Phil Trammell: Philanthropy Timing and the Hinge of History

If we want to donate money, should we give it away now or in­vest it to give away later? The an­swer de­pends on many con­sid­er­a­tions, in­clud­ing our ex­pected rate of re­turn, the chance of our per­sonal val­ues chang­ing, and the ques­tion of whether we live at the “Hinge of His­tory” — a time with high-im­pact op­por­tu­ni­ties that will soon van­ish.

In this talk, Phil Tram­mell of the Global Pri­ori­ties In­sti­tute dis­cusses how we can eval­u­ate these con­sid­er­a­tions and change our giv­ing strat­egy in re­sponse to what we learn.

Below is a tran­script of the talk, which we’ve lightly ed­ited for clar­ity. You can also watch it on YouTube or read it on effec­tivealtru­ism.org.

The Talk

To­day, I’ll be shar­ing some thoughts on how to think about the timing of philan­thropic efforts. There are a lot of con­sid­er­a­tions that go into this ques­tion.

1730 Phil Trammell 1
I don’t have time to cover most of them, but here are four that I’ll dis­cuss, along with a sim­ple model for weigh­ing them against one an­other:

1. In­ter­est [on your in­vest­ments]. Of course, if you wait to give, you’ll have more to give later.
2. What could be called the “philan­thropic dis­count rate.” There’s a risk that if you wait too long, we could all be dead, your money could be stolen, or some­thing else could hap­pen.
3. Diminish­ing re­turns. As [the amount of money you in­vest] gets very large, you ex­pe­rience diminish­ing re­turns [on what your in­vest­ments can ac­com­plish]; the value of hav­ing more re­sources is at­ten­u­ated.
4. “Hingey­ness.” I playfully call this con­sid­er­a­tion “hingey­ness” to cap­ture the in­tu­ition that peo­ple have, es­pe­cially in EA [the effec­tive al­tru­ism move­ment] that cer­tain mo­ments in his­tory are par­tic­u­larly ripe for long-last­ing philan­thropic im­pact. The most ex­treme ver­sion of this would be the hy­poth­e­sis that right now we’re liv­ing at “the hinge of his­tory,” as some peo­ple have said.

Th­ese are the con­sid­er­a­tions we’ll weigh against one an­other. What we’ll find is that in or­di­nary times, if you just as­sume that hingey­ness is con­stant, you don’t need to think about it; philan­thropists should prob­a­bly in­vest al­most all of their re­sources.

1730 Phil Trammell 2
The caveat I’ll make is that I’m us­ing a rather broad defi­ni­tion of “in­vest­ments.” An in­vest­ment is any­thing that al­lows you to spend money now and re­sults in more money go­ing to­ward your moral goals in the near fu­ture. So fundrais­ing and move­ment build­ing count. Build­ing schools doesn’t count, be­cause even though it in­creases the earn­ings of the peo­ple who go to the school [and re­ceive an ed­u­ca­tion], the money’s out of your hands once the school’s built.

But in­tro­duc­ing the hingey­ness con­sid­er­a­tion makes things am­bigu­ous. And if the hingey­ness is ex­treme enough, then you should spend a lot.

I’ll just say a few words about the model and some of its im­pli­ca­tions.

First, what is the in­ter­est rate?

1730 Phil Trammell 3
It’s quite high. The U.S. stock mar­ket has av­er­aged an in­ter­est rate re­turn of over 7% per year for the last cen­tury. At that rate, your money dou­bles roughly ev­ery decade, or mul­ti­plies by a fac­tor of two to the tenth power ev­ery 100 years, which is 1,024. That’s prob­a­bly not the rele­vant in­ter­est rate for our pur­poses, for var­i­ous rea­sons. Let’s say it’s 4.4%.

If we are pa­tient philan­thropists — that is, we want to do good but don’t care about when we do it — we should con­sider whether this in­ter­est rate [4.4%] is high or low. Is it an ap­peal­ing op­por­tu­nity for in­vest­ing, spend­ing, or even bor­row­ing?

1730 Phil Trammell 4

There’s a very sim­ple ob­ser­va­tion that should lead us to think that [an av­er­age 4.4% in­ter­est rate] is, from a pa­tient philan­thropist’s per­spec­tive, prob­a­bly high most of the time. This is what’s known as the Ram­sey for­mula, which is the back­bone of the eco­nomics around dis­count­ing. Let’s say there’s con­sump­tion growth of 2% per year. And there’s a pa­ram­e­ter, eta (𝜂), which cap­tures the ex­tent to which marginal util­ity for peo­ple diminishes in con­sump­tion, such that it gets more ex­pen­sive to help some­one at a rate of 𝜂 mul­ti­plied by the rate at which they’re get­ting richer per year. So if peo­ple are 2% richer next year, it’s 2.4% more ex­pen­sive to buy them a util [a unit of hu­man welfare].

The rea­son why peo­ple [might choose not to] in­vest $1.044 next year in­stead of $1.00 now is they dis­count their own welfare at some pos­i­tive rate — let’s say 2%. That’s what delta (𝛿) de­notes. An in­di­vi­d­ual’s own mor­tal­ity risk could be part of that. But if you’re a pa­tient philan­thropist, you don’t care whether you help some­one now or help their de­scen­dants, so you should re­move delta from the equa­tion. If you wait, you can do 2% more good than you’d be do­ing by just giv­ing this year, if your goal is to help peo­ple sim­ply by in­creas­ing their con­sump­tion. Im­pa­tience is in­cor­po­rated into mar­ket in­ter­est rates; in gen­eral, if you’re pa­tient, it’s a good idea to in­vest.

There are some ob­vi­ous caveats to this.

1730 Phil Trammell 5 You don’t want to com­pletely re­move the [so-called] pure time dis­count rate, [since it ac­tu­ally in­cor­po­rates risks] you do also care about — risks like ex­tinc­tion, ex­pro­pri­a­tion, or value drift. And you won’t ac­tu­ally do 2% more good each year than if you wait to give, be­cause if [the amount of money you set aside gets] re­ally big, then fur­ther growth to your fund will have diminish­ing re­turns. But if your goal is just in­creased hu­man con­sump­tion, the [risk] of get­ting so big that you’d ac­tu­ally push down peo­ple’s marginal util­ity prob­a­bly [is re­mote]. That wouldn’t hap­pen for a very long time.

The main take­away, there­fore, is that in nor­mal times, you prob­a­bly want to take ad­van­tage of this [op­por­tu­nity and give later].

Upon first en­coun­ter­ing this sim­ple ob­ser­va­tion, peo­ple of­ten [voice] a litany of ob­jec­tions. Un­for­tu­nately, I won’t have time to go through all of them. I’ve just listed them [on the slide be­low] with one-liner re­sponses.
1730 Phil Trammell 8

For ex­am­ple, peo­ple some­times say, “Well, what if the poor are get­ting richer so much faster than the in­ter­est rate that it’s ac­tu­ally get­ting more ex­pen­sive to help them each year?” This can only hap­pen for a tem­po­rary pe­riod of time. Even­tu­ally the good that you’ll do by in­vest­ing, if noth­ing struc­turally changes, will out­strip the good that you would have done by giv­ing to the poor now — even if the fac­tor is 250 to 1. In that case, you’ll have done more good in 275 years or so by wait­ing, if this delta equals 2% for­ever.

1730 Phil Trammell 7

If your goal is just to help peo­ple, and you think that peo­ple are im­perfect at helping them­selves — in par­tic­u­lar by be­ing im­pa­tient — then the pa­ter­nal­ism ob­jec­tion doesn’t re­ally hit home.

I want to share some in­tu­itions be­yond the some­what glib one I just gave for think­ing that pa­tience might be a su­per power that, if we make use of it, al­lows us to have a re­ally big im­pact on the fu­ture.

1730 Phil Trammell 9
Sup­pose you’re at an auc­tion house and some base­ball cards are up for auc­tion. You’re a base­ball fan; you and the friends you went to the auc­tion with are ob­sessed with base­ball cards. [If no one else at the auc­tion is a base­ball fan], you should be able to out­bid the other peo­ple at the auc­tion house for the base­ball cards.

That’s similar to the po­si­tion we’re in [as philan­thropists who care about the fu­ture]. The fu­ture, in some sense, is this large, four-di­men­sional ob­ject, and [peo­ple in the effec­tive al­tru­ism move­ment] think about it a lot — and most peo­ple don’t. There should be a way for [pa­tient philan­thropists] to buy [the fu­ture] from its cur­rent own­ers. There’d be a weird mar­ket failure if that weren’t pos­si­ble.

In fact, we can liter­ally do this. If you buy land and then sell a 100-year lease on the land, you’re buy­ing the right to say what hap­pens to the land from 100 years out un­til the end of time. And a 100-year lease on the land usu­ally costs about 90% as much as the land. Ba­si­cally, for a 90% dis­count, you’re get­ting the land, or al­most all of it [if your own­er­ship per­sists for hun­dreds or thou­sands of years af­ter the lease is up].

That’s what in­vest­ment is. It’s say­ing, “Hey, im­pa­tient per­son, you can have some re­sources just for a while, and then give more back to me later.”

Peo­ple some­times spec­u­late that we’re liv­ing at a very spe­cial time with re­spect to giv­ing — that our dona­tions can have a gi­ant im­pact on the fu­ture. But what this amounts to is the ob­ser­va­tion that we’re also liv­ing at a spe­cial time with re­spect to the value of in­vest­ing. If the dy­nam­ics of pure time prefer­ence play out, even­tu­ally pa­tient peo­ple will own more and more of the world. In­ter­est rates will no longer in­cor­po­rate pure time prefer­ence [to the same ex­tent, be­cause most as­sets will be held by peo­ple who don’t have such prefer­ences]. In a sense, the fu­ture will have been bought. So this, too, is a fleet­ing op­por­tu­nity.

1730 Phil Trammell 10

The idea [in the slide above] is to sup­pose that you face diminish­ing re­turns pa­ram­e­ter­ized by eta (𝜂). Your im­pact at a given time is some con­stant, mul­ti­plied by the amount you’re spend­ing, trans­formed by these diminish­ing re­turns. And sup­pose that a com­mu­nity of pa­tient philan­thropists got to­gether and acted as a sin­gle philan­thropic agent. What if the com­mu­nity had no in­come stream, but be­gan with a fixed pool of as­sets?

Let’s as­sume that’s the situ­a­tion we are in. How much should we spend? It turns out that it’s the num­ber rep­re­sented by the for­mula [in the slide above]. And if eta equals one, which is the case of log­a­r­ith­mic im­pact, then it’s re­ally sim­ple: It’s just your philan­thropic dis­count rate.

1730 Phil Trammell 11

If there’s a 0.5% chance per year that ei­ther we go ex­tinct or your money is stolen, then that also turns out to be the amount that you should spend ev­ery year. And you should in­vest [the other] 99.5% of your re­sources ev­ery year.

What if, on the other hand, you start out at the other ex­treme? You have no re­sources, and the com­mu­nity is a group of young peo­ple with noth­ing in their pock­ets, but they have an in­come stream. Then what frac­tion of that in­come should be put to philan­thropic pur­poses each year, and what por­tion should be in­vested?

I think most small donors think of them­selves as be­ing in this po­si­tion, but they give ev­ery­thing that they plan to give each year in­stead of in­vest­ing it in a fund to give later. But it turns out that you should spend this ra­tio: the philan­thropic dis­count rate di­vided by the mar­ket im­pa­tience rate — per­haps about 25%.

Now if this played out for­ever, then you’d grow in­definitely a bit faster than the eco­nomic growth rate.

1730 Phil Trammell 12
In 400 years [a com­mu­nity in­vest­ment pro­ject] might [yield some­thing] like 10% of the global wealth. In re­sponse to that, peo­ple will say, “If this is so easy, then why hasn’t it been done be­fore?” That would be a whole sep­a­rate talk. But peo­ple have tried to do it in var­i­ous ways. I don’t think they’ve done a great job. It prob­a­bly wouldn’t work any­way [even if it were done well], but that’s already ac­counted for in ẟp. A 0.5% philan­thropic ex­pro­pri­a­tion rate would mean that the half-life of an effort to buy the fu­ture would be 200 years [by which time, there’s a 50% chance that your in­vest­ments dis­ap­pear be­cause of some un­fore­seen event]. After 400 years, there would be only a quar­ter-chance that [your in­vest­ments] would re­main — but the fact that [this long-term strat­egy] hasn’t worked isn’t strong ev­i­dence that it is to­tally im­pos­si­ble.

Weigh­ing against this model of pa­tience is the pos­si­bil­ity that we re­ally are liv­ing in a spe­cial time. This pos­si­bil­ity was ar­tic­u­lated by the late great philoso­pher Derek Parfit. He said:

1730 Phil Trammell 13

He said that in 2011. And he was talk­ing about the next few cen­turies, so the hinge isn’t very nar­row. It’s not like we have to give money this year be­cause we’re liv­ing at the hinge of his­tory. Maybe the “hinge-iest” year will be in 400 years. It’s worth dis­t­in­guish­ing what we mean when we talk about this hinge.

You might also look back over the past few hun­dred years and ask which mo­ments have been the most pivotal, lead­ing up to the way that we’re cur­rently think­ing about what is best for civ­i­liza­tion go­ing for­ward. Maybe our val­ues would have been very differ­ent if World War II or the Amer­i­can Revolu­tion had gone differ­ently.

John Adams said some­thing similar [when he wrote] the U.S. Con­sti­tu­tion; he fore­saw that what he was do­ing would last for thou­sands of years.

1730 Phil Trammell 14

Below is the same for­mula from be­fore, but now you have this h(t) fac­tor, which scales how much im­pact you have as a func­tion of how much you’re spend­ing at a given time.

1730 Phil Trammell 15
I have no idea what the ac­tual “hingey­ness” sched­ule [of Amer­i­can his­tory] has been, but maybe it looks some­thing like this [see slide be­low]. I’m just illus­trat­ing the con­cept — it’s a con­tin­u­ous ver­sion of the idea of be­ing at the hinge of his­tory.
1730 Phil Trammell 16

We’ll let h fol­low an ar­bi­trary, finite Markov pro­cess, which just means that what­ever state you’re in, there’s some chance that you could move to any of the other states.

1730 Phil Trammell 17

Each has its own in­ter­est rate. Each has its own dis­count rate. And the op­ti­mal spend­ing policy turns out to be the one that satis­fies this for­mula:

1730 Phil Trammell 18
To shed just a lit­tle bit of light on how the “hingey­ness” con­sid­er­a­tion weighs against the [ap­proach of wait­ing to let in­ter­est ac­cu­mu­late], let’s work through the num­bers to de­ter­mine how quickly you should spend if you’re at a mod­er­ately hingey time.

1730 Phil Trammell 19
Triv­ially, if you just ratchet up the h high enough, then you should spend a lot. But let’s look at an in­ter­me­di­ate case. Let’s say the philan­thropic dis­count rate is, to my mind, rel­a­tively high: 1% per year. Let’s say eta equals one. (It turns out that the in­ter­est rate ac­tu­ally doesn’t mat­ter in this case.) And let’s say you have two states — one of which isn’t so hingey (“state 1” or s1), and one of which is 10 times hingeyer (s2). Once you’re in s1, you typ­i­cally stay there; there’s only a 10% chance that you’ll be bumped into s2. And when you’re in s2, there’s an 80% chance you’ll re­vert to s1. So, if you find your­self in s2, you have a lot of mo­ti­va­tion to spend.


1730 Phil Trammell 20
It turns out that in s1, a com­mu­nity of pa­tient philan­thropists should spend 0.3% of its re­sources. And in s2, it should only spend 2.4%. Even in mod­er­ately hingey times, the pa­tient philan­thropist spends lit­tle.

How re­vi­sion­ary is this, re­ally?


1730 Phil Trammell 21
When we’re refer­ring to a com­mu­nity of pa­tient philan­thropists, it’s a neb­u­lous crowd. But let’s think about the effec­tive al­tru­ism (EA) com­mu­nity. There’s roughly $15 billion in EA dol­lars [available to be spent] if you count the Open Philan­thropy Pro­ject (and oth­ers). You also have to count re­sources that aren’t part of ac­tual as­sets. Let’s say there are 10,000 peo­ple in­volved in effec­tive al­tru­ism, earn­ing an av­er­age of $100,000 per year each. And let’s say 10% of that, on av­er­age, is com­mit­ted to EA causes, rather than be­ing con­sumed pri­vately. That would be $100 mil­lion a year, and EA wages would grow at the same rate as ev­ery­one else’s. How­ever, we would dis­count that amount each year to ar­rive at the pre­sent value of a given year’s fu­ture wages, be­cause we’re only con­sid­er­ing the net pre­sent value. In other words, if we could bor­row against all of those wages, how much would we have?

There is a ge­o­met­ric se­ries, with $100 mil­lion di­vided by r minus g, yield­ing $5 billion of net-pre­sent value of EA wages. [When you add this to the $15 billion in cur­rently available fund­ing], that’s per­haps $20 billion. There­fore, if you think that we’re liv­ing in a mod­er­ately hingey time right now — but not a time dur­ing which an event like World War III will de­ter­mine ev­ery­thing that hap­pens af­ter­ward — then we should spend some­thing like $480 mil­lion a year.

The Open Philan­thropy Pro­ject’s grants for the first third of 2019 to­taled $125 mil­lion. Mul­ti­ply that by three and it yields $375 mil­lion — and then there’s ev­ery­one else (ex­clud­ing the Open Philan­thropy Pro­ject).

Per­haps we’re already spend­ing roughly what we would be if we [had de­ter­mined that we’re liv­ing] in a mod­er­ately hingey time. But I don’t think that’s why [we’re spend­ing $480 mil­lion]. I think there are a lot of small donors who are donat­ing 100% of the wages that they plan to give to char­ity as they earn those wages. Mean­while, the Open Philan­thropy Pro­ject is spend­ing more slowly than it plans to in the com­ing decade or two be­cause it’s ca­pac­ity-con­strained with re­spect to re­search an­a­lysts. That’s my un­der­stand­ing. There­fore, I think we might be spend­ing about the right amount, but for the wrong rea­sons.

1730 Phil Trammell 22
One jus­tifi­ca­tion for this could be that we re­ally are liv­ing in a very hingey time, so we should be ratch­et­ing up our spend­ing. Another could be that much of the ex­pen­di­ture [qual­ifies as] what I’m call­ing in­vest­ment [in move­ment-build­ing]; it’s a form of fundrais­ing, like giv­ing to 80,000 Hours, to [re­cruit] more peo­ple who will give later. But that doesn’t seem to be the case if you [look at] what EAs spend their money on.

Another jus­tifi­ca­tion might be that in the fu­ture there’ll be so many peo­ple who are sym­pa­thetic to EA ex­oge­nously, or who are part of a long-term trend of so­cial val­ues mov­ing to­ward EA’s cur­rent val­ues, that we should spend a lot now. In a sense, there’s a lot of low-hang­ing fruit now that will soon be picked.

But if you don’t think that — if you think that we’re cur­rently a bit in­clined to over­spend — where should we go from here?

1730 Phil Trammell 23 One thing we might re­con­sider is earn­ing to give. I know it was pop­u­lar in EA be­fore this in­flux of money came in — par­tic­u­larly from the Open Philan­thropy Pro­ject — and now many peo­ple have de­pri­ori­tized it. But maybe that was too strong of a con­clu­sion to draw. We might also re­con­sider the im­por­tance of fru­gal­ity. Even if there aren’t a lot of seem­ingly great giv­ing op­por­tu­ni­ties right now, you can always just in­vest. And ac­tu­ally, there’s quite a lot of value in do­ing that.

As an in­di­vi­d­ual donor, I think it might be good to con­sider [in­vest­ing in] a donor-ad­vised fund in or­der to re­duce the rate of value drift. Donor-ad­vised funds are tax-ex­empt funds that you can set up for your­self such that you give to the fund, let your money grow tax-free, and then even­tu­ally donate it to a char­i­ta­ble cause. You can’t take the money out for your­self.

I think the fi­nal im­pli­ca­tion is that if we re­ally are plan­ning on these cen­tury-long timescales, we should think hard about how best to de­sign long-last­ing philan­thropic in­sti­tu­tions and co­or­di­nate [how differ­ent in­sti­tu­tions plan their giv­ing] as time pro­ceeds.

Moder­a­tor: Thank you. Does your con­clu­sion mean that EAs should con­sider find­ing good in­vest­ment op­por­tu­ni­ties or [are you recom­mend­ing that they] set up a [new] fund?

Phil: Those both seem like worth­while things to do. I think some­thing that should hap­pen even­tu­ally — and that a lot of peo­ple have ex­pressed in­ter­est in — is [to set up a] fund that ex­plic­itly meets the giv­ing de­sires of peo­ple who think that we should save for a very long time.

But I’m not re­ally recom­mend­ing that. I’m just say­ing that if you con­ceive of all EA ac­tivity as this fund over time, maybe it should con­sider al­lo­cat­ing more to efforts that in­volve sav­ing for a long time. Does that make sense?

Moder­a­tor: Okay. To fol­low up, how do you take into ac­count the fact that [if some­one in­vests for the long term], other peo­ple or agents will be the ones ac­tu­ally spend­ing [their] money?

Phil: Yeah, that [cre­ates risk from] in­sti­tu­tional value drift. If we think hard about it, we can lower the risk of that. We’ll never [re­duce the risk] to zero. But one thing you might do is that, in­stead of en­trust­ing this gi­ant fund to one per­son and hav­ing them hand it down through the gen­er­a­tions, you could set up a com­mit­tee of peo­ple. They might vote on whom to swap out ev­ery 10 or 20 years.

You could also look at other in­sti­tu­tions — for ex­am­ple, re­li­gious in­sti­tu­tions that have been faith­ful to their val­ues over time — and see what might be learned from them. It is a big em­piri­cal ques­tion. I haven’t thought about it as much as I would like to.

Moder­a­tor: Thank you. There is a clar­ify­ing ques­tion from the au­di­ence: Can you ex­plain what you said about buy­ing land and rent­ing it?

Phil: You can buy a patch of land and then sell a 100-year lease on it. You give some­one else the right to do what they like with it for the first 100 years, and you’ll be paid about 90% of what you paid to pur­chase the land. In a sense, you just paid 10% of the price of the land for the right to say what hap­pens to it [af­ter 100 years have passed].

Moder­a­tor: All right. Thank you so much, Phil.

No comments.