Data
2015 10 05
Today’s post recapitulates some of the points made in Improvements to Macroeconomic Data back in early 2024, but:
a) I have more and different readers now and those needed improvement have still not occurred 😊
b) Things have happened
a. The President fired the head of the Bureau of Labor Statistics
b. The ways that housing and long-life durable goods like automobiles are treated in price statistics has been questioned
c. I’ve thought of more improvements
Let’s go.
Short term Inflation Expectations.
Since inflation affects decision-making about future transactions, I take as given that more accurate expectations about inflation given the information that is available at the time of the decision is useful. As the common expectation of many people is more likely to be accurate than any one person, it is useful to have such a common expectation.
To that end the Treasury already issues “Treasury Inflation Protected Securities (TIPS) at 5- and 10- year tenors. These pay no “interest” but instead pay the accumulated percentage increase in the Consumer Pirce Index. These are paired with securities of the same tenor and other characteristic that do pay interest. Both are traded and by comparing the prices of the two securities one can deduce by how much inflation the transactor of a TIPS expects the CPI to increase. In this way we can know what traders expect for inflation over a 5- or 10-year period. Those data are available here and here.
Knowing expectations also for shorter periods would be useful, especially for making and critiquing monetary policy, but for other natural and policy shocks as well. Movement in a
5-year indicator does not have much information about expectations for the next year. Treasury therefore should issue additional TIPS at 1-, 2-, and three-year tenors so that these shorter-term expectations would become visible.
Wages and Prices
Economic news and analysis often refer to changes in wages, especially in the context of inflation. Is wage growth incompatible with the Fed’s inflation target? Has inflation reduced real wages? Have wages of one kind of worker – wage level, geography, racial/ethnic categorization, immigration status -- increased more than another kind? These are all interesting questions. Unfortunately, none can be well answered because the Bureau of Labor Statistics does not produce wage indexes! The Bureau produces wage income unit value [1] indices, but as is well known (but seldom mentioned) the two are not the same.
No matter how narrowly defined the group under examination (unless all the members have identical prices or wages), a unit value index suffers from defects of composition compared to a price index. This was especially bad during the pandemic. Initially many low-wage workers like restaurant serving staff lost their jobs because of lockdowns but fewer higher-paid office workers did. If you have seen a graph that showed a spike in “wages” during the pandemic, that is the explanation. The labor income unit value index rose sharply although no one’s wages increased.
Bureau of Labor Statistics should develop and publish wage indexes based on comparison from one period to the next of wages paid for the same task.
In general price information for goods do not suffer from composition effects as goods can be defined precisely enough that all items have identical prices, but imports and exports are exceptions. Because of the way the data are collected, basically at the point of entry/exit, it is difficult to define the imported or exported item precisely and so the customs classification is used to calculate the unit value. Difficult, but not impossible and BLS should put resources into doing so.[2]
Data Categories
Both price and quantity data follow an activity classification. This is all that is necessary for taxation and regulation in a closed economy, no imports or exports, or in which there is no overlap between domestically produced goods and traded goods as tends to be the case for very simple if open economies and more complex but largely closed [3] economies. In a large and (still) largely open economy like the US, distinguishing tradeable from non-tradable goods and services can be useful, particularly now that tariffs can change from week to week and policy generated uncertainty can affect exchange rates.
Expectations of Economic Growth
It would be useful to have an indicator of market opinion of expectations for the economy. People try to use the stock market for this purpose but as best it is only an indicator of expectation of stock returns. Therefore, the US Treasury could sell a security, a “Trillionth,” that pays a proportion (x trillionth) of GPD at some future date instead of interest.
Mechanically it would be like the TIPS. Payment is made against the value of future economic data, the GDP instead of the CPI. GDP is around 30 trillion so the Treasury could sell a security that promises to pay 25 Trillionths of future GDP at some future date with a face value of, say, $1000 and the premium/discount would indicate the expected rate of growth. Like TIPS it could be sold for various tenors as TIPS ought to be.
Such an indicator could be especially useful to the Fed, as TIPS already is, for making monetary policy, but it would be useful as well as a real time reading of market “opinion” of economic management. In addition, it would be a good investment vehicle for pension funds and individuals’ retirement savings.
Image Prompt: Person holding a bucket as to collect rainwater (data), but only a few drops are falling
.
[Standard bleg: Although my style is know-it-all-ism, I do not really think that, knowing that I can be mistaken or overstate my points. I’m also aware of the amazing range of views and experiences among readers. Bring those to be bear by commenting on these posts. Both other readers and I will benefit]
[1] A “unit value” of a category is the aggregate value of that category divided by the number of individual items in the category.
[2] Part of the difficulty is that “customs” classifications of goods (on the basis of what they contain) does not correspond to the “industrial” classification (on the basis of how things are is made)
[3] The aim of tariffs generally is to prevent much if any such overlap and the the partial effect is the corresponding implicit tax on exports products.