Inflation is not an Aggregate Quantity

[John Creighto]

Economists try to reduce inflation to a single quantity in order to plot changes in such things as real income and real GDP. However, the changes in buying habits do not change equally for people of all income levels. 85% of productivity increases of the 35% increase in productivity we have scene since 1989 has gone to the top half of the population. Consequently if the inflation index is based on the average buying choices then this measure of inflation would much better represent the inflation scene by the top half of the income earners then the bottom half.

While this top half might see a small rate of inflation the bottom half of the population could see a much larger rate of inflation due to increased housing costs which arises from the greater proportion of the buying power which goes to the top half of the population.

Consequently the rate of inflation is not an aggregate quantity but rather a function of income. Further real changes in income for each earning bracket do not map linearly to any single aggregate quantity representing inflation such as a price deflator or price inflator.

http://en.wikipedia.org/wiki/Median_household_income

 

[talk2glenn]

Consumer demand models can be aggregated by using summation indexes. That is, you are correct that you cannot simply graph an aggregate budget curve using P1, P2, and m (where P1 and P2 are the prices of two hypothetical goods, and m is the net income in the economy), but you accurately graph aggregate demand using the sum from 1 to i, where i is the number of consumers in the economy/model, of the function D(P1,P2,m_i).

The summation method accurately reflects the differences in aggregate consumption patterns given particular income distributions (so two economies where m net = 1000, but with different income distributions between rich and poor, will plot different aggregate demand curves).

I can assure you that any possible scenario you can consider has been worked into the econometric models. The Consume Price Index accurately reflects the effective change in aggregate prices for the economy as measured along consumer demand curves, after the income and substitution effects.

 

[John Creighto]

Consumer demand models can be aggregated by using summation indexes. That is, you are correct that you cannot simply graph an aggregate budget curve using P1, P2, and m (where P1 and P2 are the prices of two hypothetical goods, and m is the net income in the economy), but you accurately graph aggregate demand using the sum from 1 to i, where i is the number of consumers in the economy/model, of the function D(P1,P2,m_i).

The summation method accurately reflects the differences in aggregate consumption patterns given particular income distributions (so two economies where m net = 1000, but with different income distributions between rich and poor, will plot different aggregate demand curves).

I can assure you that any possible scenario you can consider has been worked into the econometric models. The Consume Price Index accurately reflects the effective change in aggregate prices for the economy as measured along consumer demand curves, after the income and substitution effects.

I think my point may have been missed. The substitution effects will be different for each income bracket and consequently the effective inflation will be different for each income bracket. All the aggregating does is weights consumers by demand which has the effect of giving more weight to higher income earners then lower income earners when their is a large disparity between the average and median income.

 

[John Creighto]

Not only is inflation not an aggregate quantity but some people believe the indexes are deliberately manipulated:

Burns was disingenuous. The U.S. Treasury "prints" money but Burns’ Fed bought deficit-funding Treasury securities at a price convenient to the national purse. This is the chief mechanism available to a Federal Reserve chairman that fulfills Burns’ warning to his students: "Excess government spending causes inflation."

Fed officials, including Bernanke, have taken to blaming the federal deficit for our ills, but the same holds true today. Bernanke has bought over a trillion dollars of mortgages and continues his "quantitative easing". This is a deceptive name to fulfill the Fed’s role as waste dump for discredited securities and euthanasist of the People’s currency. These are crimes against humanity.

As inflation rose, Burns applied tactics then current among Stasi counterintelligence colonels. After oil prices quadrupled in 1973, Burns told his staff to remove energy costs from the Consumer Price Index. Burns’ rationale was the Yom Kippur War, over which the Fed had no control. A few months later, with food prices raging, Burns told his staff to remove them from the CPI calculation. Burns claimed the disappearance of anchovies off the coast of Peru was the cause of food inflation, and beyond the Fed’s jurisdiction. In time, Burns discarded used cars, children’s toys, jewelry and housing – about half the costs consumers battled in their daily struggle with rising prices.

http://www.creditwritedowns.com/2010/10/central-bankers-are-paid-to-lie-buy-corn.html

 

[CRGreathouse]

I think my point may have been missed. The substitution effects will be different for each income bracket and consequently the effective inflation will be different for each income bracket. All the aggregating does is weights consumers by demand which has the effect of giving more weight to higher income earners then lower income earners when their is a large disparity between the average and median income.

High-income earners aren't really counted much when calculating the CPI-U (and not at all for the CPI-W, but I digress). The CPI-U "is based on the expenditures of almost all residents of urban or metropolitan areas, including professionals, the self-employed, the poor, the unemployed, and retired people, as well as urban wage earners and clerical workers", while the CPI-W measures only the last group (urban wage earners and clerical workers).

You seem to be confusing the CPI with the GDP deflator, which is a good measurement of the cost of production in the country but not a good measure of expenditure changes, except possibly for the wealthy.


In fact, for the reasonably well-to-do (say, top 15%), the CPI is no longer an appropriate measurement. That's why organizations like AIER are able to collect and sell data for inflation for these high earners.

 

[talk2glenn]

I think my point may have been missed. The substitution effects will be different for each income bracket and consequently the effective inflation will be different for each income bracket. All the aggregating does is weights consumers by demand which has the effect of giving more weight to higher income earners then lower income earners when their is a large disparity between the average and median income.

No, you have missed my point. Aggregating does not simply weight consumers by income; it includes the consumption of each individual from the index. Are you familiar with summations in statistics and algebra?

Effective inflation may vary individually from the aggregate. This does not imply that an aggregate cannot be found.

 

[John Creighto]

No, you have missed my point. Aggregating does not simply weight consumers by income; it includes the consumption of each individual from the index.

The amount of demand someone contributes to the sum will depend on their income. If you what the result to be income independent then you should sum their fractional demand (demand divided by income (I just made this word up.)). However, people in different income brakets will make different consumption choices so their is no guarantee the aggregate maps well to a demographic.

Are you familiar with summations in statistics and algebra?

You are being patronizing.

Effective inflation may vary individually from the aggregate. This does not imply that an aggregate cannot be found.

It is not a question of weather the aggregate can be computed it is a question of if it is meaningful. If it doesn't accurately reflect the changes in living costs scene by half the population then I would think their should be a more meaningful way to report the statistics.

 

[John Creighto]

High-income earners aren't really counted much when calculating the CPI-U (and not at all for the CPI-W, but I digress). The CPI-U "is based on the expenditures of almost all residents of urban or metropolitan areas, including professionals, the self-employed, the poor, the unemployed, and retired people, as well as urban wage earners and clerical workers", while the CPI-W measures only the last group (urban wage earners and clerical workers).

You seem to be confusing the CPI with the GDP deflator, which is a good measurement of the cost of production in the country but not a good measure of expenditure changes, except possibly for the wealthy.


In fact, for the reasonably well-to-do (say, top 15%), the CPI is no longer an appropriate measurement. That's why organizations like AIER are able to collect and sell data for inflation for these high earners.

I am glad that economists are making some attempts to break down the CPI into separate demographics. I am not sure if this is sufficient resolution but suspect it is not. I mention the gdp deflator because because it is based on inflation indexes. It is relevant because it is used to graph things such as real wages. I would like to see this broken down into separate demographics.

 

[brainstorm]

I am glad that economists are making some attempts to break down the CPI into separate demographics. I am not sure if this is sufficient resolution but suspect it is not. I mention the gdp deflator because because it is based on inflation indexes. It is relevant because it is used to graph things such as real wages. I would like to see this broken down into separate demographics.

Would different demographics have mutually exclusive "shopping baskets" of goods? If different levels of inflation would be reported, how would you expect this to influence policy and investment decision-making?

 

[talk2glenn]

I am glad that economists are making some attempts to break down the CPI into separate demographics.

Apologies, but you still don't seem to understand how CPI is constructed. It is itself an aggregate function. If you're interested in measuring price changes for an individual member of the index, you don't start with the CPI and work backwards. You start with the consumption basket for the individual or individuals of interest, and build a relevant index.

This is done regularly at the request of policy makers. The United States CPI (and its respective components) is simply the most widely reported of the indexes generated.

This is the generic function for a price index. As you can see, the chosen components and their assigned weighting determines the output value. The components of the official Consumer Price Index are chosen and weighted according to the aggregate demand functions for the measured economy (called the Consumer Expenditure Survey). This is a function of available data for consumption patterns by region, type, and income. This data is collected using statistical sampling. Economic units are represented in the survey at the same rate as their representation in the actrual economy. To address your concern specifically, if 10% of the population earns approximately $20,000/year, then 10% of the sample would be wage earners making approximately $20,000.

The practical consequence of this is that, to repeat myself, two economies of an identical scope (as measured in GDP) but with different relative distributions of wealth would generate two different CPI's, using the BLS methodology.