U.S Savings Forecast based on Unemployment



College of Business

Western Carolina University




The purpose of this paper is to forecast United States savings rate for the next 5 years starting January of 2000 through December of 2004.  This forecast is based on the civilian unemployment rate of the United States since January of 1959.  United States savings rate is expected to rise in January to 6.96 percent and then gradually drop to around 2 percent in late 2004.  This forecast of United States savings is based on the assumption that interest rates, consumption spending, government spending, and net exports do not influence unemployment or United States savings rate.  (JEL: E24, E21)   


Part 1. Introduction


This paper’s purpose is to forecast United States real personal savings based on the rate of unemployment in the United States.  The explanatory variable will be United States Unemployment Rates.  The approach to forecasting will be the New Classical Macroeconomic Theory and will make use of the civilian unemployment rate (UNRATE), sampled monthly from January of 1959 (1959.01) to December of 1999 (1999.12), for a total of 499 observations. 


This forecast will be useful to all lenders and borrowers alike.  When people k now what the market may show in the future, it may help them make better financial decisions for themselves now.  Thus, when people are making better financial decisions, the economy as a whole benefits.  Savings is necessary for good economic growth, but if the United States citizens continue to trust in the almighty paycheck, they will tend to save less in the future. 


The rest of the paper is organized as follows: Part 2. presents the data used to forecast United States Personal Savings; part 3. explains the theoretical basis for the approach adopted in forecasting the variable U.S. Savings; part 4. presents the forecast of U.S. Savings; part 5. evaluates the importance of the forecast for the economy, and part 6. discusses the conclusions for economic policy.


Part 2.  Data


All variables are taken from the Federal Reserve Bank of St. Louis Federal Reserve Economic Data (FRED).  Civilian unemployment rate (UNRATE) and personal saving rate (PSAVERT) are both real variables.  All variables used are monthly.  The sample period of the data is from January of 1959 to December of 1999.  The observed forecast horizon is 5 years, or 60 months.  None of the data was transformed or proxied in any way since both sets of data were observed monthly. 




Part 3. Economic Theory


Assumptions made in this forecast include all other variables are not influential in the forecast for the next 5 years and inflation will remain the same.  Essentially, this forecast assumes there are no other undue influences on personal savings. 


People tend to save less when unemployed since they have no steady paycheck to depend on.  They sometimes don’t know where their next dollar is coming from, so they will spend what they get on necessary things without saving.  This is because unemployed people usually don’t have the money left over after normal spending to save.  This is why people save less when unemployed.


Assuming unemployment is a function of Real Personal Savings:


St = f (St-5, ut-5)




St = a + bS t-5, + C u t-5 + et


Real personal savings will, according to this forecast, rise sharply in January and then decline slowly until late 2004 if inflation remains the same.  When unemployment stays as low as it is now, real personal savings will decline so people can continue their standard of living at the same level. 


While people are spending more than they can afford (in other words, the savings rate going down), they will have to dip into the debt market, thus hopefully slowing spending.  This will include taking out loans in order to continue spending normally.  With people spending less, inventories will rise pushing up “cost of goods sold” for businesses.  This will cause businesses to lay off employees, thus driving up unemployment up toward its natural rate.  This, in turn, will bring the savings rate back towards a more normal rate. 


Part 4. Estimates of Real Personal Savings Rate


The adjusted R-square of the estimate is 0.223061, indicating that 22.3061% of the difference in real personal savings rate is explained by the difference in the unemployment rate.  In other words, about 22% of savings is based on unemployment, or has been in the past 40 years.


St = 5.57 + .669St-5 - .516ut-5 + et


Part 5. Forecast Implications: Good Constant Rates


Real personal savings are interesting to everyone from large lenders, like bankers, to everyday citizens looking to know what to do with their money for the next several years.  Savings can affect everything from interest to funds available for borrowing.  Since this forecast uses unemployment rate to base a relationship on savings rate, forecast is only feasible if unemployment continues on its current trend.  Also, since unemployment is based on government spending, interest rates, consumption spending, and net exports, these and other things must remain inconsequential or remain at the same rate of change.


According to records of the past forty years, unemployment has gradually declined with short periods of gain.  Because of this, forecast assumes that unemployment will continue to decrease over the next five years.  Savings, according to this forecast, will decline slowly over the next five years as well.  When unemployment continues to decline, industries retain workers and add more jobs, thus lending the belief to people that their jobs are stable and will continue to be stable in the future.  When jobs are stable and people trust their employers, they will spend more, thus reducing their savings rate. 


Part 6.  Policy Conclusions


For the past several years, unemployment rate has declined approximately 0.3 percent to 0.6 percent per year, depending on economy (e.g. 1997.01 unemployment rate was at 5.3, when 1998.01 unemployment rate was 4.7).  Real personal savings has historically declined from 0.8 percent and 0.1 percent annually since 1996.01.  Savings should, according to this forecast, decline at an increasingly slower rate towards the naturally occurring equilibrium stated in the New Classical Macroeconomic Theory.


Assuming this forecast is correct, employees all over the United States will become very secure in their jobs or worry about being unemployed.  As unemployment declines, so will savings until the lack of savings will force people to go into debt or show their spending.  In turn, this will increase unemployment, thus pushing it towards its natural rate.


With employees spending more and more, employers may face the responsibility of their employees trust growing stronger.  With employees trust growing, businesses must be careful not to take advantage of the situations with people demanding work for more money and probably people begging for overtime to take advantage of more people wanting to work more.  With this comes a problem, however, when the savings rate returns to normal, this will also send unemployment reeling back toward normal thus putting many out of their jobs while still in debt from their spending.  Most will have no reserves after the spending if this forecast is determined to become correct.


However, if the government (the Fed) were to step in and try to correct the savings rate declining, they would be too late, since any noticeable effects of the government’s intervention would not be seen for several years.  This would make the Fed ineffective because then they would be trying to fix what had happened already in the past and not working on the current problem since they would be lagged several years.  If, in fact, the Fed stepped in, their intervention might actually be counterproductive to the economy.  The possibility of the Fed trying to restrict the money supply to induce real savings would slingshot the entire process much worse than it would become if simply left alone to the natural equilibrium.




Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED) (12-31-99), http://www.stls.frb.org/fred/ (2-27-00)


Thomas, Lloyd B. Money, Banking, and Financial Markets, New York NY; McGraw-Hill Company, 1996