How Government Budget Deficits Reduce Wages and Raise Profits
In recent years there have been growing complaints over slow growth in wages compared to profits. Those who make the complaints usually offer little in the way of explanation. Here is a part of the explanation: growing government budget deficits.
Government budget deficits are financed partly by the creation of new and additional money. But for the rest, they are financed by selling government securities to the citizens, who pay for the securities with money that already exists and which is part of their savings. If the government had not been running at a deficit and had not needed to sell these securities, the citizens would have used most of the savings with which they buy the government securities to buy corporate securities and in other ways to make their funds available to business firms.
Those savings, in the hands of business firms would have been used to purchase capital goods and to pay wages. These wages, however, never come into existence if the savings out of which they would have been paid are diverted to the government to finance its deficit. Thus, wage payments in the economic system are smaller because of government deficits.
Yes, it is true that the government itself pays wages to some extent. But it is unlikely to do so to the same extent as do business firms. And to whatever extent the additional wage payments it makes out of the proceeds of its securities sales are less than the wage payments that business firms cannot make because of the diversion of part of what would have been their capital funds to the government, total wage payments in the economic system are reduced.
In addition to this likely reduction in overall wage payments in the economic system, the effect of government budget deficits is an increase in the aggregate, i.e., total, economy-wide, amount of business profits. Here’s why.
Whether business gets money to finance its purchase of capital goods and payment of wages, or the government gets money to buy goods and services, including meeting its own payroll, and, of course, simply to give money away in carrying out various welfare-state functions, the amount of business sales revenues in the economic system will be pretty much the same. This is because a dollar from the sale of a good to a business firm and a dollar from the sale of a good to the government is still a dollar of sales revenues. A dollar from the sale of a good to an employee of business or dollar from the sale of a good to a government employee is again still a dollar of sales revenues. And finally, a dollar from the sale of a good to someone who has received his money under one or another of the government’s welfare-state programs is no less a dollar of sales revenues than a dollar spent by someone who had earned that dollar.
But here’s a crucial difference: The dollars spent by business firms in buying capital goods and in paying wages sooner or later show up as costs of production. Those costs of production, of course, must be deducted from sales revenues in calculating profits. Aggregate profit in the economic system reflects their subtraction.
To the extent that government budget deficits divert funds from the purchase of capital goods and the payment of wages by business firms, their effect is sooner or later to reduce the magnitude of costs of production in the economic system and equivalently to increase the aggregate amount of profit in the economic system. Costs reflect prior outlays of money. To the extent that those outlays are less, the costs will be less. The reduction in costs of production raises profits equivalently, because, as we have just seen, the amount of sales revenues in the economic system is the same either way; it’s the same with or without the budget deficits. In the face of the same aggregate sales revenues, reduced aggregate costs mean equivalently higher profits.
So growing budget deficits are part of the explanation of profits rising relative to wages.
Before concluding, it’s necessary to point out that a rise in profits achieved in this way is not a cause for joy. What is present is an illustration of the dichotomy, identified by Ricardo, that often exists between monetary value and physical wealth. (See his Principles of Political Economy and Taxation, chap. XX.) Aggregate money profit is up, but in large part that is the result simply of the expenditure for capital goods being down. What that signifies is less previously produced wealth being available to serve in the production of future wealth. A part of the output of the economic system that would have gone into the production of future output is instead diverted to the government’s consumption and to the consumption of those to whom the government gives money.
The effect of this cannot fail to be that the productivity of labor in the economic system will be less than it otherwise would have been and that real wages in the future will consequently suffer from production being less than it otherwise would have been and thus from prices being higher than they otherwise would have been. And, ironically, as an integral part of these developments, while total costs of production in the economic system, are lower, unit costs of production will be higher than they otherwise would have been, because of the reduced output that results from the smaller total outlays of business and the consequent reduction in the supply of capital goods available for use in further production.
In sum, the effect of government budget deficits is lower money wages, higher money profits, and lower real wages to the extent that the deficits serve to increase prices and unit costs on top of reducing money wages.
This article is copyright © 2006, by George Reisman. Permission is hereby granted to reproduce and distribute it electronically and in print, other than as part of a book and provided that mention of the author’s web site www.capitalism.net is included. (Email notification is requested.) All other rights reserved. George Reisman is the author of Capitalism: A Treatise on Economics (Ottawa, Illinois: Jameson Books, 1996) and is Pepperdine University Professor Emeritus of Economics. This article is an application of his theory of profit/interest presented in Capitalism, and is based specifically on the discussion on pp. 829-830 of his book.
Comments (26)
Another great article by Dr. Reisman! Thanks! This is a good example of unintended the consequences of bad economics, especially the budget deficit loving Keynesians.
Published: July 25, 2006 12:06 PM
How Government Spending Reduces Wages and Raises Profits.
Taxes, if not collected, would have been used to purchase capital goods and to pay wages. Because taxes come directly from wages, taxes matter more than deficits. All spending, not just deficit spending, matters.
Published: July 25, 2006 5:35 PM
Dr. Reisman,
I looked at the correlation between corporate profits and gross government savings (negative gross government savings are government budget deficits) using quarterly data from the BEA. The correlation coefficient between the two is 0.06 for the 1st quarter of 1947 to the 1st quarter of 2006. If I reduce the sample to include only the 1st quarter of 1980 to the 1st quarter of 2006 then the correlation coefficient is still only 0.16. Since there isn’t very much correlation between the two series empirically, I don’t see how you can conclude that budget deficits have increased profits.
Regards,
Tom
Published: July 26, 2006 11:39 AM
Tom, I don't know if Dr. Reisman will respond to your post, but as a statistician, I think there are a few things to consider in your analysis:
Did you do a lag analysis? The correlation might be stronger if a lag exists between cause and effect.
Also, budget deficits might not be the most important factor in explaining corporate profits. I'm sure you can think of a lot of others. Was the correlation significant at say the 95% level? If so, then you have discovered that deficits do effect profits, but the effect is weaker than other things that effect profits.
Published: July 26, 2006 12:32 PM
Roger M,
I have thought about what I have done and I believe that I need to make an adjustment. So what I have done is to make both variables measured as a percentage of GDP. This has increased the correlation substantially between the two. The correlation between profits and government savings (negative savings are government budget deficits, which include Federal as well as State and Local) is 0.69 from 1947 to 2006. However, if I reduce my sample to cover only 1980 to 2006 then my correlation coefficient drops to 0.26. The other problem is of course that that the sign of the correlation coefficient is wrong. If deficits lead to increased profits were correlated then the sign should be negative. Next, will be to lag profits to see if that changes matters.
Tom
Published: July 26, 2006 1:07 PM
Roger M,
With a one year lag in profits, the correlation between lagged profits and government savings is 0.75 over the entire sample. And if I reduce the sample to 1980 and 2007, the correlation between lagged profits and government savings is 0.57. This means there is a moderate to strong correlation between government budget surpluses and corporate profits, which seems counter to Dr. Reisman's claim.
Tom
Published: July 26, 2006 1:24 PM
Tom,
I must say that your "empirical" study is not in the least convincing. There are other things (variables) that can and do effect profits. In order for your empirical study to even begin to be methodogically correct, you would have to control for ALL of these variables. Since there is no way to put the economy in a lab and control these variables (and apparently, you have not attempted to even *statistically* control these variables), your calculations are therefore void of any meaning or substance. Perhaps I am wrong about this, but I don't think so.
Furthermore, a "moderate to strong correlation between government budget surpluses and corporate profits" makes no economic sense to me. Perhaps you have a "theory" as to why this would be so? I would be very interested in hearing such a theory, because I am always ready and willing to learn, even if it overturns my perception of how the universe (in this case, economics) operates.
Published: July 26, 2006 6:29 PM
Don,
Dr. Reisman's article claims that "the effect of government budget deficits is an increase in the aggregate, i.e., total, economy-wide, amount of business profits." I am simply looking at those varibles to see if there is any relationship between government budget deficits and corporate profits.
To give an example of what I am doing, someone may claim that there are more beach goers when the temperature is higher. The first thing one might want to do in evaluating such a claim is to look at the relationship between the data collected on temperature at the beach and the number of people on the beach. If you had 100 people on the beach when the tempature was 90F and 50 on the beach when the temperature was 70F and 10 on the beach when the temperature was 50F, then one would say there is a strong positive correlation between temperature and beach goers. So, a claim that there is a relationship between temperature and the number of people on the beach would a least seem reasonable.
But what if I said the opposite, that there are less people on the beach when the tempature is higher. Given the data above it would seem like a very questionable statement.
I am merely testing Dr. Reisman's statement that there is a positive relationship between corporate profits and budget deficits. Budget deficits should be positively correlated with corporate profits. That is, when budget deficits are high profits should be high and when there are budget surpluses profits should be lower.
One may also object to the beach example and say there are other things than the temperature that cause people to go to the beach and I have not control for these other variables. In addition, I do not have a theory as to why the temperature and beach goers would be related. But my objective was much simpler: to merely test the relationship between two variables.
Published: July 26, 2006 9:07 PM
Tom,
If you don't mind calculating them, I am curious about the result(s) is you do not measure the variables as a percentage of GDP but still include the one year lag.
I will present the rest of my arguments/concerns later.
Thanks,
Don
Published: July 27, 2006 3:44 AM
Don,
No problem. I have taken the quarterly data series of corporate profits from 1947:01 to 2006:01 and lagged it one year and the government gross savings (negative government gross savings is a budget deficit, while positive government gross savings are budget surplus) and find the correlation coefficient. The correlation coefficient lies between 1 and -1. 1 being strong positive (linear)correlation, -1 being strong negative (linear) correlation and zero being no (linear)correlation.
So the correlation between gross government savings and corporate profits lagged one year (not measured as a percentage of GDP) is 0.14 for 1947 to 2006 and 0.29 if I restrict the sample to 1980 to 2006. This number (0.14) means that there is a weak positive linear correlation between gross government savings (budget surpluses) and one year lagged corporate profits.
As I posted above, if I adjusted the two data series to be a measure of GDP the correlation between the two adjusted data series becomes a moderate to strong positive correlation.
Published: July 27, 2006 10:10 AM
Tom,
Good work! I realize that Libertarians abhore statistical analysis such as you're doing, but as a statistician with a corporation, I think they have a role to play in making more objective decisions.
The reason that the correlation is higher when the variables are a % of GDP is that your including the effect of GDP on both the deficit and profits. In other words, controlling for GDP changes, profits and deficits are positively correlated.
Dr. Reisman often assumes a constant money supply in his analyses, which we won't ever see in the real world, but it's an excellent tool for analysis. For your analysis, you probably also want to include the money supply as a factor in order to control for it and better simulate Dr. Reisman's assumptions.
Finally, you might try shorter lags of 1,2,3 quarters. I often analyze lags and find that the sign of the correlation and the correlation itself can change quite rapidly with different lags. You want to use the lag with the highest correlation. Keep up the good work!
Published: July 27, 2006 12:12 PM
Roger M,
I have tried shorter lags 1,2,3 quarters and longer lags 8, 16 quarters on corporate profits and have correlated that with gross government saving (both adjusted by GDP). The correlation coefficient is similar typically around 0.74 (the high being 0.75 [4 lags] and the low around 0.67 [16 lags]).
The bottom line seems to be that government budget surpluses are positively correlated with lagged corporate profits.
Published: July 27, 2006 2:49 PM
Next, one might ask why would corporate profits and gross government savings (budget surpluses) be positively correlated?
The answer might be that when the economy is good, incomes are up and profits are up; the taxes generated from incomes and profits are also up; and government (Federal, State, and Local) tends to generate budget surpluses in good times. When the economy tanks, incomes are down; corporate profits are down; and the taxes generated from income and profits are also down. In addition during, bad times people are using more government services, such as welfare and unemployment insurance.
So, increased use of government services in bad times combined with the reduction in tax revenue lead to budget deficit at the same time business is bad and corporate profits are down. In good time, there is less use of government services combined with higher tax revenues leading to government surplus during good times, while business is good and corporate profits are up.
Published: July 27, 2006 3:12 PM
I'd be careful about interpretations of the data at this point. You still need to include the money supply as a factor because it may be masking the effect of deficits. Also, are you using deflated data? If not, adding the money supply should kill two birds.
Years ago I worked for a trucking company and tried to determine the effect of pricing on demand. I often came up with models that showed a positive correlation between prices and demand, meaning price increases caused greater demand. Of course that was silly, but that's what the statistics showed. But what was happening was that during the peak seasonal demand for trucking, the summer, we were raising rates and then lowing rates when demand fell in the winter. Seasonality was the main driver, masking the effects of rates and giving me the wrong sign. When I included seasonality in the model, rates achieved the correct, negative, sign.
Published: July 27, 2006 4:22 PM
I agree that may not be a correct intrepretation of what is happening. But looking at the data there doesn't appear to be a negative correlation between corporate profits and gross government savings.
If I were to deflate the data I would use a price deflator, such as cpi, ppi of the gdp deflator, instead of the money supply. However, I don't think deflating both original data series will have much of an impact on the correlation coefficient. I don't think dividing the original data by the money supply would be correct. For one thing, which money supply definition would one divide by. Second, I think a price deflator would be more appropriate in any case.
In your case above, with price and quantity both increasing or decreasing together, that would definitely be a case of a shifting demand curve.
Published: July 27, 2006 4:53 PM
Tom, I think you're going to have to go farther than correlation analysis. You'll at least have to do a linear regression of the deficit on profits, lagged by one year, with GDP and the money supply as separate variables. Since GDP and money are probably correlated, you may want to combine them in some fashion. If you use CPI instead of money, or if you just deflate GDP/Deficit/profits with a price index, you still don't hold money supply constant as Dr. Reisman usually assumes in his analyses, because the money supply may grow at the same rate as the GDP, in which case the CPI would be zero. Also, you might consider using the GO instead of the GDP. The GO is probably a more accurate estimate of economic growth. If you'll tell me which data set you're using, I'll try some of it myself.
Published: July 27, 2006 5:03 PM
Roger M,
I can go futher. Although, actually I have never thought about a relationship between government surpluses/deficits and corporate profits. But if I were to test a model like you have suggested then I will test this:
Profit= constant + GGS + M2 + GDP
where, profit is corporate profits and GGS is gross government savings. M2 and GDP are self-explanatory.
I expect GGS to have a positive sign, as well as M2 and GDP.
Let me run the regression and I will report the results.
Published: July 27, 2006 7:04 PM
CORPPROFIT(-4) = -0.92 + 0.072*GGSAVINGS - 0.13*M2SA + 0.15*GDP
Prob: GGS 0.058, M2SA 0.0000, GDP 0.0000.
R2 is 0.979 and the DW is 0.22.
Because of the serial correlation the results are not reliable. The serial correlation will have to be corrected.
So,if we assume an AR(1) process, then we get:
CORPPROFIT(-4) = -49.621 + 0.175*GGSAVINGS + 0.013*M2SA + 0.086*GDP + [AR(1)=0.941]
Prob: GGS 0.0004, M2SA 0.8220, GDP 0.0049 and AR(1) 0.0000.
R2 is 0.996 and the DW is 2.15. Also, the inverted AR roots are less than 1 at 0.94.
So, now the variables have the expected signs and the serial correlation is eliminated, although M2SA is not statistically significant. This shows that if the GGS goes up by $1 billion, then corporate profits lagged one year will increase by $175 million holding the other variables equal.
Published: July 27, 2006 8:23 PM
Tom,
I wish I had the time to continue adressing my concerns and arguments as I stated I would in my prior post. However, time constraints simply do not allow me to do so.
In closing, I would like to state that the equation Profit= constant + GGS + M2 + GDP, while having mathematical viability, does not seem to be economically viable or even logical, because there is no *direct* relation between GDP, M2, and GGS, which this equation seems to assume. I wish I had more time to present my case, but I do not. (Even Reisman's analysis implies the rlation is indirect).
However, even though I would not be able to take the time to engage you, I would be very interested in hearing an explanation as to why you think it is valid to equate profit as you have do. (If you have the time, of course)
By the way, don't get me wrong, I think you are doing very good work, I just think statistics can sometimes be misleading, especially when the data is poor (has as much noise) as BEA data usually is. (Many economists will admit this, if pushed hard enough).
Published: July 28, 2006 1:45 AM
BTW, where is this data on the BEA website, I can't seem to find it?
Published: July 28, 2006 1:48 AM
Tom, Nice work! Very impressive! The effect on profits is small, but it's there.
Last night I thought of a simpler way to test Dr. Reisman's theory. One of the reasons for the positive correlation is the trends in both profits and the deficit; as the profits have been increasing along with the deficit, at least for the past 2 decades. So a quick check would have been to take the first differences, which eliminates the trend, and test for correlation.
Don writes that the equation "...does not seem to be economically viable or even logical, because there is no *direct* relation between GDP, M2, and GGS..." Actually, we don't want a direct relationship between those variables, because then they would correlate with each other and violate the assumptions of the regression technique. Regression assumes that predictor variables are orthogonal, that is, they aren't related to each other. GDP and M2 were included in order to separate out their effects on corp profits so that we could more clearly see the effect of gov savings.
Keep up the good work!
Published: July 28, 2006 7:34 AM
Don,
The BEA collects a lot of the data, but a great place to access data is from the St. Louis Fed, FRED 2.
http://research.stlouisfed.org/fred2/
The regression I ran above uses quarterly data. The M2SA data, however, is monthly and the series starts begining 1959. The M2SA monthly data was transformed to quarterly data.
As far as the model goes, I would like to include variables that influence corporate profits. Does the GDP influence corporate profits? Or how about the money supply M2SA or Gross Government Savings? Reisman has argued that decreases in Gross Government Savings influences corporate profits by increasing them.
I have not seen any studies done, which I could have built upon, that involves statistically testing the relationship between GGS and corporate profits. So, this is new ground for me. Others, who are interested may wanted to build on what I have done by testing other variables they think should be added that influence corporate profits.
Published: July 28, 2006 7:57 AM
Tom, Now that I think about it, isn't the sign for GGSAVINGS still the opposite of Dr. Reisman's theory? If the coefficient for GGSVINGS is positive, then when GGSAVINGS turns negative (a deficit), then it will reduce corp profits. The sign needs to be negative in order for deficits to increase profits, I think. Maybe we need to regress first differences.
Of course, anyone who has bothered to stay with this thread for long will accuse us of fishing until we get the answer we want. But what we're really trying to do is consider all the possibilities and eliminate all objections that someone might raise. If the sign of GGSAVINGS is still positive in a regression with first differences, then I would conclude that we can't find emperical evidence for the theory.
Published: July 28, 2006 8:55 AM
Roger M,
That is true, the sign for GGS is opposite of Dr. Reisman's theory. To look at an other way, the regression says that if there is a $1 Billion increase in government deficits, then corporate profits will decrease by $175 Million.
I guess I could try to difference the data, which I might do, but it could be that his theory is incorrect. Actually I am skeptical of his reasoning that "government budget deficits...reduce the magnitude of costs of production in the economic system and equivalently to increase the aggregate amount of profit in the economic system." It would seem to me that budget deficits (caused by increased spending) would show that the government is using more real resource than otherwise, and are bidding those resource away from the private sector. Total costs are not necessarily changing only who is in command of those resources has changed. But what if budget deficits are caused by tax cuts. If the private sector has more disposable income they may end up commanding more resources. However, again I don't think this will necessarily change the total cost of those resources.
It might make more sense that when the economy is good, corporate profits are up and deficits (federal, state and local) are down. And when the economy is bad, corporate profits are down and deficits are up. That is, deficits and profits move in opposite directions or, equivilently, GGS is positively related to profits.
Published: July 28, 2006 9:50 AM
D(CORPPROFIT(-4)) = -0.404 + 0.175*D(GROSSGOVTSAVINGS) + 0.0887*D(GDP) + 0.032*D(M2SA)
Prob.: GGS .0002, GDP 0.0168, M2SA 0.6191.
The R2 is 0.15 and the DW is 2.229.
So, here I have first differenced the data and I get similar results as above and the sign remains the same (isn't that Led Zeppelin).
Published: July 28, 2006 10:22 AM
That's interesting. I can't think of any other way to analyze it. You've done a good job with the statistics, though.
Published: July 28, 2006 10:30 AM