Registered Investment Advisor

Scott Burns' Articles -- Recent and Archived
Print Article Email Article

Where does your money go?

070121_photo2.jpgHow does your spending compare with other people?

It isn't difficult to find out. Just visit the Bureau of Labor Statistics Web site. Then prowl around the figures from the Consumer Expenditure Survey (see links below).

I did that and here is what I learned.

When we have more income, we spend it. Households in the middle quintile spend $5,295 a year on food. Households in the top quintile spend twice as much, $10,051. Middle quintile incomes range from $33,381 to $53,358 but average $42,622. Top quintile incomes are at least $85,147 but average $147,737.

The same pattern holds for housing ($13,234 for middle income, $28,006 for top quintile), transportation ($7,437 for middle income, $15,691 for top quintile), apparel ($1,509 for middle income, $3,704 for top quintile), and donations ($1,222 for middle income, $3,869 for top quintile).

The proportions of how we spend our income don't change dramatically as income rises. Middle income households spend 13.5 percent of their total consumption on food. Top quintile households spend 11.1 percent. As a percentage, food away from home remains virtually level at about 5.6 percent.

Middle income households spend 33.8 percent on housing. Top quintile households spend 31 percent. This figure hides major increases in mortgage and real estate tax spending as income rises. Middle incomes spend 11.7 percent on mortgages and 6.6 percent on property taxes. Top quintile households spend 15.2 percent on mortgages and 8.9 percent on property taxes.

The percentage of income we spend on common categories falls slightly as income rises. The most dramatic change is the decrease in health-care spending. It falls from 6.6 percent for middle income households to only 4.4 percent for top quintile households.

Retirees spend less as they age. Two-person households age 55 to 64 have incomes of $68,953 before taxes and spend $54,058. Two-person households age 65 to 74 have incomes of $52,406 before taxes and spend $44,471.

Households age 75 and older show a further drop. They have $37,969 in income before taxes and spend $34,104. They cope with declining income by paying dramatically less in mortgage payments and property taxes. They also chop their new vehicle purchases in half. And they all but eliminate vehicle finance charges.

Retiree spending on health-care nearly doubles, but they also increase their cash donations.

I can't believe some of the numbers. The survey shows that spending on apparel and services is only $1,509 a year for middle income households, rising to only $3,704 for top quintile households. I know we've got the most efficient distribution system in the world, but if that's what we spend, what's supporting all those Wal-Marts, Kmarts and Targets -- not to mention a zillion other clothing retailers?

What we see on television, in movies and in advertising is well beyond the top quintile. Membership in the top quintile -- the $85,147 and up group -- doesn't buy an automatic entry to Luxury Land. Only $807 a year for alcoholic beverages doesn't leave much room for Dom Perignon. And $7,107 for vehicle purchases won't go far at the local Aston Martin dealer. Nor will the $1,568 the top quintile spends on fees and admissions buy many tickets to see Celine Dion in Las Vegas ($444 each for orchestra, $143 for second mezzanine).

Big-time debt isn't universal. One household in four owns its home mortgage-free. An impressive 61 percent of households age 65 and over have no mortgage. Middle income households pay only $302 in vehicle finance charges. Top quintile households pay only $550. Either way, it's not a lot of debt -- at 6 percent $550 translates into $9,167 in debt, a fraction of the average new car cost.   

ON THE WEB Bureau of Labor Statistics, 2005 Consumer Expenditure Survey by quintiles of income

Same, showing percentage of income in each category

Bureau of Labor Statistics, 2005 Consumer Expenditure Survey by age group

Same, showing percentage of income in each category

Bureau of Labor Statistics, spending by age group for two-person household

Bureau of Labor Statistics, Consumer Expenditure Survey general access page

  

Only published comments... Jan 20 2007, 06:00 PM by scottb
Filed under: ,

This article contains the opinions of the author but not necessarily the opinions of AssetBuilder Inc. The opinion of the author is subject to change without notice. All materials presented are compiled from sources believed to be reliable and current, but accuracy cannot be guaranteed. This article is distributed for educational puposes, and it is not to be construed as an offer, solicitation, recommendation, or endorsement of any particular security, product, or service.

Performance data shown represents past performance.  Past performance is no guarantee of future results and current performance may be higher or lower than the performance shown.

AssetBuilder Inc. is an investment advisor registered with the Securities and Exchange Commission.  Consider the investment objectives, risks, and expenses carefully before investing.


Comments

 

ABModerator03 said:

Do you have any information on the "typical spending trends" of a middle-class couple after they retire at the age of 65.

I would assume that the first 5 years are the greatest with the next 5 years declining somewhat and every 5 years thereafter. What is the typical percentage decline in each of the successive 5 years.

Thanks in advance.

Ken

From Scott Burns:

There is good research on this based on regular household surveys. When the data is sliced by age it shows that expenditures tend to decline significantly by age 75.   Here are links to three columns on the subject, each with additional links:

Consumption, It's Just So 1950

In the future, you'll need less money--- the sequel

In the future, you'll need less money (DMN)

Many people can't believe these findings or want to believe that people are spending less because they have less to spend. In fact, most people were experiencing increases in net worth even as they were spending less.
January 15, 2007 10:44 AM
 

ABModerator03 said:

This is interesting, but like you I can't believe some of the numbers. I find it hard to understand how in every quintile "Income before taxes" and "Income after taxes" are almost the same. For me (top quintile), income after taxes is 61% of income before taxes, and that ignores sales tax. For the average person in that quintile, according to the data, income after taxes is 95% of income before taxes. How do so many people manage to be in the top income quintile and yet pay so little tax? I want to learn their secret.

From Scott Burns:

So do I. As I said, I don't believe all of these numbers.

If you go to the links provided, however, you will find part of the reason. The tax burden referenced is the federal income tax. Income spent on "pension and Social Security" is not included in taxes but is included in the total consumption figure. It's also broken out.

You can be in the top quintile for income and still have an average income tax rate that is much, much lower than your marginal tax rate due to the structure of our taxes. A couple filing jointly, for instance, can have an income well over $80,000 before they move from the 15 percent tax bracket to the 25 percent tax bracket due to the standard deduction, personal exemptions, and tax-deferred plans like 401(k) plans.
January 20, 2007 8:24 PM
 

ABModerator03 said:

The problem with averages, as I am sure you know, is that you can drown in an average of six inches of water. Here is how I distort the numbers, although it would not even register on the blip screen.

I am 67, single, and live in an Uptown apartment. The numbers would show that I am in the minority of people in my age bracket who do not own their housing. However, what the numbers don't show is that I sold my house in 2001 and invested the proceeds in CDs, bonds, and stocks. The interest and dividends that I earn from these investments cover the rent, the utilities, and my transportation operating costs. Moreover, I don't incur the costs of cutting the grass or painting the house or fixing it.

I enjoy your columns. They are very informative; I turn to them as soon as I seen the latest offering.

Paul

From Scott Burns:

Good move! I believe you are the front line of a new wave. When you do the math, owning doesn't look very beneficial for older people.

Mr. Burns,

Thanks for your note.   I appreciate it.

I am considering a move to San Angelo or Midland.   Now that I am retired, I don't have any reason to stay in the Metroplex, and I have a number of reasons to head for the exits, i.e. pollution, traffic congestion, crime, etc.

I am looking at retirement communities in San Angelo - Rio Concho West and in Midland - The Village at Manor Park.   I considered buying a patio house in either community, but I ruled it out.   I don't want any more houses and, as you surmised, the numbers don't support owning one at this stage of the game.   I am also considering a rental similar to the one that I now occupy.   Or even stepping up to a two bedroom apartment, which I can easily afford, but I don't need two bedrooms.

At the end of ten years, assuming my assumptions and estimates are correct or nearly correct, I would be approximately $70,000 ahead by renting as opposed to buying a house.   I would be ahead, although by a lesser amount, with the retirement community option.   Of course, as you know, I had to make a lot of estimates, which may or may not pan out.   But using a range of scenarios, with different estimates, I still come out ahead by renting.   And I am talking about renting up market.

I owned a house in North Dallas for 24 years.   I sold it for approximately 4.5 times what I paid for it.   It sounds like a good deal.   However, after considering inflation, mortgage interest, property taxes, maintenance, insurance, real estate agency fees, etc., I just about broke even.   Moreover, I put an average of six to eight hours a week of sweat equity into the house.   If I had paid myself what I was being paid at TXU, I lost money.   Big time!   Had my wife and I stayed in a nice two bedroom apartment and invested the difference in an S&P 500 Equity Fund, I would have been at least $50,000 ahead of where I came out after selling the house.

I have spent the better part of the year since I retired researching retirement living options.   I put my findings in a memo that I sent to Bob Moos, since he writes a lot of columns for seniors.   I would be happy to share a copy of it with you if you like.

Cheers,

From Scott Burns:

  I've come to the same conclusion indirectly. My wife and I sold our Dallas townhouse two years ago to spend full time at our house in Santa Fe. I have so much going on in Dallas, however, that we decided to rent an apartment rather than stay in hotels. Last week my wife leased a 2 bedroom, two bath apartment in North Dallas with a wonderful open view of the Bent Tree golf course. We'll use the 2nd bedroom as a sitting room/study. It's 1130 square feet, is well maintained, has great light, high ceilings, and an attached one car garage. Price $1,270 a month plus bills.

My former assistant at the DMN just sold her townhouse of the same square footage for about $175,000 so she could live full time in Big Sky, MT. The monthly operating costs on that property were about $1,100.

So my wife and I are getting the same level of amenity (actually slightly better if you value the view that other facilities in the complex) for $170 a month additional with zero capital invested or at risk. That compares nicely with either investing $175,000 or borrowing it.
January 21, 2007 8:34 AM
 

ABModerator03 said:

I have always been curious re: the income figures used in articles ! Are they normally : Gross Income,AGI,Taxable Income or Net Income after Taxes ? Your column of 1/21/07 mentions Income (?) in quintiles in one section and then income after taxes in another section !

Thanks,

Martin

From Scott Burns:

There is no standard measure for income measurement. So different studies will have different figures. In some studies, for instance, researchers try to measure--- and include--- "in kind" income in goods or services that people receive. This is particularly important at low income levels where the household could receive benefits in Medicaid, subsidized housing, the Earned Income Tax Credit, etc. It's also important when considering the income of older people because they are likely to have "imputed income" from assets they own such as their home.

The same difficulty in measuring also applies to assets. The Survey of Consumer Finances, which the Federal Reserve does every three years, measures all kinds of assets but it does not include corporate or public pensions or Social Security. All of these are income rights, not personal assets, but they make an enormous difference to people. One of the questions many readers puzzle over is how they should treat their pension income when they measure their net worth and how it should influence their asset allocation.
January 21, 2007 1:18 PM
 

ABModerator03 said:

Scott------

Please help two people settle a disagreement regarding your article in the Globe today-

"A peek at how and when we open our wallets"-----On your last point, you comment on vehicle finance charges for middle-income $302 vs. top quintile paying $550. Would you clarify your thought process on this matter? Does the $550 translate into $9,167 in vehicle cost---include finance charges for payments on the car, insurance cost, fuel, registration, etc. or, just how do you come about this figure? I claim your are dealing with yearly figures, on finance charges, my partner believes it includes all of the afore-mentioned.

I disagree with him, as you can tell.

From Scott Burns:

The figures are averages across a broad sample. That has a lot of implications for what you can pull out of the figures. Just as a fair number of households, particularly older households, own their houses free and clear, many households own their cars free and clear. Those households reduce the apparent interest burden other households appear to pay.

My wife and I, for instance, own two vehicles free and clear, a 2003 Prius and a 2002 Suburban (which makes us automotive bi-polar). Another couple could have the save inventory, worth about $30,000, and be on the hook for as much as $2,400 in interest. Average the two couples and you get $1,200.

Which applies to "real life?" Neither.
January 21, 2007 3:34 PM
 

ABModerator03 said:

I couldn't believe that the highest 20% income category. At $140,000 plus yearly income, the average tax people paid is less than $10,000. This can't be true - I know it is probably true - I just wonder how most other people can achieve such feat!

George

From Scott Burns:

I also thought their income tax bill was low. A household can, roughly, have an income of about $105,000-115,000 and have a federal tax bill of $10,000, assuming they itemize deductions, save 5 to 6 percent in a 401(k), file a joint return and have additional personal exemptions.
January 21, 2007 7:17 PM
 

ABModerator03 said:

Click here: Bureau of Labor Statistics Data

Scott,

The main thing we learn from this survey is that taxpayers are wasting a lot of money on the gathering and the deciphering of this data. It is clearly inaccurate and unrealistic. As you said in your article, I cannot believe some of the numbers. This entire survey seems to be totally flawed.

How did you compute the proportions spent on food to be 13.5% for middle income versus 11.1% for top quintile?

Matt

From Scott Burns:

I wouldn't be so harsh. The top quintile may spend a lower percentage on food but their DOLLAR spending doubles. Similarly, the amount of income lost to taxes appears far lower than what most people lose to taxes but when you look at the tables closely you find that employment taxes are counted as consumption under pensions and Social Security.

The statistics may have limits, but so does everything else. I wouldn't throw the baby out with the bathwater.
January 22, 2007 1:01 PM
 

ABModerator03 said:

This is a good start to what I am interested in. I have carefully tracked my spending for the six years I have been retired. I am in the upper quin-tile. I am looking for a financial model or spreadsheet that identifies expenditures by category and percent of income. Such as dining out, groceries, utilities, vacations, gifts given, etc. Any tips on where I may be able to find this information? My annual income is about $175,000 and spend about $30,000 on groceries (includes all stuff bought at a grocery store such as paper products, cleaning supplies as well as food), and dining out.
January 22, 2007 4:33 PM
 

ABModerator03 said:

Please disregard my previous post. I reread the article and reviewed the references. Much of the information I am seeking is in the tables from the BLS.
January 22, 2007 4:41 PM
 

ABModerator03 said:

Dear Mr. Burns,

Where in the world did our government get these figures? They are average for who-where? I have had the pleasure and difficulty of being in the top,lower and middle quintiles over the past ten years trying to educate,feed and clothe myself and two children. I track my expenses pretty regularly and these figures do not add up!

Carla

From Scott Burns:

You are not alone in that response. The figures, however, are averages. The different income bands, for instance, mix households of different size, age, and geographic location.

Readers in Boston and California, for instance, regularly write to tell me that the rents readers from Ohio, Tennessee, and Texas talk about are unbelievably low.
January 22, 2007 6:39 PM
 

ABModerator03 said:

Clarification from Scott:

Celine Dion fans please note. The $444 dollar price for orchestra seats quoted in last Sunday's January 21st column was a reseller price, found by Googling "Celine Dion tickets" on the web. A representative for A New Day assured me that the direct purchase price from Ticketmaster, www.ticketmaster.com, ranges from $87.50 for second mezzanine to $225 for front orchestra.
January 23, 2007 5:49 PM
 

ABModerator03 said:

I just began to look at the data that was linked to your article and noticed something that looked very unusual. For all brackets but especially the top 2 brackets the tax rates seemed unbelievably low (about 4-6%). Do you think this is possible? Just FICA and Medicare are higher than this. This doesn't include State and federal taxes. Any ideas about why this apparent error?

Thanks for your columns.

Rich

From Scott Burns:

Part of the low figure (but NOT all!) can be explained by the fact that the federal income tax is subtracted off the top but the employment tax is in the consumption figure--- its under pensions and retirement saving toward the bottom of the list.

With incomes under, say, $100,000 the average income tax tends to be much lower than the marginal tax rate. It is possible, for instance, to pay taxes at a marginal rate of 25 percent but have an average tax rate well under 15 percent. Again, this still doesn't explain the absolutely low numbers in the survey data.
January 25, 2007 8:09 AM
 

ABModerator03 said:

Scott,

The basic math is very confusing and seems flawed. Take the food spending that you cited:

"When we have more income, we spend it. Households in the middle quintile spend $5,295 a year on food. Households in the top quintile spend twice as much, $10,051. Middle quintile incomes range from $33,381 to $53,358 but average $42,622. Top quintile incomes are at least $85,147 but average $147,737."

"The proportions of how we spend our income don't change dramatically as income rises. Middle income households spend 13.5 percent of their total consumption on food. Top quintile households spend 11.1 percent."

I can't figure how the top quintile's food percentage is calculated to be 11.1% using the figures that were cited. It seems that $10,051 is no where near 11.1% of $147,737 average income.

Very confusing.

Matt

From Scott Burns:

You'll understand it if you go through the itemized expenditure lists in the original documents.

http://www.bls.gov/cex/2005/share/quintile.pdf   This one, for instance, shows the percentage breakdowns. It also shows that households with income averaging $147,000 spend only $90,000 on consumption covered in the survey. That's how they can spend 11.1 percent on food.  
January 25, 2007 7:24 PM
 

ABModerator03 said:

SCOTT,

In last Sunday's story about data from the 2005 Consumer Expenditure (CE) Survey, you questioned the spending on apparel and services. Your concern is right‑on. We estimate the spending per household on apparel and services for 2004, for income ranges $30,000-$59,999, to be $2,227 per year. That is not quite the same income range you reported, which was for the middle quintile ($33,381-$53,357), but it's close.

You were citing 2005 data; ours are 2004. Using the CE published data exclusively (accessible on BLS' website), for incomes $30,000-$69,999 the mean spending per household ("consumer unit," in CE terms) on apparel and services is $1,666 in 2005. We use the CE microdata to build the data by customized income ranges. The 2005 microdata CD is expected to be released at the end of January. Our estimate is based on market sales in the US, thus somewhat comparable to PCE (Personal Consumption Expenditure) data estimated by BEA. In particular, it excludes purchases by US households made outside the US and includes purchases by foreigners made inside the US.

The problem with the CE data is underreporting by respondents and sampling errors, as documented in studies published by BLS. The most recent study was published in Monthly Labor Review in September 2006 (pp. 20-46). In this report, BLS researchers compared total spending on categories such as apparel and services from the CE data to PCE. For apparel and services, the totals estimated from CE data range from 49% (men's apparel) to 71% (shoes) of the PCE estimates (p. 37). There are problems with the accuracy of PCE data as well, mainly the delay in incorporating the 2002 Census of Retail Trade in the PCE estimates. For these reasons and others, we use a customized econometric model to estimate market size by merchandise line. The developers of the model are the experts at Moody's Economy.com, but the basic ideas were developed at JCPenney years earlier.

Despite the data problems, I appreciate you bringing CE Survey data to the attention of the business community. It is an important input into our model and the key input in comparing spending on merchandise we sell to other spending, such as for gasoline and health care. We also use for monitoring expenditures over time, since the underreporting bias is unlikely to vary much over time.

Both my boss and I really liked The Coming Generational Storm. Are you going to be issuing a new addition? There were a few issues in demographics I think need fine tuning because the ethnic, health, and income mix will be quite different in 30 years than in the immediate future. Otherwise, I see nothing in the near term to change the outlook or your prescription for remedy. Thanks for brining this seminal work to the public.

BRUCE

From Scott Burns:

Thanks for your note and the information. I doubt there is a single government statistic that doesn't have important caveats about its use or accuracy, witness the never ending concern about the actual rate of inflation or the less widely discussed differences in measuring employment and unemployment. Like real life, most everything is messy when you put it under a microscope.

We can, however, recognize trends and it appears that clothing is declining as an object of our spending. Why is another question. The nice story is that incomes are rising enough that clothing no longer commands such a large portion. A visit to any mall or multiplex movie theater, however, tells another story--- more and more people dress as though they were going to a car wash or laudromat.
January 26, 2007 9:03 AM
 

ABModerator03 said:

Dear Scott,

I was lucky to hide your column last week before my wife got to see it.   If she had seen it, she would want us to increase spending in two or three categories to keep up with the average in our income quintile.

Personally, I enjoy spending like the next lower quintile and if we never replaced a piece of furniture, a carpet, or curtains, I would be very happy.

Of course, buying new tools & gadgets, computer software & stuff, or piece of office equipment is different. So is making another trip to South America or to Europe.

Except for last week's information that would be dangerous for my wife to read, I enjoy your columns.

Hugh

From Scott Burns:

I have the same problem in my house, too.
January 28, 2007 4:35 PM
 

ABModerator03 said:

Scott,

I am retired with a pension, substantial 401k assets and no debt. I am using a planning tool (Quicken) to look into the future to determine how much I can spend without dying at age 105 broke. In looking toward the future it becomes clear that you must guess at inflation, tax rates, rate of return on investments and age at death. None of those are easy but a good tool lets you try a number of combinations quickly.

I am seeking more information on spending by age group. Do people in there 90's spend more or less than people in there 70s.

Members of our family have lived to over a 100. It seem they travel less but spend more on heath, services and the like. Is there a data source for general spending by age group for those over 60. Your column of the past month showed the government data for spending by all ages groups but over 65 was lumped into a single group.

Thanks

Jim

From Scott Burns:

Here's a link to a column that discusses retirement spending, http://assetbuilder.com/?p=116. When you scroll to the bottom of the column you'll find several more links with have survey data tables. The surveys show that spending starts to decline in the 50s, with a fairly dramatic decrease by age 75. After age 85 (the "old-old", as opposed to the "young-old" who are 65-75) the data is too thin for researchers to make any conclusive observations. In addition, medical expenses and health vary wildly after that age.

By the way, you can also experiment with long term portfolio survival with an online calculator at http://fire-calc.com.
February 4, 2007 3:38 AM

About scottb

Scott Burns has covered the changing world of personal finance and investments for nearly 40 years. Today, he ranks as one of the five most widely read personal finance writers in the country. Scott began his career as a newspaper columnist at the Boston Herald in 1977 where he was also the financial editor. Nationally syndicated in 1981 and now distributed by Universal Press, the column appears in newspapers from Boston to Seattle. In 1985 he joined the staff of the Dallas Morning News where his column quickly became one of the most widely read features in the paper. He left the Dallas Morning News in 2006 to become one of the founders of AssetBuilder and its Chief Investment Strategist. Burns is a graduate of Massachusetts Institute of Technology (1962). He has written four books, including "The Coming Generational Storm" (MIT Press, 2004) coauthored with economist Laurence J. Kotlikoff. His fourth book, also coauthored with Kotlikoff, was published in 2008 by Simon & Schuster. The paperback edition will be available in January, 2010.  "Spend Til' the End" uses consumption smoothing to demonstrate the errors of conventional financial planning. His business experience includes working as a staffer for a major consulting company and service as a director and audit chairman of a NASDAQ listed manufacturing company. He and his wife now live in Dripping Springs, a "hill country" town about 25 miles outside of Austin.


© 2007-2012 AssetBuilder Inc - DFA Funds Registered Investment Advisor. All Rights Reserved.

Contact Us

Open Monday-Friday
9 a.m. - 5 p.m. (CST)

ph. 972.535.4040
fx. 214.556.3848
Email Us

1255 W. 15th Street Suite 240 Plano, Texas 75075