MaxedOutMama
Wednesday, January 26, 2005
Do It Yourself
This is going to be another unbelievably boring post. I can't stand reading more articles such as this one in the Washington Post about Social Security any more. It is literally painful to me to watch the media adopting these fake numbers. Only another person required to do very accurate computations for a living - computations that will cost my customers huge sums of money if they are wrong - can understand my feelings.
These people obviously don't understand numbers at all. I suppose the authors of these types of articles could be conniving duplicitious scum, but I prefer the more generous-spirited explanation of concerted incompetence. Regardless, they should know better, and many people are very understandably and reasonably deceived by the latest incarnation of voodoo economics. The authors of the above-linked piece of misrepresentation write:
When the Social Security Administration begins to run deficits from current income and trots over to the Treasury to redeem those bills, the taxpayers of that year will have to pay for them. If the bills did not exist, the Social Security Administration would have to get money to pay retirees from the Treasury and general funds. Any income paid on those bills now is paid from the general funds, and goes right back into general funds to pay current expenses (because Social Security is running a surplus) in the type of transaction that should be familiar to anyone who has read much about Enron and similar corporate accounting escapades. We have spent all the Social Security surpluses and will continue to spend them unless we amend our wicked ways.
After rolling around on my mental floor in horror infused with despair for a few days, I decided to cope by showing you how very ordinary people can do a set of simple calculations to figure about what's really happening and going to happen with Social Security for themselves. All the information used is publicly available. Spend half an hour with me, and then you will be far more qualified than a lot of column writers to discuss the issue of whether Social Security is in a crisis or nearing one.
It's really rather simple once you ignore the funny money. The most important variables in this system are workers per retiree, average monthly worker wage, and average monthly retiree benefit. The further you go out in years, the more uncertain each of these variables becomes, so we will concentrate on the years up to 2025. In order to make matters simpler, we will render everything in today's dollars, assuming that both wages and benefits don't change with time. They do, but at the current time they are pretty closely tied together when calculating benefits, so we can ignore that change quite safely for our purposes.
So come with me, children, and calculate with Mama. Leave the pundits to enjoy their sojourn at Walt Disney; the voters can catch up with them later, and we will.
We will get our figures for projected workers per retiree from Social Security's own latest projections, which are as good as they can possibly make them. They offer three sets of figures, and we will use the intermediate ones. We don't want to alarm ourselves unnecessarily.
Coming up with average worker wage is only slightly more challenging. The government does surveys, and we will use them. We will check our figures with two different methods, because we want to make sure our numbers are reasonable. From this survey (the latest available, slowly loads in pdf) we find that the average hourly wage across private, state and local governments is $17.75. Ominously, we note that government wages are higher than private wages. This will become important later. $17.75 * 39 hours a week * 52 weeks a year gives us about $35,997, or about $36,000 annually. From this survey (much faster txt), we find (it's right at the top) that the average wage nationally is $36,000.
These numbers are slightly old, but close enough for our purposes. These are not mean wages, but average wages, meaning that if you multiplied the number of wage-earners by the average wage you would end up with total wages for the entire nation. So in reality, somewhere around 4 to 6 percent of these wages would not be subject to the payroll tax. Certain state employers (like some university systems) are excluded from their half of the SS tax, and some workers exceed the wages subject to Social Security tax. This will offset the increase in wages we may be missing by not having figures for 2004 or 2005.
$36,000 a year breaks down to $3,000 per month. Social Security taxes on subject wages are 12.40, half paid by employer, and half by the employee. So the monthly contribution per worker is 12.4 % of $3000 which equals $372 dollars. We'll round that up a bit to $380 for convenience and because we want to be sure not to exaggerate the problem.
Now we come to the next item, which poses slightly more of a challenge: how do we figure out average monthly benefit per retiree? Here we will revert again to two useful pages here and here provided by the Social Security Administration. On the second page we find two examples of how a current retiree's benefits would be calculated. The lower number (for a low wage earner retiring at 62) is $1273.74, and the higher number (for a more typical wage earner retiring at 65) is $1508.80. An example for a really high wage earner is not given, but they represent a small percentage of the workforce anyway. So we will interpret between these two numbers and derive a thumbnail figure of $1,400 monthly benefit per current retiree. Certainly 20 years in the future most retirees receiving benefits will have retired in 2004 or later, so when projecting forward in today's dollars this is probably a pretty accurate figure. If anything, it should be a bit low.
Still, we want to check ourselves, so we will try to calculate this in a different way. You can see a historical list here of average monthly benefits up to 2002. The last number given is $895 over all retirees. Extrapolating from the past five years and giving more weight to the lesser increases in more recent years, we come up with an average increase per year of about $25.00. $25 * 10 years equals $250 per decade, and two decades yields an increase of $500, which added to $895 in 2002 gives $1,395 in 2022. Not a bad fit, so we can feel pretty comfortable with our $1,400 average monthly benefit figure. Remember, we are not being paid for this - all we are trying to do is figure out whether we need to worry about Social Security deficits or not.
Okay, now it's all simple.
Currently there are about 3.3 people paying for each retiree, and on average each month, each one is paying $380 into SS. 3.3 * $380 = $1254. Adding $50 (updating 2002 figure to 2004 app. monthly benefit) to $895 means we need to pay about $945 per retiree, so we're running a nice surplus. No crisis now.
But if current workers were paying a higher benefit to each retiree, we would have a problem. $1,400 / 3.3 = $424.24 per worker - call it $425. That doesn't looks so good.
When the ratio drops to 2.5, it will be $560 per worker.
When the ratio drops to 2.0, it will be $700 per worker.
How much would the average monthly salary in today's dollars have to be in order for 2 workers to pay for one retiree's average earnings 20 years in the future, expressed in today's dollars? .124 * X = $700; $700 / .124 = $5,645, expressed in today's dollars, which is equivalent to yearly earnings of $67,740. The actual number in the future would be much higher of course.
It's not going to happen, honestly, no matter how much productivity grows. And remember, given the current method of calculating retiree benefits, when average wages go up so do retiree benefits, so actually the benefit per retiree would be much higher. This does not look promising; it appears there is no way under today's system to catch up with ourselves once the ratio of workers per retiree drops much below 2.7. If this seems confusing, go back to this page and look at how they are adjusting past wages in light of current wages.
But let's try to calculate this another way, to quantify the effects of the expected problem.
In 2010 we expect to have about 3.2 workers paying for each retiree. Moving our average benefit up 8 years (8 * $25 = $200, $200 + $895), we get an estimated average retiree monthly benefit of $1,095. 3.2 workers each paying $380.00 per month gives us $1,216. We are still running a decent surplus in 2010.
In 2015 we expect to have about 2.9 workers per retiree, which gives us 2.9 * $380 = $1,102. To get our expected benefit per retiree, we add another 5 years of increases ($125) to $1,095, equaling $1,220 rounded. It looks like we will be crossing over into deficit territory. We could be a few years off either way, but not by too much.
In 2020 we expect to have about 2.6 workers per retiree, which gives us 2.6 * $380 = $988. Adding another 5 years of increases, we can guess at an average monthly retiree benefit of $1,345. We have a deficit per retiree of at least $350 per month. What the heck - let's be conservative and say it's $300 per month per retiree.
How many of these retirees are there going to be? 67,987,000 of them, according to the intermediate table. But over 10,000,000 of that number receive disability only and on average receive less monthly. So we'll round down to 60,000,000 retirees and multiply by our $300 deficit per retiree to get a monthly deficit of $18,000,000,000. To get the annual deficit multiply by 12, which gives us $216,000,000,000. That is conservatively $216 billion that social security will cost us above what workers pay into the system in 2020. Pulling off our shoes and socks, we count anxiously on our toes and the fingers of one hand, and realize that 2020 is only fifteen years from now.
This is getting ugly fast. There is no trust fund - if T-bills are redeemed either those 2.6 hardworking wage-earners per retiree, corporations, or the retiree is going to have to pay for those T-Bills in extra income taxes. Probably all will be pay higher taxes. The effect on the budget gets worse when you realize that in 2003 the Social Security surplus reduced the deficit by over $70 billion (ignore those stupid interest earnings). The net effect on the budget will be, very conservatively stated, an additional $280 billion in deficit spending compared to current figures.
Let's move forward to 2025, because we said we were going to go out 20 years. We will use the figure of $1,400 monthly benefit per worker, which is definitely too low. But again, we are lowballing. We expect to have 2.3 workers per retiree. 2.3 * $380 = $874. Not good. We now have a deficit per retiree of over $525 per month, but we'll be conservative and use $450.00.
How many retirees do we expect to have in 2025? The total expected is approximately 76,000,000, but about 11,000,000 of those are projected to receive disability only. So we'll take a conservative figure of 68,000,000 retirees and multiply by our $450 deficit per retiree to get a total projected monthly deficit of about 65,000,000 of those will receive Let's use just the number 30,600,000,000. To get the annual deficit we multiply by 12, which gives us $367,200,000,000. That is about $367 billion extra that those 2.3 workers, corporations and retirees will have to come up with in 2025. It could easily be substantially more. The net effect on the budget versus current figures is approximately $440 billion in additional annual deficit, which is about equal to the ENTIRE deficit now. Remember, we are doing all of this in today's dollars.
When you start to consider the effects of Medicare and Medicaid, which are also highly dependent on the same worker/retiree ratios, you begin to see what a massive hole we will be in 20 years from now. Now it's up to you whether you want to consider this sort of thing a crisis or not, but I question the sanity of anyone who claims we don't have to worry about this until 2042. There is no trust fund. We have been using SS taxes to pay for general expenditures that we should have been funding from current income. The longer we keep doing it, the worse the situation gets.
What are our options?
Things that might help to reduce the SS shortfall would be more immigration, higher birth rates, cutting back on our overall federal spending sharply, or changing the way Social Security benefits will be calculated in order to give a lower average benefit. This is one reason why I favor more legal immigration. Higher birth rates can only be achieved by cutting real taxes on the younger and poorer of our population. As I explained in this prior post, we are taxing lower wage earners to death and most of the taxation on younger or less affluent workers comes from Social Security.
However, if we cut taxes for these people we will increase the overall budget deficit now. But we certainly should not be raising taxes on the lower 1/2 of wage earners, and we should be embarrassed at the amount of taxes many of these people are paying today. As I pointed out in this post, raising payroll tax rates on higher wage earners is getting close to counterproductive already. We are sending jobs overseas to cheaper locations at a very high rate.
We are running out of options to address the situation with each passing year. What we are going to do will end up being cutting benefits, raising taxes, and cutting benefits very sharply for the top echelon of retirees or those with significant other assets. There is no other way out, but certainly we can't spring this on people at the last moment. We need to allow people time to prepare, and we need to figure out what we are going to do for the lower echelon of wage earners who are being taxed at historic highs and have little chance of saving with what they have left.
As for raising the full retirement age, that has already been done. People can still take early retirement at age 62, but if you increase that age you will force a lot of older people to go on disability. That will end up costing more, because those persons will get not only disability but Medicare.
Note: A careful reader will note that I have slightly earlier negative dates, even lowballing, than the Social Security projection to which I linked. That is partly because I ignore the funny money revenue from the T-Bills in the trust fund and so on. The point is, the tax-payers pay that money into SS just as they pay their payroll taxes, so it is a wash and need not be considered. I have also ignored the "revenue" from these T-Bills in my calculation of how SS offsets the budget deficit currently. I believe my figures are quite accurate - in fact I believe they are underestimates by about 15% by 2023. But that's a much longer set of calculations.
Factors which the social security projections don't address is that high unemployment rates among older people and lower medical insurance coverage rates (largely related to high medical insurance costs) are forcing far more retirees to take early retirement at age 62 (they use their SS check to pay for medical insurance or offset medical costs). This does lower the average benefit, but it increases retirees per worker, so the effect is pretty much another wash. In 20 years, almost all of the retirees in the program will have retired under the current formula, so the average benefit in today's dollars per retiree is a pretty good figure.
Update: The CBO's latest release is out, and Donald Luskin read it and pulled out the relevant passage. The Congressional Budget Office has SS cash flow going negative a couple of years later than I do. But the facts are clear - we will be in the hole before 2020.
The Congressional Budget Office writes:
You can't honestly make that claim without looking at the budget as a whole, and the budget as a whole has a lot of sinkholes in it already. Last year, although the precise figures haven't been released yet, Medicare and Social Security taken together ran a deficit and some contributions from general revenue were required in order to make payments to beneficiaries.
These people obviously don't understand numbers at all. I suppose the authors of these types of articles could be conniving duplicitious scum, but I prefer the more generous-spirited explanation of concerted incompetence. Regardless, they should know better, and many people are very understandably and reasonably deceived by the latest incarnation of voodoo economics. The authors of the above-linked piece of misrepresentation write:
The latest Social Security trustees' report, whose numbers even the White House uses, predicts that the Social Security program can pay all promised benefits for the next 38 years -- with no changes at all. The June 2004 estimate from the nonpartisan Congressional Budget Office projects that Social Security can pay all promised benefits without changes for even longer, until 2052. That's nearly half a century.Half a century and quite a fantasy! The numbers above are generated by pretending (as the Social Security Administration must do under law) that the Trust Fund is a reality instead of a pile of paper. Oops. If we were running and could expect to run large surpluses into the foreseeable future, this would be a reasonable assumption. But we haven't been running surpluses (the few surpluses during Clinton's era were cash-flow surplus generated by Social Security surpluses, but real deficits remained), and we most certainly won't be running surpluses. In fact, we wouldn't be running surpluses even if there were no problem with Social Security. The Medicare and Medicaid programs alone assure large deficits far, far into the foreseeable future no matter what rosy projections we make for growth in the economy, etc. In light of this indisputable fact, the T-Bills held by the Social Security Administration are as valuable as Confederate dollars.
When the Social Security Administration begins to run deficits from current income and trots over to the Treasury to redeem those bills, the taxpayers of that year will have to pay for them. If the bills did not exist, the Social Security Administration would have to get money to pay retirees from the Treasury and general funds. Any income paid on those bills now is paid from the general funds, and goes right back into general funds to pay current expenses (because Social Security is running a surplus) in the type of transaction that should be familiar to anyone who has read much about Enron and similar corporate accounting escapades. We have spent all the Social Security surpluses and will continue to spend them unless we amend our wicked ways.
After rolling around on my mental floor in horror infused with despair for a few days, I decided to cope by showing you how very ordinary people can do a set of simple calculations to figure about what's really happening and going to happen with Social Security for themselves. All the information used is publicly available. Spend half an hour with me, and then you will be far more qualified than a lot of column writers to discuss the issue of whether Social Security is in a crisis or nearing one.
It's really rather simple once you ignore the funny money. The most important variables in this system are workers per retiree, average monthly worker wage, and average monthly retiree benefit. The further you go out in years, the more uncertain each of these variables becomes, so we will concentrate on the years up to 2025. In order to make matters simpler, we will render everything in today's dollars, assuming that both wages and benefits don't change with time. They do, but at the current time they are pretty closely tied together when calculating benefits, so we can ignore that change quite safely for our purposes.
So come with me, children, and calculate with Mama. Leave the pundits to enjoy their sojourn at Walt Disney; the voters can catch up with them later, and we will.
We will get our figures for projected workers per retiree from Social Security's own latest projections, which are as good as they can possibly make them. They offer three sets of figures, and we will use the intermediate ones. We don't want to alarm ourselves unnecessarily.
Coming up with average worker wage is only slightly more challenging. The government does surveys, and we will use them. We will check our figures with two different methods, because we want to make sure our numbers are reasonable. From this survey (the latest available, slowly loads in pdf) we find that the average hourly wage across private, state and local governments is $17.75. Ominously, we note that government wages are higher than private wages. This will become important later. $17.75 * 39 hours a week * 52 weeks a year gives us about $35,997, or about $36,000 annually. From this survey (much faster txt), we find (it's right at the top) that the average wage nationally is $36,000.
These numbers are slightly old, but close enough for our purposes. These are not mean wages, but average wages, meaning that if you multiplied the number of wage-earners by the average wage you would end up with total wages for the entire nation. So in reality, somewhere around 4 to 6 percent of these wages would not be subject to the payroll tax. Certain state employers (like some university systems) are excluded from their half of the SS tax, and some workers exceed the wages subject to Social Security tax. This will offset the increase in wages we may be missing by not having figures for 2004 or 2005.
$36,000 a year breaks down to $3,000 per month. Social Security taxes on subject wages are 12.40, half paid by employer, and half by the employee. So the monthly contribution per worker is 12.4 % of $3000 which equals $372 dollars. We'll round that up a bit to $380 for convenience and because we want to be sure not to exaggerate the problem.
Now we come to the next item, which poses slightly more of a challenge: how do we figure out average monthly benefit per retiree? Here we will revert again to two useful pages here and here provided by the Social Security Administration. On the second page we find two examples of how a current retiree's benefits would be calculated. The lower number (for a low wage earner retiring at 62) is $1273.74, and the higher number (for a more typical wage earner retiring at 65) is $1508.80. An example for a really high wage earner is not given, but they represent a small percentage of the workforce anyway. So we will interpret between these two numbers and derive a thumbnail figure of $1,400 monthly benefit per current retiree. Certainly 20 years in the future most retirees receiving benefits will have retired in 2004 or later, so when projecting forward in today's dollars this is probably a pretty accurate figure. If anything, it should be a bit low.
Still, we want to check ourselves, so we will try to calculate this in a different way. You can see a historical list here of average monthly benefits up to 2002. The last number given is $895 over all retirees. Extrapolating from the past five years and giving more weight to the lesser increases in more recent years, we come up with an average increase per year of about $25.00. $25 * 10 years equals $250 per decade, and two decades yields an increase of $500, which added to $895 in 2002 gives $1,395 in 2022. Not a bad fit, so we can feel pretty comfortable with our $1,400 average monthly benefit figure. Remember, we are not being paid for this - all we are trying to do is figure out whether we need to worry about Social Security deficits or not.
Okay, now it's all simple.
Currently there are about 3.3 people paying for each retiree, and on average each month, each one is paying $380 into SS. 3.3 * $380 = $1254. Adding $50 (updating 2002 figure to 2004 app. monthly benefit) to $895 means we need to pay about $945 per retiree, so we're running a nice surplus. No crisis now.
But if current workers were paying a higher benefit to each retiree, we would have a problem. $1,400 / 3.3 = $424.24 per worker - call it $425. That doesn't looks so good.
When the ratio drops to 2.5, it will be $560 per worker.
When the ratio drops to 2.0, it will be $700 per worker.
How much would the average monthly salary in today's dollars have to be in order for 2 workers to pay for one retiree's average earnings 20 years in the future, expressed in today's dollars? .124 * X = $700; $700 / .124 = $5,645, expressed in today's dollars, which is equivalent to yearly earnings of $67,740. The actual number in the future would be much higher of course.
It's not going to happen, honestly, no matter how much productivity grows. And remember, given the current method of calculating retiree benefits, when average wages go up so do retiree benefits, so actually the benefit per retiree would be much higher. This does not look promising; it appears there is no way under today's system to catch up with ourselves once the ratio of workers per retiree drops much below 2.7. If this seems confusing, go back to this page and look at how they are adjusting past wages in light of current wages.
But let's try to calculate this another way, to quantify the effects of the expected problem.
In 2010 we expect to have about 3.2 workers paying for each retiree. Moving our average benefit up 8 years (8 * $25 = $200, $200 + $895), we get an estimated average retiree monthly benefit of $1,095. 3.2 workers each paying $380.00 per month gives us $1,216. We are still running a decent surplus in 2010.
In 2015 we expect to have about 2.9 workers per retiree, which gives us 2.9 * $380 = $1,102. To get our expected benefit per retiree, we add another 5 years of increases ($125) to $1,095, equaling $1,220 rounded. It looks like we will be crossing over into deficit territory. We could be a few years off either way, but not by too much.
In 2020 we expect to have about 2.6 workers per retiree, which gives us 2.6 * $380 = $988. Adding another 5 years of increases, we can guess at an average monthly retiree benefit of $1,345. We have a deficit per retiree of at least $350 per month. What the heck - let's be conservative and say it's $300 per month per retiree.
How many of these retirees are there going to be? 67,987,000 of them, according to the intermediate table. But over 10,000,000 of that number receive disability only and on average receive less monthly. So we'll round down to 60,000,000 retirees and multiply by our $300 deficit per retiree to get a monthly deficit of $18,000,000,000. To get the annual deficit multiply by 12, which gives us $216,000,000,000. That is conservatively $216 billion that social security will cost us above what workers pay into the system in 2020. Pulling off our shoes and socks, we count anxiously on our toes and the fingers of one hand, and realize that 2020 is only fifteen years from now.
This is getting ugly fast. There is no trust fund - if T-bills are redeemed either those 2.6 hardworking wage-earners per retiree, corporations, or the retiree is going to have to pay for those T-Bills in extra income taxes. Probably all will be pay higher taxes. The effect on the budget gets worse when you realize that in 2003 the Social Security surplus reduced the deficit by over $70 billion (ignore those stupid interest earnings). The net effect on the budget will be, very conservatively stated, an additional $280 billion in deficit spending compared to current figures.
Let's move forward to 2025, because we said we were going to go out 20 years. We will use the figure of $1,400 monthly benefit per worker, which is definitely too low. But again, we are lowballing. We expect to have 2.3 workers per retiree. 2.3 * $380 = $874. Not good. We now have a deficit per retiree of over $525 per month, but we'll be conservative and use $450.00.
How many retirees do we expect to have in 2025? The total expected is approximately 76,000,000, but about 11,000,000 of those are projected to receive disability only. So we'll take a conservative figure of 68,000,000 retirees and multiply by our $450 deficit per retiree to get a total projected monthly deficit of about 65,000,000 of those will receive Let's use just the number 30,600,000,000. To get the annual deficit we multiply by 12, which gives us $367,200,000,000. That is about $367 billion extra that those 2.3 workers, corporations and retirees will have to come up with in 2025. It could easily be substantially more. The net effect on the budget versus current figures is approximately $440 billion in additional annual deficit, which is about equal to the ENTIRE deficit now. Remember, we are doing all of this in today's dollars.
When you start to consider the effects of Medicare and Medicaid, which are also highly dependent on the same worker/retiree ratios, you begin to see what a massive hole we will be in 20 years from now. Now it's up to you whether you want to consider this sort of thing a crisis or not, but I question the sanity of anyone who claims we don't have to worry about this until 2042. There is no trust fund. We have been using SS taxes to pay for general expenditures that we should have been funding from current income. The longer we keep doing it, the worse the situation gets.
What are our options?
Things that might help to reduce the SS shortfall would be more immigration, higher birth rates, cutting back on our overall federal spending sharply, or changing the way Social Security benefits will be calculated in order to give a lower average benefit. This is one reason why I favor more legal immigration. Higher birth rates can only be achieved by cutting real taxes on the younger and poorer of our population. As I explained in this prior post, we are taxing lower wage earners to death and most of the taxation on younger or less affluent workers comes from Social Security.
However, if we cut taxes for these people we will increase the overall budget deficit now. But we certainly should not be raising taxes on the lower 1/2 of wage earners, and we should be embarrassed at the amount of taxes many of these people are paying today. As I pointed out in this post, raising payroll tax rates on higher wage earners is getting close to counterproductive already. We are sending jobs overseas to cheaper locations at a very high rate.
We are running out of options to address the situation with each passing year. What we are going to do will end up being cutting benefits, raising taxes, and cutting benefits very sharply for the top echelon of retirees or those with significant other assets. There is no other way out, but certainly we can't spring this on people at the last moment. We need to allow people time to prepare, and we need to figure out what we are going to do for the lower echelon of wage earners who are being taxed at historic highs and have little chance of saving with what they have left.
As for raising the full retirement age, that has already been done. People can still take early retirement at age 62, but if you increase that age you will force a lot of older people to go on disability. That will end up costing more, because those persons will get not only disability but Medicare.
Note: A careful reader will note that I have slightly earlier negative dates, even lowballing, than the Social Security projection to which I linked. That is partly because I ignore the funny money revenue from the T-Bills in the trust fund and so on. The point is, the tax-payers pay that money into SS just as they pay their payroll taxes, so it is a wash and need not be considered. I have also ignored the "revenue" from these T-Bills in my calculation of how SS offsets the budget deficit currently. I believe my figures are quite accurate - in fact I believe they are underestimates by about 15% by 2023. But that's a much longer set of calculations.
Factors which the social security projections don't address is that high unemployment rates among older people and lower medical insurance coverage rates (largely related to high medical insurance costs) are forcing far more retirees to take early retirement at age 62 (they use their SS check to pay for medical insurance or offset medical costs). This does lower the average benefit, but it increases retirees per worker, so the effect is pretty much another wash. In 20 years, almost all of the retirees in the program will have retired under the current formula, so the average benefit in today's dollars per retiree is a pretty good figure.
Update: The CBO's latest release is out, and Donald Luskin read it and pulled out the relevant passage. The Congressional Budget Office has SS cash flow going negative a couple of years later than I do. But the facts are clear - we will be in the hole before 2020.
The Congressional Budget Office writes:
Thus, in CBO’s baseline projections, the capacity of the Social Security trust funds to offset some of the net deficit in the rest of the budget—as they do now—will begin to dwindle during the coming decade. Shortly thereafter, Social Security is projected to begin adding to deficits or reducing surpluses.The reality is quite clear, even if one looks at the numbers for an hour or two. I hope that I have helped one or two people to understand the difference between what is false and what's not. It is very confusing to suddenly have this burst of unrealistic articles claiming that Social Security will be able to pay benefits until 2052.
You can't honestly make that claim without looking at the budget as a whole, and the budget as a whole has a lot of sinkholes in it already. Last year, although the precise figures haven't been released yet, Medicare and Social Security taken together ran a deficit and some contributions from general revenue were required in order to make payments to beneficiaries.
Comments:
<< Home
Lancelot,
I'm not ignoring you, but I'm pressed for time. Wage increases will increase income faster than outgo for a while, but then the effect may well reverse itself later on, as the ratio of workers to retiree drops.
So, for instance, if the bulge of workers retiring begins to push wages up by 2010, while we still have a relatively high worker/retiree ratio, for a while things will look better. But what this will also push the average monthly retiree benefit up, and later on that higher average benefit becomes more significant. If you couple this trend with the likely effect that the Medicare/Medicaid/SS deficits are likely to have on the economy by 20 years out (pushing growth lower as taxes or deficit spending increases), then things look ugly in 2025 - 2030, because now you have low or negative real wage growth, considerably higher average benefits, and fewer workers per retiree.
If you can construct a scenario under which the workers per retiree stay higher, the effect of wage indexing is more benign. Low wage growth in the next 15 years is not necessarily bad for the long term forecast if growth resumes later on, because it shifts the debt burden earlier and lessens the total debt, and then increases inputs to the system later when we need them the most. But almost the worst possible scenario is high wage growth for a decade or two, and then plummeting wage growth after that - and if we don't get our house in order that is the most likely scenario by far. In that scenario, workers are also more likely to begin drawing benefits earlier.
If I have time, I'll try to show later several scenarios and how they might work out.
I had glanced at the Cato Institute's analyses and suggestions. I'm trying to ignore various suggested solutions right now, because I'm trying to figure out if I'm corect about what has happened in the last two decades. My hypothesis is that we have decapitalized the "up-and-coming" segment of our population, thus suppressing economic growth and stratifying economic classes. This is a problem. If it is as large an effect as I think it is, then we will have to figure out a solution to reduce it, because by any standard, new immigrants fall into the up-and-coming segment.
The other effect purely numeric analyses don't pick up is the falling rate of benefit coverage in the private sector. Big companies are now eliminating retiree medical coverage (I think Motorola, for example, did last year) - so that will have a sharp effect on when people retire. In order to improve the SS problem you probably have to adopt some reforms in the medical insurance field, and one of the most important is full HSA participation coupled with trade organization and private association rights to provide insurance coverage to cross-state pools under federal rather than state law.
I'm not in market research, I'm a programmer working in banking compliance. I do a lot of loan and regulatory calculations. Bankers have to worry about things like disparate effects across demographic segments they service, and long term investment and interest rate risks. It's population dynamics of a sort.
I think going into market research focused on overall trends is probably a good idea. Domestically, the population bulge moving through its retirement years alone will have huge effects. Internationally, the ability to find and open up new trade markets for companies will be crucial. We are entering an era of hyper-competitiveness.
One interesting idea to get the type of immigration we need is to offer open immigration to entrepeneurs from other countries, allowing them to move their companies lock, stock and barrel, along with most of their employees, their employees' families, and their own families. Because so much industry is now computer-based, we could probably pull in a very high quality people and companies from the Asian region at relatively little transitional expenses to those companies.
We could offer 7 years of no income taxes to help them with transitional costs, but we would get immediate return from the additional wage income taxes and social security taxes. The best asset the US has to sell is a very stable business environment (if we try to maintain that as a conscious goal), open opportunity in a classless society (if we strive to maintain that), and peace.
If we define the problem, we will be able to define the solutions. If we keep wandering around in the land of denial, we are in a deep, deep hole.
Post a Comment
I'm not ignoring you, but I'm pressed for time. Wage increases will increase income faster than outgo for a while, but then the effect may well reverse itself later on, as the ratio of workers to retiree drops.
So, for instance, if the bulge of workers retiring begins to push wages up by 2010, while we still have a relatively high worker/retiree ratio, for a while things will look better. But what this will also push the average monthly retiree benefit up, and later on that higher average benefit becomes more significant. If you couple this trend with the likely effect that the Medicare/Medicaid/SS deficits are likely to have on the economy by 20 years out (pushing growth lower as taxes or deficit spending increases), then things look ugly in 2025 - 2030, because now you have low or negative real wage growth, considerably higher average benefits, and fewer workers per retiree.
If you can construct a scenario under which the workers per retiree stay higher, the effect of wage indexing is more benign. Low wage growth in the next 15 years is not necessarily bad for the long term forecast if growth resumes later on, because it shifts the debt burden earlier and lessens the total debt, and then increases inputs to the system later when we need them the most. But almost the worst possible scenario is high wage growth for a decade or two, and then plummeting wage growth after that - and if we don't get our house in order that is the most likely scenario by far. In that scenario, workers are also more likely to begin drawing benefits earlier.
If I have time, I'll try to show later several scenarios and how they might work out.
I had glanced at the Cato Institute's analyses and suggestions. I'm trying to ignore various suggested solutions right now, because I'm trying to figure out if I'm corect about what has happened in the last two decades. My hypothesis is that we have decapitalized the "up-and-coming" segment of our population, thus suppressing economic growth and stratifying economic classes. This is a problem. If it is as large an effect as I think it is, then we will have to figure out a solution to reduce it, because by any standard, new immigrants fall into the up-and-coming segment.
The other effect purely numeric analyses don't pick up is the falling rate of benefit coverage in the private sector. Big companies are now eliminating retiree medical coverage (I think Motorola, for example, did last year) - so that will have a sharp effect on when people retire. In order to improve the SS problem you probably have to adopt some reforms in the medical insurance field, and one of the most important is full HSA participation coupled with trade organization and private association rights to provide insurance coverage to cross-state pools under federal rather than state law.
I'm not in market research, I'm a programmer working in banking compliance. I do a lot of loan and regulatory calculations. Bankers have to worry about things like disparate effects across demographic segments they service, and long term investment and interest rate risks. It's population dynamics of a sort.
I think going into market research focused on overall trends is probably a good idea. Domestically, the population bulge moving through its retirement years alone will have huge effects. Internationally, the ability to find and open up new trade markets for companies will be crucial. We are entering an era of hyper-competitiveness.
One interesting idea to get the type of immigration we need is to offer open immigration to entrepeneurs from other countries, allowing them to move their companies lock, stock and barrel, along with most of their employees, their employees' families, and their own families. Because so much industry is now computer-based, we could probably pull in a very high quality people and companies from the Asian region at relatively little transitional expenses to those companies.
We could offer 7 years of no income taxes to help them with transitional costs, but we would get immediate return from the additional wage income taxes and social security taxes. The best asset the US has to sell is a very stable business environment (if we try to maintain that as a conscious goal), open opportunity in a classless society (if we strive to maintain that), and peace.
If we define the problem, we will be able to define the solutions. If we keep wandering around in the land of denial, we are in a deep, deep hole.
<< Home
