Canada expenditure data for 65+

[music-and-ski]

I'm trying to figure out if I can retire. I don't know how much money I'll spend. To get an idea of how much I might need retired in Canada, I downloaded the public use micro sample from Statistics Canada's Survey of Household Spending and processed it using the excellent Wizard Statistics program on the Mac. I'm hoping some of you will also find this information useful, and I'd like to hear your reactions.

Here is the distribution of the expenditures of 2 person households in 2017 both people are 65+:

Mean value 67.5k, but a lot of variation.

Here is the distribution of consumption, which is a little less than expenditure, because it doesn't include income taxes, buying investments, giving money to your grandkids, purchasing life insurance, and things like that:

Mean value of 52.7k. (This includes mortgages, so I also filtered on people who own their homes without mortgages, and it's almost identical.)

Here is consumption for 2 person households when both people are 65+ and at least one person is 75+

Mean value 46.7k, supporting the notion that you need less when you're 75+ than when you're 65 to 75.

Here is consumption for 1 person households 65+

Mean value 32K.

I'm curious as to how this compares to people's real experience after 65? Do these seem high or low to you? We talk a lot about income in retirement, but I think we should talk about expenditures and consumption.

A lot of older people I know have plenty of income, because of luck and planning, but they don't spend it. There is some USA research referenced on FIRECalc that tries to disentangle whether lower spending at higher ages is a voluntary choice or simply a product of how much income is available and the culture in which one was raised. Unfortunately, I don't have an indication of wealth in this data, but I do have income. Here is a graph showing the relationship between income and consumption for 1 person 65+:

You can see that expenditure goes up with income, but only at a rate of 41cents of extra consumption for each dollar of extra income. The linear regression model of consumption as a function of income is:

	outcome variable	explanatory variable	coefficient	standard error	t-statistic	p-value	95% confidence interval (lower)	95% confidence interval (upper)
	------------------	----------------------	--------------	----------------	-------------	----------	---------------------------------	---------------------------------
	TC001	HH_TotInc	0.413910	0.030988	13.357239	0.000000	0.353008	0.474813
	TC001	constant	14761.503042	1551.074070	9.516956	0.000000	11713.050267	17809.955816

Here is the relationship for 2 person households all 65+, again only 41 cents of additional consumption for every extra dollar of income:

	outcome variable	explanatory variable	coefficient	standard error	t-statistic	p-value	95% confidence interval (lower)	95% confidence interval (upper)
	------------------	----------------------	--------------	----------------	-------------	----------	---------------------------------	---------------------------------
	TC001	HH_TotInc	0.408017	0.020069	20.330552	0.000000	0.368586	0.447449
	TC001	constant	22592.314670	1766.936849	12.786147	0.000000	19120.694855	26063.934486

 

[music-and-ski]

I could spend all day on this. (Well, I've spent many days already ...)

Below are the coefficients for an interesting model. To calculate the average consumption for Canadian households in your situation with everyone 65+, take $15.4K of base spending (8161 + 7284) plus an additional $7.28k for each additional person beyond yourself. Add another 40.7 cents for each dollar of your income. Then add $7.16k if you have a mortgage, or $1.76k if you are renting. Subtract $3.14k if any of your household members are 75+.

The R^2 on this is only 0.493, meaning there is a lot of variation among households. I know we're all supposed to plan for our own retirement, not the Canadian average for "people in our situation". But at least this gives me some clues.

outcome variable | explanatory variable | coefficient | standard error | t-statistic | p-value | 95% confidence interval (lower) | 95% confidence interval (upper)
TC001 | constant | 8161.332130 | 2407.115034 | 3.390504 | 0.000727 | 3437.366384 | 12885.297875
TC001 | HHSize | 7284.429556 | 1405.333365 | 5.183417 | 0.000000 | 4526.461373 | 10042.397739
TC001 | HH_TotInc | 0.407158 | 0.016759 | 24.294618 | 0.000000 | 0.374268 | 0.440048
TC001 | Tenure = 1 | 7160.351883 | 2105.176105 | 3.401308 | 0.000699 | 3028.941598 | 11291.762168
TC001 | Tenure = 2 | | | | | |
TC001 | Tenure = 3 | 1762.105190 | 1644.105544 | 1.071771 | 0.284099 | -1464.453643 | 4988.664023

TC001 | P75plusYN = 1 | -3140.602269 | 1318.505907 | -2.381940 | 0.017421 | -5728.171477 | -553.033062
TC001 | P75plusYN = 2 | | | | | |

 

[sengsational]

I think the statistics provide a data point, but your current spending should probably guide your decision. What I did was go down the list of past spending and think about if it would have reason to change when retired. Taxes, yes. Housing, no. Travel, yes. Groceries, no, etc. For the changed ones, I did various levels of effort to estimate. Tinkered around a bit with taxes, and took a quick guess at some other things that might change. So that's the spend side. Being that you're obviously analytical, you might consider spending an hour or two on the i-orp site, understanding the details (find the advanced analysis page), and getting various annual spend amounts (as driven by your current asset balance and growth estimates). The default is level spending, but you can also choose other spending patterns, like where, as you get older, you spend less.

 

[Sunset]

What other people spend really has nothing to do with what you will spend in retirement.

The best way I know to be able to figure out if a person can retire is to know their current expenses, by tracking every penny they spend, as estimating is just guessing.

Once you know you spend $X per year for the past 2 or 3 years, then you can figure out if your savings and CPP and Pension will cover it or not.
Of course you also have to factor in rare expenses like: roof, furnace, car, etc.

I'll admit I didn't take my own advice, as I didn't know to do this until after I retired. I now do it all the time for the past 5 years (phone app makes it easy).
What I have learned is:

1. I could have retired 2 years earlier
2. My estimates of my spending were far off for a few things.
3. It's relaxing to actually know how much I have spent as the year progresses, knowing that I'm not accidentally overspending.

 

[Nemo2]

What other people spend really has nothing to do with what you will spend in retirement.

The best way I know to be able to figure out if a person can retire is to know their current expenses, by tracking every penny they spend, as estimating is just guessing.

Once you know you spend $X per year for the past 2 or 3 years, then you can figure out if your savings and CPP and Pension will cover it or not.

+1 We know (pretty much) what we're bringing in, so we spend less. Following low expenditure years, we've 'traditionally' boosted our travel the following year, while still staying below income, and never, ever, touching 'the nut'........and I'm not even a rocket sturgeon....so it's not fishy.

 

[cathy63]

I'm trying to figure out if I can retire. I don't know how much money I'll spend.... We talk a lot about income in retirement, but I think we should talk about expenditures and consumption...

Welcome! I think if you stick around here for a while, you'll find that we talk quite a lot about expenses and less about income than you might think.

When I was starting out, this thread helped me a lot: Some Important Questions to Answer Before Asking - Can I Retire?. Take a look, and you'll see that the very first question is "What are your expenses? No, your real expenses that you have tracked carefully over a period of at least two years, not some rough estimate that you just pulled out of thin air."

You might also get some good advice if you introduce yourself in the "Hi, I am..." forum.

 

[music-and-ski]

What other people spend really has nothing to do with what you will spend in retirement.

The best way I know to be able to figure out if a person can retire is to know their current expenses, by tracking every penny they spend, as estimating is just guessing.
[/LIST]

Well, I'm thinking of cutting back my spending, so I can retire. I see some other people who are retired, and seem to be spending a lot less than we are spending, and they seem happy. I'm not comfortable going up to them and asking them how much they are spending, or what they are spending on. But, I feel their reality might actually be a better vision for my retirement years than some modified hypothetical version of my own current reality.

Put another way, I don't trust our current consumption patterns as an indication of what we want to spend in retirement, and definitely not an indication of what we need to spend in retirement. If I'm going to be hypothetically slashing budgets thinking "we could probably spend 1/3 as much on this" I'd like some confirmation that the average retired household actually does spend that sort of money on that sort of category.

But, yes, I'm tracking our own spending and looking at it critically.

 

[music-and-ski]

the very first question is "What are your expenses? No, your real expenses that you have tracked carefully over a period of at least two years, not some rough estimate that you just pulled out of thin air."

Yes, I'm tracking again. The only long-term data I've got is income less RRSP contributions and taxes for 5-6 years. But, we have detailed tracking from a few years ago for about a year. And, I have expenses for the first half of this year, which is a bit misleading since the ski season was abruptly cancelled and everything else fun and expensive was also cancelled.

You might also get some good advice if you introduce yourself in the "Hi, I am..." forum.

I will!

 

[walkinwood]

Your situation is unique of course, but knowing trends of the population at large has value too. But you have to start with tracking your spending & categorizing it appropriately. The trend of reduced spending as we age - even taking into account additional healthcare spending (which may be moot in Canada) - is worth considering in your plan.

These papers speak about the reduction in spending as you age.

https://www.retirement-insight.com/estimating-true-cost-retirement/
https://www.i-orp.com/help/RealityRetirementPlanning.pdf

Full disclosure: We don't use this in our calculations, but it is good to know that we may be overestimating our financial needs as we age. We don't count SS either - yet another contingency plan. However, we currently live very well on less than we planned to spend, so being conservative in our planning isn't hurting us.
https://www.retirement-insight.com/estimating-true-cost-retirement/

 

[tmm99]

Welcome.

In my case, my overall spending has increased since retirement(by design. Didnt want to skimp as much on retirement.). So if I were in your shoes, I wouldn't try to tighten my budget just so I could retire. Only if I could live on less now, I might consider retirement, but I am glad I could splurge more now without worrying about exceeding my budget.

Either way, knowing your current spending is essential. It really doesn't matter how much everyone is spending. I had kept track of my spending for several years before I retired. And I added some extra discretionary spending on top of my spending to set my budget, to ensure I could weather market downturns, etc.

I see a lot of people here whose WRs are really low, but those are folks with very high NW numbers. My NW/asset is not very much in comparison but my WR is still around 2.8% unless I go into the discretionary spending area with a lot of traveling. I want to keep this pace until my SS starts (another 5 years) My WR is what I feel comfortable at. (I put aside 3.5 minus 2.7% for the rainy day fund for big one time purchases or as future spendings in case the market performs poorly and I don't want to take any money out of the market...)

 

[music-and-ski]

.

I see a lot of people here whose WRs are really low, but those are folks with very high NW numbers. My NW/asset is not very much in comparison but my WR is still around 2.8% unless I go into the discretionary spending area with a lot of traveling

Firecalc is convincing me I can have a much higher withdrawal rate then you have right now now. Even if I keep spending to 6% of the portfolios, it’s still giving me a very good success rate, since the withdrawals needed to sustain that spending go down when the CPP and OAS kick in at age 65.

 

[obsessed]

With a little effort, I would think you can go back and pull together at least a couple years of spending data if most of your spending is via credit card or check. I use about 25 categories but 10 would probably suffice. With most banks, you can download the transactions or at least scrape them off of a PDF statement if you have those. If you just have paper, just categorize the big ones and bunch the rest. A spreadsheet program will be your friend.

 

[Navigator]

Well, I'm thinking of cutting back my spending, so I can retire. I see some other people who are retired, and seem to be spending a lot less than we are spending, and they seem happy. I'm not comfortable going up to them and asking them how much they are spending, or what they are spending on. But, I feel their reality might actually be a better vision for my retirement years than some modified hypothetical version of my own current reality.

Put another way, I don't trust our current consumption patterns as an indication of what we want to spend in retirement, and definitely not an indication of what we need to spend in retirement. If I'm going to be hypothetically slashing budgets thinking "we could probably spend 1/3 as much on this" I'd like some confirmation that the average retired household actually does spend that sort of money on that sort of category.

But, yes, I'm tracking our own spending and looking at it critically.

It’s fun to play around with numbers and your analysis is certainly interesting. However, as others have noted, your current spending is the best indicator of what you will spend in retirement. I saw a study that said expenses go down 10% on average in retirement. However, 16% of people actually increase spending, usually due to a large travel budget. If you have a good handle on your current spending, it is straightforward to see where cuts in spending will be when you stop working. For example, since I didn’t need to commute to work after retiring, what I spent on gasoline decreased. I didn’t know exactly how much it would be, but made a best guess estimate. After doing this for all spending categories, the resulting budget proved to be pretty accurate. If you are planning to drastically cut spending in order to be able to retire, I would suggest being very cautious about this. Retirement is a big life change, but retiring to a different life style is an extreme change.

 

[tmm99]

Firecalc is convincing me I can have a much higher withdrawal rate then you have right now now. Even if I keep spending to 6% of the portfolios, it’s still giving me a very good success rate, since the withdrawals needed to sustain that spending go down when the CPP and OAS kick in at age 65.

Yes, true, mine shows I can have a much higher spending level using Firecalc as well. I also use a retirement planner by fidelity.com however (which is used by many here on this forum), and my numbers aren't as good when I used the "If the market performs significantly below average" setting, my spending is much closer to that number than the Firecalc number. Not sure if you have access to it as you're in Canada, but it's a very good, comprehensive tool in my opinion, and as I said, a lot of people from this forum use it, in conjunction with Firecalc. I also leave some money at the end of life in my calculation, because I would not want to put myself in a situation where I have nothing left at the end of my life except for government-supported pensions. That's just not realistic.

When DJIA dipped down to the 18K's just last month (was it April? or May? Can't remember), I thought that it was probably good to keep my WR the way it is (I could still sleep soundly) until my SS kicks in...

I look at the graph of Firecalc... The one with a bunch of lines? I don't like it when a lot of those lines are bunched up hitting close to zero or getting really low. I want the graph lines densely populated way above the low range. I also check with my NW going down by up to 25%. (I'm around 55/45.) I would benchmark against a drop, like the drop that we had just experienced back in April.

And oh, when I say NW, I actually mean investable assets...

 

[music-and-ski]

With a little effort, I would think you can go back and pull together at least a couple years of spending data if most of your spending is via credit card or check. I use about 25 categories but 10 would probably suffice. With most banks, you can download the transactions or at least scrape them off of a PDF statement if you have those. If you just have paper, just categorize the big ones and bunch the rest. A spreadsheet program will be your friend.

Oh, I like this idea. With only 10 categories I could combine “beer” into the “food” category, and combine “skiing” into the “‘fitness and health” category. In this way I won’t be tempted to cut back on beer and skiing.

 

[street]

When I retired 4 years ago I thought I had done enough tracking my expenses and was confident that they were very accurate. Well, each year since I have been 26% to 30% less in spending then what I thought I would be.
Why I was so far off is that I did pad my expenses thinking of inflation on thing costing more each year into retirement.
I'm glad I over estimated my costs of living wish which gave me more confidence I can make it through the worst times.
It really boils down to what it is going to cost you to live and that is something you will need to come up with a number.
Good luck.

 

[NoiseBoy]

Oh, I like this idea. With only 10 categories I could combine “beer” into the “food” category, and combine “skiing” into the “‘fitness and health” category. In this way I won’t be tempted to cut back on beer and skiing.

Yes you can. I have tracked my spending on a monthly basis since the market implosion in 2008. I quickly determined that I wasn't disciplined enough to itemize down to the line items on each store receipt so beer, groceries, and anything else that I buy on a trip to the supermarket all go into a single "household items" category. I put as much as I can on my credit card and that makes it pretty easy to update my spending spreadsheet.

I am still working, but now I do it 100% from home. One result is that I'm getting a partial preview of my spending levels in retirement. For instance, I used to commute to the office three times a week (~250 miles), but I did not have even one fill up at a gas station on the CC last month. My two year average spending on auto fuel was about $165 per month. In pre-retirement, it's looking like $30 per month and now I have to worry about keeping the battery levels up in both cars! I am still waiting to see the car insurance rate cuts State Farm promised

 

[brett]

We live in Calgary. That type of data was and would be meaningless to us.

It depends very much on your lifestyle, wants and needs, etc. We spend a fair amount on travel so we put our finger in the air and made a guess for an increase post retirement.

We tracked our after tax spending for two-three years or so prior to retirement. Simply added up our current account outflows each month-a five minute exercise. We added a guess for travel and a guess for increased medical...mostly dental work. Increased those guesses by 10 percent. Adjusted for inflation. Then added a bit more to be safe.

Nine years later our after tax spending estimates were a little high compared to actuals. Inflation was lower, investment returns higher. Tax installments were a little higher because of investment returns. We spent more on some things, less on others.

We still track after tax spending. It is the only budgeting we do. Just want to ensure that we are still on track and plan for the years draw downs from investments. Do not really care about the spending categories because of how we shop/buy. We are interested in the annual number. We have spending variations by month depending on when we book/pay for travel plus some recurring annual numbers. We spent less for the last two months so we simply carry the excess forward. We close the number of at the end of the year..positive or negative. Then we decide if our cash spend projection needs to be adjusted up or down. So far only one slight increase because of a change residence.

I realize that it is/was a rather simplistic approach but I cannot help but think that if we spent hours on it, and budgeted by umpteen categories, the bottom line result would have been that much more accurate. I spent my career working with numbers. I did look at several methods but in the end decided to focus on the actual numbers rather that the formula
Perhaps we just lucked out.