Hacker News Reader

  "unlimited for life with a fixed initial cost" products

Interestingly, this is the same problem that bedevils Social Security and Medicare/Medicaid. Consider the change in perspective that AA underwent when it realized how much these passengers were costing the company:

  Rothstein, Vroom and other AAirpass holders had long been 
  treated like royalty. Now they were targets of an 
  investigation.

The same thing happened with Medicare. For any costly program that doesn't put financial caps on utilization, cost can ultimately be capped either by queues or by various kinds of fraud charges and investigations into the users of the program. This is why we have the Stark Law and Medicare fraud investigations:

http://en.wikipedia.org/wiki/Stark_Law

http://en.wikipedia.org/wiki/Medicare_fraud

Some of the asserted frauds are actually crimes. Some of them are in a gray area and arguably abusive, like Rothstein. Some are really just attempts at innovative business models. Whatever the reason, though, if you are costing the government a lot of money, you are going to get hit with a Medicare fraud investigation. It doesn't matter if it's because your product is amazing and the cost is due to organic growth. All that matters is how much money it's costing the government.

I do not know about medicare per se, but when social security was designed to support people 65 and over, I think life expectancy was around age 62. So it was designed to support a relatively small number of retirees in their extreme old age. Then people began living longer.

Somewhat similarly, welfare was aimed at a tiny fraction of "the deserving poor" but was poorly worded/thought out and thus inadvertently changed the social contract and thereby actively grew the number of poor moms. At the time it was created, having a kid out of wedlock was strongly taboo. Poor single moms were thus mostly widows. The authors apparently could not imagine it becoming more socially acceptable to have children out of wedlock.

I worked at bigco for five years. Similarly shortsighted wording for contracts and the like is pretty rampant from what I have seen.

> I do not know about medicare per se, but when social security was designed to support people 65 and over, I think life expectancy was around age 62. So it was designed to support a relatively small number of retirees in their extreme old age. Then people began living longer.

This is true, but the real problems with Social Security began when Lyndon Johnson combined the Social Security fund, then entirely separate, with the rest of the federal budget. Instead of maintaining a very close eye on how much money was coming in and going out for Social Security, there was no longer any urgency to keep that money around. In other words, money taken out of paychecks for Social Security was spent on everything else, and Federal income taxes were paid out as Social Security.

From what I have read, it is complicated, probably too complicated for me to ever debate off the cuff. And I am especially not qualified right now.

But thanks for contributing.

> I think life expectancy was around age 62

You are thinking of life expectancy at birth, which averages in childhood death.

SS only pays people who worked for at least a few years, so life expectancy for people who reach age ~20 is much higher.

The Republican-invented meme of social security insolvency because it didn't account for life expectancy increases is tired and worn out. The reality is that the architects of it were nowhere near that stupid, and they accounted for evolving life expectancies with a fairly high degree of accuracy. Had they not, the fund would have been bankrupt long ago.

Even now, the system faces not a crisis, but a need for adjustments. In the 2030s, incoming taxes are expected to be at approximately 75% of payouts. If no changes are made, once the trust fund is exhausted, people will continue receiving at least 75% of their entitlement indefinitely.

Not being able to pay your outgoings is the definition of insolvency, friend.

On the other hand, the government is not quite a 'corporation,' and isn't entirely subject to those kinds of notions, is it?

> Not being able to pay your outgoings is the definition of insolvency, friend.

Didn't say it wasn't. Try not randomly reshuffling the words. You'll get a much clearer picture of what I actually said.

Only paying part of your outgoings is considered 'not paying,' or grounds for a default, typically? So that 'meme' is pretty much factual reality, if your statistics are correct?

disclosure: non-American with no knowledge of your welfare scheme

Not being able to pay something twenty years from now given current income is much different than defaulting. Specially given historically low tax rates and an economic depression, and a spenditure that has mostly not been made yet.

I agree, but that doesn't seem particularly relevant. The system in question would generate significant revenue shortfall if that fellow's comments are accurate. Typically in public policy we make plans for these kinds of scenarios, instead of just assuming that Future Murika will have a tax base robust enough to pilfer from and a political scenario permitting us to actually legislate.

On the other hand, that strategy is working fine for Twitter...

Look very closely at what I actually said. The meme I speak of is not "insolvency" generally, but insolvency due to a specific cause.

> disclosure: non-American

Which may explain why you're not familiar with the particulars of the meme I refer to, but does not excuse randomly shuffling my words.

> with no knowledge of your welfare scheme

Social Security is not a "welfare scheme", it is a government-administered pension program. Everyone who has paid into the system, from the very poor to Bill Gates and Warren Buffet, is entitled to benefits at retirement in proportion to their contributions.

You are badly misinformed. There is no trust fund, only government IOUs to itself. The program already requires revenues from the general Treasury, and the accounting collapse of the fictional trust fund is now in 2020, not the 2030s.

http://www.ssa.gov/oact/TRSUM/index.html

If USA government "IOUs" are worthless you better tell market traders who consider them one of the safest investments out there.

The treasury bills owned by SSA are special non-marketable ones that are only sold to the SSA. They do not have any market value.

I have no opinion regarding the solvency of social security.

What this critique amounts to is saying that the US will always honor its commitments to the market but not to it's tax payers. But why would that be? Are the financial elite more powerful than millions of retiring voters? Social Security taxes were raised in 1983 with the expectation of the baby boomers retiring 30 years later. But in those 30 years the surplus SS money was used to paper over deficits caused mostly by tax cuts on the top income percentiles. In effect the wealthy have raided the fund, but now they are wealthier than ever and we should hold their feet to the fire to pay the money back.

But in those 30 years the surplus SS money was used to paper over deficits caused mostly by tax cuts on the top income percentiles.

This is getting way off topic, but the lower and middle class portions of the Bush tax cuts, for example, contribute way more to the deficit than those for the upper class. And your claim doesn't really square with the trend of more % of all income taxes being paid by the top percentiles over time.

Well of course they are paying more taxes, they are earning way more money.

A Ponzi scheme is also really safe, if it's clear that there are a ton of gulliable fools willing to support the pyramid long enough for you to get in and get out. :)

I am aware of all the information behind your assertions, so no, I am not "badly misinformed", thanks, I simply reach very different conclusions than you do.

What it did not account for is a shift in the wealth gap and a stagnation of wages, relative to inflation, for a majority of people.

The wage base will need to be adjusted to reflect these changes. The last time it was changed was 20 years ago under Reagan.

20 years ago under Reagan

30 years ago, to the nearest decade. That was the change in Social Security policies that raised my retirement age and the tax rate that I pay into the system. My standard retirement age is just after Medicare "exhausts," that is runs out of accumulated funds to pay full benefits to people who have been paying into the system for their entire working lives.

Yes and no.

The original New Deal guys had mechanisms in place to deal with those sorts of things. But in order to keep funding in place for the Vietnam War and other things, the Congress accelerated cost of living adjustments in the 60s and 70s -- basically using Social Security as a political tool.

Since adjustments were made long ago, many other tweaks have affected Social Security. Welfare reform pushed folks who were chronic public assistance recipients onto the SSDI rolls. The ceiling for payroll tax adjustments hasn't kept up with inflation. And of course, the Obama administration decided to cut the payroll tax in the short term to stimulate the economy.

I hear this argument all of the time here in conservative Idaho, that social security provides more of a benefit than what's been paid in. This is fundamentally NOT how social security works. It has always been self-funded, and if not for the manipulations of unscrupulous politicians who borrowed from it to pay for 30 years of wars, it would be doing just fine today Thank You Very Much.

Most people my age (35 and younger) believe that they can't rely on social security for their retirement. It takes some meta thinking to go a step above and realize that this line of reasoning is leading to a self-fulfilling prophecy, meaning we are unlikely to fight for a system we think is doomed.

All hogwash of course. Productivity is at least an order of magnitude or two higher today than it was in the depths of the Great Depression. Most staples like food, clothing, transportation and even shelter are cheaper now than they were 75 years ago, adjusted for inflation. We have no way to predict how high productivity will be in 2030, when social security will supposedly encounter "problems".

This all dwarfs any increase in lifespan, desired affluence, or cost of medical care we are likely to encounter (once real progress and cures begin happening, instead of the phantom progress we are experiencing in tech bubbles).

People complain more about the self-funding social security payroll tax of 6% than they do about the 3% rents charged by credit card companies. Heck some people spend more on their smartphone, cable and internet plans than they do on social security.

I really worry for the future of our society, when we freely give up all of the gains made in the 20th century like 40 hour work weeks, rights to privacy, shared public land and other resources like the airwaves, immigration for the huddled masses, and most importantly shared sacrifice and that our society is only as strong as the weakest among us, all the things that Americans used to believe in. We think we are somehow better or wiser than the people some respectfully call the Greatest Generation, who survived the depression, then fought in WWII and died for the standard of living we enjoy today.

We're better than this, especially here in such a mecca of ideas and enlightenment.

There is some truth to this, but not enough. Social Security provides a fixed rate of return, and provides benefits throughout old age, regardless of how long one lives. How? Not through investment, through fiat. And ultimately that's the core of the problem, because the only way to achieve that is to have more workers paying in to the system than there are retirees pulling from it. Whether it's dollars or government bonds sitting around in a vault somewhere the problem of making them multiply by fiat will always exist. And that's the problem that Social Security ultimately faces. Within the next decade we will face the "we've been stealing from Social Security" problem, but about 15 years after that we will face the absolute "Social Security can't pay for itself at all" problem.

This is taken directly from a Metafilter post[1]. In short, it seems the "social security is broke" trope isn't an accurate depiction of reality. The post is lengthier than this snippet, includes plenty of citations, and is worth reading in its entirety.

"Someone just told me that living outside the US for 5 years could disqualify me for Social Security benefits - I told them that I never expected to get them anyway.

It's been said upthread, but this pervasive attitude shows that the rich and their propagandamiesters are winning.

Straight talk on Social Security[2]:

'Here’s what Social Security is not:

- going broke;

- a Ponzi scheme;

- expected to stop paying out benefits in your lifetime;

- bankrupting our nation or future generations.'"

[1] http://www.metafilter.com/115423/My-FaithBased-Retirement#43...

[2] http://www.rollingstone.com/politics/blogs/national-affairs/...

That's nice, but it doesn't reflect reality. In reality it will require a great deal of budgetary pain just to get the government to pay out the social security trust fund. That represents a debt the US has made to itself but we haven't been very good at paying off even those loans. If we don't make any hard choices such as raising other taxes or cutting overall spending then we'll end up with an even greater increase in our national deficit and an accelerating buildup of our national debt. And that will cause the whole system (the economy, the government, everything) to come crashing down long before the so-called social security trust fund "runs out".

Right now a social security payments surplus is subsidizing many other government programs, when that subsidy ends within the next decade it will cause hardship.

More so, social security is on a glide slope to begin running an absolute deficit in about 25 years. At that point the trust fund will be emptied and there won't be as much money coming in as going out. That's a pretty serious problem.

Medicare is even worse off, by a huge margin.

Yes, we can change medicare and social security and perhaps fix these things, but right now the train is on a set of tracks which are aimed directly at the edge of a cliff. We can also potentially balance the whole budget and pay down our debt as well, but that isn't happening either. Look at the EU and their budgetary crises right now. The public doesn't want to pay down their debts, they don't want "austerity measures", they don't want to pay their bills, they want benefits to remain forever high regardless of whether that's financially feasible. And they are willing to vote and riot to have their way.

Social Security and Medicare are not sustainable in their current forms. Period. They need fixing. Period. Fixing them will be really motherfucking hard (politically and financially) for our country. Period.

I hear this argument all of the time here in conservative Idaho, that social security provides more of a benefit than what's been paid in.

For many people (the subset of the working poor who end up living a long time) this is absolutely the case. For many other people (those with a below average lifespan or substantially above average career earnings) Social Security is a terrible deal and has no chance under the current formula of paying them back what they put in. Even if you don't count what their employers had to contribute, which might otherwise have been spent on higher wages.

It seems inevitable that the SS payroll tax cap will be lifted without substantial changes to the benefit formula, which will make it an even worse deal for, say, the median earner with a post-graduate degree.

That doesn't demonstrate that it's a bad program, just that's it's redistributive and we should be honest about that fact. A smart, productive 35-year old should absolutely not rely on Social Security for her retirement, assuming she doesn't want her lifestyle to drop off a cliff at age 68 or whatever the retirement age is by then.

Social Security is redistributive much in the same way that flood or fire insurance is redistributive: those who never need it have overpaid, and those who do need it have underpaid.

The flaw that exists in the current discourse regarding Social Security is that the program is framed as a retirement program. It is not. It is "end of life insurance". In fact, one correct term to use in describing programs such as Social Security is "Social Insurance":

http://en.wikipedia.org/wiki/Social_insurance

No, it's not redistributive merely like fire or flood insurance. My $300,000 house will cost the same amount to insure as my similarly-housed neighbor who makes twice as much. And the additional cost to insure a $400,000 house isn't typically greater than the reduction in cost to insure a $200,000 house.

That's not how social security works. The benefit curve drops off dramatically in relation to the "pay-in" curve:

http://www.ssa.gov/oact/COLA/piaformula.html

And again, things will get worse when, seemingly inevitably, the payroll tax cap is lifted without substantially modifying the "bend points". A dollar you contribute to SS on your earnings over $56,000 is worth half the benefits of a dollar contributed under that amount.

  People complain more about the self-funding social 
  security payroll tax of 6% than they do about the 3% rents 
  charged by credit card companies. Heck some people spend 
  more on their smartphone, cable and internet plans than 
  they do on social security.

But I can choose not to use a credit card, buy a smartphone, or sign up for cable or internet. I can't opt out of Social Security.

We have to be clear on whether it's supposed to be a great deal for everyone, or whether it's meant to be a redistributive program.

If it's supposed to be a great deal for everyone, ok, well, I have a different opinion. Let me opt out. You can continue to use it.

Alternatively, if it's supposed to be a redistributive program, well, then you might see why people complain about it. If it's redistributive, by definition some will pay in more than they get out, and others will pay in less than they get out.

All that the rational critics of Social Security, Medicare, and Medicaid want is the ability to opt out, in the same way we can switch from AT&T to Verizon. Right now you can only opt out via the very high friction process of moving and giving up citizenship.

I certainly think the frequent flyer miles thing was a mistake on the part of American Airlines (AA). But one of the things that bugs me about the article is that the 'marginal' cost of flying these folks is the same as the marginal cost of flying someone in coach, or flying someone in general. Weight * cost of gas to carry that weight + consumables such as food, tissues, Etc.

So when AA argues they 'lose millions' they are really saying that if they could sell this persons First Class seat to someone else on short notice they would make more money. But that assumption glosses over the fact that folks fly First with mileage upgrades, or get 'bumped' from coach on an over sale situation, or book way in advance and pay less. Since there isn't any discussion about what the average seat price is for First Class on the same flight when no AirPass member is flying, it didn't feel like they were being particularly objective with regard to price.

The fact that they could not sell AirPass + Companion for $3M (which according to them would have the owner 'ruining' AA in year 2 at this rate) seems to me to validate my feeling that they are over pricing their 'loss' here.

The lesson here is not that lifetime-for-fixed-up-front is inherently idiotic but that it must be sensibly discounted. AA discounting at the anomalous 1981 rates is akin to American public pensions assuming the stock markets will throw 8% at them ad infinitum.

An example of a lifetime-for-fixed-cost product that works is the annuity. Even British consols, perpetuities which pay out forever, are seen as a smart financing method (zero roll-over risk) because they were sensibly priced.

The British War Loan is not a comparable that supports your point, because it is callable.

Fair point, though it stands to be mentioned that this would require an Act of Parliament and hasn't happened since consols were first offered in 1751 (though the coupon was reduced through 1903).

If the purchase price of these tickets were 350,000, and if the typical purchaser was in his 30's at the time of purchase, that gives an annuity of almost $2000 a month, assuming an average lifespan of 70's and a 6% interest (yeah, that may be a tad bit high on the interest assumed, but remember this was the 80's). So in effect, these passengers are paying 2G a month for their high end "bus pass". And if they travel once a week, that is $500 per ticket. I really don't think AA is losing much money here.

Edit: This also explains why these tickets didn't sell recently at $3 million a piece -- for that money, again assuming a 40 year annuity but a 3% interest rate, you can buy a $1200 ticket twice a week.

You get 2 first class tickets for 350,000 and no costs for cancellations and rebookings. That's an incredible service. Booking 3 flights because you're not sure when you want to leave or whatever -- that's pretty huge.

Quibble: the "companion pass" to allow a second ticket was an additional $150,000.

Assuming a lifetime flying pass purchased for $350 000 in 1981 by a 35-year old and the 1981 US life expectancy of 74.01 years [1], one would have to fly only $15 505 each year to beat the 1981 - 2011 CPI CAGR (inflation) of 3.1% [2]. If we instead use the 2009 US life expectancy of 78.09 that number drops to $14 745.

I fly once or twice a month and don't buy first class tickets very often; it wouldn't take too much for me to cross $15k (and as a 21-year old my threshold would be $13 061 assuming constant life expectancy). Remember, the marginal cost to you of these flights is more or less zero.

AA saw this as a source of financing so one should really see what rate they seem to have financed themselves at. Let's assume we're slicing above travellers who fly 100 000 miles (between Delta's Platinum and Diamond [3]; I wasn't able to find what the average of the top two tiers in 1981 would have been). Airline price per mile for domestic coach ranges from $0.06 to $1.22 [4]. Let's use $0.58/mile as our conservative guess for the cost of a first class flight. Thus we have an airline financing $350 000 for 39 years at $58 000/year at an internal rate of return of about...16.5%.

If American Airlines really believed in a 16.5% floor on US interest rates they should have started by dropping the first part of their name.

Note that we've ignored that $350 000 buys both you and your cuddly friend a ticket.

[1] http://www.google.com/publicdata/explore?ds=d5bncppjof8f9_&#...

[2] http://www.usinflationcalculator.com/inflation/consumer-pric...

[3] http://www.delta.com/skymiles/about_skymiles/benefits_at_gla...

[4] http://online.wsj.com/article/SB1000142405274870454090457545...

This kind of stuff goes into management books as classic case of why merely quarter to quarter of focus on company performance is so dangerous.

I am sure the then CEO and his fleet of execs even got a good deal of rewards for rolling out this scheme.

If companies really want to make up on profits in such a low margin industry the way really is to make travel cheaper and easier. Not offering unlimited plans and then expecting customers to not treat them as such.

This serves as a lesson to startups as well

Never offer something for a lifetime

Even if the cost of supporting it "forever" is minimal (that is, the complete opposite of what AA has done)

Things like GMail, for example. You may be tempted to do it. Don't

Always put in your terms of use a provision of termination. Of course, be reasonable, for example, allow the user to backup the data given a reasonable timeframe (e.g. Google Health)

Makes me a bit worried for pinboard.io. It's not free, which puts it in a better position than its competition, IMO, but it's a 1 time only payment. I hear they're profitable now, though, and have vowed to not go away like delicious did.

Pinboard(.in) has a $25/yr "pro" level; believe he's said the pro service is enough to keep it running for everyone.

I just bookmarked this very article in delicious.

Heh, perhaps some Joyent staff may have something to contribute to this conversation. They offered a "lifetime" hosting deal with shell access a while ago for a one-off fee, and I have heard that this is becoming a bit of a millstone for them. Certainly my brothers 'lifetime' account with them is not getting a lot of love and feels a bit like it's being left to rot on decrepit hardware.

I wonder would they do it again with the benefit of hindsight, or did it serve their purpose?

Upgrading users is one of those "damned if you do, damned if you don't" scenarios. If you do: some users are unhappy about old hardware. If you don't: some users are unhappy about uptime and changing software.

Joyent's CEO was on HN a while back and mentioned that if users on the old FreeBSD boxes want to run on the new illumos boxes, they should send an email to support.

I don't think GMail offers anything contractually "forever". The service is always offered "as-is", and you have the option of either continuing to use it or taking your data elsewhere. (Hence the furor over the new UI.)

If you'd said "Never offer something for free", which seems to be what people are responding to, I'd have a whole other set of arguments...

GMail generates revenue, so it doesn't work in your example as a "free" service. You pay for it by agreeing to have your email trawled to pull relevant ads.

"Never offer something for a lifetime"

Unless cost decrease exponentially over time, like with gmail, in witch storage cost falls down like a stone.

Stone falls with no faster than quadratic speed and has a maximum limit if not in the vacuum.

Until there is a major flooding event.

> It boggles the mind every time I read about one of these "unlimited for life with a fixed initial cost" products.

"For life" is a particularly egregious example, but these sort of conflicting incentives are everywhere. It's one reason I've been so happy with the popularity of pay-as-you-go models in, e.g., AWS.

I've always been leery of "unlimited" hosting for a monthly fee because your incentives aren't exactly aligned with the host. NearlyFreeSpeech.net, for example, highlights that fact in their marketing materials that since it's pay-per-use, the more successful your site is, the more money they make, so it's win-win.

Netflix and home broadband also suffer from these problems.

if you do enough research, the "unlimited" hosting isn't unlimited at all. There are certain restrictions in the fine print. which is exactly what American should have done.

I don't know how TiVo feels, but I always thought their lifetime subscriptions were a good way to seed their initial customer base. Lifetime is a little misleading because it's tied to the TiVo hardware, not the lifetime of the customer, but I still think it's relevant. TiVo started with a big problem of knowing they had a product that people would love if they tried it; in that instance, getting people past a hurdle of paying of monthly costs kind of makes sense, especially if they become devoted customers who tell their friends. I'm sure it also didn't hurt their cash flow.

On the flip side I am still using my first generation tivo because I have a lifetime on it and if I get a new device I would have to buy a new lifetime subscription.

I don't know what their current policy is but we've rolled our initial Tivo S1 "lifetime" subscriptions over to S2 and now one S3/Premiere Tivo. Seems to be an occasional special they do.

Even then, lifetime subscriptions were rendered useless in the UK when TiVo dropped support for the original product and teamed up with a cable company. So it was neither a lifetime subscription for the hardware nor the customer.

I'm not sure if it's such a bad idea in this case though. I don't know how much the marginal cost of a seat on a flight to London is, but it's a hell of a lot less than retail. Provided this handful of AAirpass customers aren't, in-and-of themselves, generating more flights that AA has to schedule, their actual sticker price to the company cannot be accounted for at retail prices.

But if the AAirpass people displace other passengers who would have paid retail and instead went to another airline, this opportunity cost should be assessed at retail.

I can't help but think of the deal GM struck with the unions back in the mid 20th century. Lower immediate wages for a nice retirement pension. Little did GM know about the trend of longer life spans.

Well, the people making that deal didn't have their incentives aligned with the incentives of the people on the shop floor.

More like "lower immediate wages for someone else's problem in the future -- maybe we'll have plenty of money to pay, maybe we'll just file bankruptcy, maybe we'll get a bailout"

Also worth noting that GM, like almost all US corporations, chose to under-fund their pension system. Assuming that most of your retirees will die early is like assuming that your stock return will average 8% annually but they've waged a successful PR campaign convincing people that this wasn't an epic managerial botch.

"4. Life memberships are silly. If they want to attract the business flyers they were talking about it should be an annual charge."

I'm not sure an annual charge would necessarily fix the situation for heavy users. It's very similar to the situation of cable and phone companies losing money on heavy users of their service.

An annual charge would give the company the freedom to adjust the terms on an annual basis in response to problems that are discovered. For a high-end product targeted at a very small market and a small part of the overall customer base, I think that makes a big difference.

The problem cable and phone companies have is that lots of people sign up for unlimited plans when available because of the mental transaction costs of doing anything else. That makes adjusting terms to manage heavy users harder because those adjustments also affect a large slice of customers where the unlimited deal is working as designed.

Deutsche Bahn (German railways) provides this option, you can buy a "100% discount" yearly card to either first or second class. But these are quite expensive.

Just hop on the next train: BahnCard 100 offers 12 months travel on all DB services. It only costs 3,990 EUR for standard class and 6,690 EUR for 1st class.

They also have 25% and 50% discount cards available. The 50% is quite a good deal, as I think I paid mine back within the first week of getting it.

http://www.bahn.com/i/view/DEU/en/prices/germany/bahncard.sh...

Railways have an even lower marginal cost than planes. DB would rather have a person travelling in the seat for 'free' and encourage ridership than have it empty.

Airlines used to think this way - the marginal fuel cost for a seat was something like $10 (essentially it only costs you at takeoff). But then airport improvement fees, security fees, faa upgrade fees meant that they were paying out $100s for passengers who were travelling for free.