size of economy–to determine which was the most expensive campaign?

ANSWER:  To start with, the $3.5 million was only what the campaign raised. One of McKinley’s biographies quoted in Wikipedia reports that by including fundraising by state and local committees,  “Estimates of what Republicans may have raised in total have ranged as high as $16.5 million.” As an aside, J.D. Rockefeller himself donated $25,000, a relative income value of about $7 million today.

McKinley ran the campaign from his front porch in Canton, Ohio; he did not go to rallies or take whistle-stop train trips. Much of the money was spent on printing pamphlets and hiring hundreds of stump speakers to campaign on his behalf. We can think of all these stump speakers as the 1896 version of today’s TV ads. (At least then you could choose not to go hear them instead of having them in your living room 24 hours a day.)

The relative value of $3.5 million today using the CPI index is $108 million; for $16.5 million, it is $509 million. If campaigns then and now spent most of their money on printing, then that would be a reasonable measure, although looking at relative printing costs would be better. Printing, however, is a very small cost of campaigns today.

Today’s campaigns pay millions for ads on TV and social media as well as rallies and personal appearances. I am going to guess that TV ad rates are proportional to the cost of producing them and that cost is proportional to the skilled wage. In 1896, the campaign was spending large amounts hiring stump speakers and their cost would also be proportional to the skilled wage. So if we look at the relative values of $3.5 million and $16.5 million using the production worker compensation index, we get totals of $877 million and $4.7 billion, respectively.

We do not know how state and local Republican committees spent their funds; it could have been support for McKinley or candidates for other offices. In any event, we have that the McKinley presidential campaign raised close to a relative worth today of nearly $5 billion for the 1896 election.

QUESTION: I’m trying to determine the actual value of 6 000 000 CAD$ of public debt in 1941. Since you have references for only US, UK, Australia and Spain, I figured relative output/economic cost for any of those nations does not work since it’s not relative to Canadian GDP. What do you suggest?

ANSWER: If the relevant data isn’t on MeasuringWorth, one has to look elsewhere, in this case Statistics Canada and the Historical Statistics of Canada, both of which are online.  For example, if one looks at GNP at market prices, the ratio of that figure in 2018 to 1941 is roughly 200.  Assuming the $6 mil Canadian public debt is a nominal figure, that compares to $1,200 million Canadian today.  But, more to the point, if users identifies the measure(s) to use, then they need to find the relevant statistics for the chosen years and do the calculation themselves.

QUESTION:  I wish to compare the relative value of several large donations made to American universities between 1829 and 2015. Most of these were bequests and the funds were used for both operating expenses as well as the construction of new facilities. There is no indication of the relative “split” and clearly it will vary with the individual circumstances of the receiving institution. If it cannot be determined what the bequest was used for — which is the case here — what criteria should be used to determine the best method to use?

ANSWER:  You supplied several examples; let’s concentrate on just three:

1829. James Smithson to found the Smithsonian Institution gave $500,000 in 1829.
1830. Stephen Girard bequeathed $2,000,000 to a number of education causes in 1831.
1831. John Hopkins gift of $7,000,000 in 1873 was used to found a University bearing his name.

A price index would not be very useful since the commodities bought then and now are so different.  If we knew how much of each was used for construction costs, then using one of the wage series might provide a useful comparator.

Because the gifted funds were used to purchase both consumer and producer (capital) goods, the two choices I would recommend are 1) the income value (using the relative GDP per capita) and 2) the economy share (relative share of GDP).  The first is a more general measure of how the gifts compare to the average earnings of the day. The second shows their relative value to the economy at the time.

For Smithson, the relative value in 2018 for each of those two values is $420 million and $11 billion.

For Girard, they are $1.6 billion and $39 billion.

For Hopkins, they are $2.1 billion and $16.2 billion.

By comparison, Michael Bloomberg has given $3.3 billion to John Hopkins over the years; the Bill and Melinda Gates’ foundation is worth around $50 billion.

Hello. We are a newspaper based in Barcelona, Spain, founded in 1881. We are writing an article about the wealthiest people in history, inspired by an article in Bloomberg Businessweek on May 21st citing your site. I would like follow up that pieces with a few questions to help us with our article. Thank you in advance.

QUESTION  1,  Is the ratio of wealth to GDP the best way of comparing fortunes through history? Which are the main alternatives?

ANSWER.  It is the most often used and has the advantage of not having to deal with the mix of output.  (There were no antibiotics available to the richest person on earth before they were invented.)  By using the share of GDP index you are comparing what we call the “economic power” of the individual.  That power can be used in the market to control the use of resources.  Of course absolute monarchs have this power and do not have to use the market, so in some sense the size of their wealth is not as important as their power to expropriate.

QUESTION 2.  Can we consider John D. Rockefeller as the wealthiest person in history? Some rankings cite Genghis Khan, for example. Do you consider this kind of rankings as accurate? Can we really compare wealth through history or we have to limit it to its context?

ANSWER: It is useful in comparing two private individuals that have a large amount of wealth in a market economy and cannot expropriate for comparison over time.  Using a wage or income index could also be instructive. We seem fascinated at how much more CEOs make today as a proportion of their workers.  It could easily be of interest in the past.

As for Genghis Khan, he was an absolute ruler, he did not need monetary wealth to command resources, so I do not see an easy way to compare him with Rockefeller.  Perhaps with Alexander the Great?  And then there is Gaius Appuleius Diocles, the Roman chariot racer who was regarded as the richest athlete in history and perhaps the richest person in the Roman Empire.  How do we compare him?