Economic Policies that Optimize for Future People

BLUF: This isn’t a profoundly deep post– it just shows my current, general views on a variety of current economic issues.

I do not believe future people are intrinsically less valuable than the people existing today. In fact, I think they might be more valuable because their lives will intrinsically be more worth living as their well-being will probably be greater. I also respect the validity of the 20% chance of extinction by 2100 that is the average of a number of researchers’ estimations, so I think that even considering a variety of extinction scenarios, there are many more future people expected to exist in the future. Thus, I think we should optimize our political and economic policies to serve their interests even more than the selfish interests of those people alive today. What does this look like in concrete policies?

  • Steep carbon taxes
  • Taxes on essential ecological service destruction in general priced at the cost of replacement
  • A land-value tax
  • Universal Basic Income at a cost to less-efficient government social programs like Social Security

Principles behind optimizing our economy for the long-term future

  • A willingness to bear the temporary economic losses, as a society, of implementing steep carbon taxes and essential ecological service destruction taxes.
  • More deliberate experimentation to test policies via states and charter cities.
  • Beyond concerns about environmental destruction, being willing to optimize for economic growth more than redistributing resources to satisfy the preferences of everyone who happens to be alive today. Social security recipients are no longer actively contributing to the economy, so we should cut their funding to give everyone a UBI.

Beyond optimizing for the long-term, I generally support:

  • lifting economically stifling regulation; we should make entrepreneurship as easy as possible. One shouldn’t need to consult a lawyer to start many personal businesses.
  • lifting barriers to competition like government-mandated licensing (e.g. taxicabs)
  • Free trade
  • Much greater immigration, especially of educated people, but not quite open borders.

The DoD can Further Optimize its Retirement System

BLUF: The DoD could probably save money, increase servicemember compensation, and better optimize talent management if it further increased TSP contributions while reducing 20-year retirement pensions.

Introduction to the Military Retirement: For decades, the DoD had a retirement system where after 20 years of active-duty service (and upon leaving the service) a servicemember (SM) was entitled to an immediate annuity based on years of service and basic pay using a 2.5% multiplier. For instance, after 24 years of service, a SM essentially earned 2.5% * 24 = 48% of their base pay for the rest of their life.  In 2016, Congress created a new system called the Blended Retirement System (BRS) that reduces the base pay multiplier to 2.0% but adds two new components: 5% matching Thrift Savings Plan (TSP) contributions and continuation pay (CP) (a few month’s pay for signing and continued service obligation for 8-12 year SMs).

The traditional 20-year retirement model (T20R) served as a powerful retention tool.  It takes a lot of time and resources to develop a senior Non-Commissioned Officer (NCO) or a field grade officer, and the T20R is a strong incentive for retaining talent who often have lucrative civilian career options. The problem with the T20R is that it is expensive to the Department of Defense (DoD), accounting for about $52 billion of its budget annually.[1]

The BRS was developed in the interest of saving money while also strengthening talent management.  Costs are initially higher under the BRS because of expenditures for CP and TSP matching contributions, but costs will eventually lower due to the lowered retirement pensions.[2]

The BRS promotes talent management because it improves incentivizes for short-term and mid-term service as well as long-term, 20+ year service.  Some individuals who would be valuable assets to the DoD are not necessarily careerists, and matching TSP-contributions is an incentive to attract them. Additionally, too many servicemembers tend to leave at 20 years; the BRS would better incentivize additional service.

My Argument: The military/Congress decided on a matching contributions rate of only up to 5% of base pay, but I argue that a higher percent of matching contributions, such as 10%, can further save the DoD money over its T20R system, increase average SM compensation, and increase talent management by offering more optimal incentives for SM retention.

Current Retirement System Example: A typical compensation scenario under the traditional retirement system: If an officer retires at a paygrade of O-5 at 20 years of service, it will cost the DoD an additional $10,278 per year compared to the BRS model which has a pension set at 40% of final base pay.  With future improvements in medical care, 20-year retirees (approximately 42 years old) might very well live up into their 90s, which may be 50 years after their retirement. Under the T20R, this 50-year pension would cost the DoD $2,569,530, and under the BRS, this would cost $2,055,624. Out of this $513,906.00 in savings comes the TSP matching contributions and continuation service pay in the new BRS.

Blended Retirement System Example:  Under the BRS, the military plans to match up to 5% of a SM’s contribution to their TSP. Over a 20-year career where a SM chose to donate at least 5% of their base pay to the TSP, this would cost the DoD $75,573.80 over the 20 years.  Continuation pay, an additional retention incentive, costs the service 2.5*base pay. For an O-4 with 12 years in service, this would cost the DoD $18,075.45. Under the new BRS, the military is not spending approximately $513,906 due to the difference in pensions (at 50 years after retirement), while additional expenditures for a 20-year officer retiree would generally only amount to $93,649.254.  Just as the DoD is saving $420,256.74, SMs who transfer to the BRS lose approximately this much income over their lifetime due to smaller monthly pensions (assuming they would have donated the TSP matching contributions themselves).

Proposal: While the DoD ought to save money and cut costs, I argue that it is even more important to offer competitive incentives for the sake of talent management and retention.  To illustrate a point, if the military were to maximize TSP contributions (limit of $18,500) for a 20-year career officer, this would cost the DoD $370,000. This is still less than the $420,256 in savings the DoD makes under the new BRS. This would allow the SM to open up a private investment account like a Roth IRA, which one would expect to match the returns of the TSP.  While the DoD would still be saving $50,256, the SM still investing $18,500 a year could now reasonably expect to earn an additional $550,894 over 20 years assuming a market growth of 4% over inflation. If the SM did not touch this money and stopped contributing to the Roth IRA after retiring at 20 years TIS, they could expect to have both their TSP and Roth IRA each worth $3,915,032 after 50 years.

This is an extreme argument to exemplify a point: the DoD can save money and increase SM compensation by utilizing market forces and compounding interest.  Only about 17% of SMs end up serving 20 years and retiring [3], so it would be very costly to the DoD if it were paying everyone an additional $18,500 a year for their TSPs. The military, however, is interested in increasing short-term service incentives for talented individuals who know a 20-year career is not for them.  The military also may be interested in reducing the incentive for mid-career (8-15 years) SMs to force themselves to continue service (out of fear of walking away with no retirement benefits; this the military does not want to retain members who really are ready to go). Additionally, the military is interested in reducing the overly strong incentive for talented senior SMs to get out shortly after 20 years of service.[3] See Appendix A for a chart of officer retention.[3]

To balance these incentives and the cost behind them, the DoD can increase matching TSP contributions to, say, 10%, while paying for this through reducing the pension percentage of base pay after 20 years to say, 35%.  For an officer who allots the full $18,500, and would allot the extra 10% of their Base Pay saved by the DoD’s 10% matching contributions, this would yield an extra $55,089.45 after 20 years.

According to my model (detailed below), a SM will always have less total compensation than the T20R until after 24 years after retirement at 20 years (assuming a RR < .5).  After 24 years, BRS retirees will see their total compensation surpass that off T20R retirees. If a BRS retiree is most interested in maximizing their lifetime compensation (perhaps for the sake of setting up a trust for a good cause) and they are willing to bear the lower pension up until their 24th year after retirement (approximately 66 years old), they are actually best off with a lower RR.  The best RR for that frame of mind is actually 0.

Model

Problem: Find an optimal TSP matching contributions rate and pension rate (as a final percentage of final base pay) to maximize lifetime earnings of service-members (SMs) without increasing costs to the DoD.

Variables:
        RR, the new 20 year retirement rate to optimize ;
LSEn, lifetime service earnings at year n ;
        TBPn, total base pay, the summation of base pay earned in all ranks prior to year n ;
        BPn, base pay at year n ;
        TCn=MCR*BPn ;
        RDn, retirement pay difference n years after being retired ;
        TSPn, value of TSP account at year n after starting account ;
        TSPr, value of TSP at exit from government service ;
       Yr, year of military retirement ;
        Yc, year of cessation of benefits ;
        n, years (of service or retirement) ;

Parameters:
         SMBRS, the number of service members (SMs) in the BR ;
         I, inflation rate of the US dollar, assumed to be 2% ;
         MRR, market return rate and rate a balanced TSP portfolio is predicted to grow.  For a particular solution, I assume 6% annual growth (1.06) ;
         MCR, the BRS matching TSP contribution rate ;
         AYURD, Average number of years lived until from retirement to death

Assumptions: For simplicity’s sake, I assume the TSP index fund growth will remain constant.  I also ignore future rises in the cost-of-living with the presumption that the military will continue to make cost-of-living-adjustments to its pay and retirement system.  I am also speculating an increased rate of SM retirement from 17% to 20%.

Model
General model for lifetime service earnings at year n due to TSP account growth:
             LSEn= TBPn – n * TCn + TCn * ( 1 – MRRn ) / ( 1 – MRR )

 Model for TSP account growth after retirement:
            TSPn = TSP*  MRR * ( Yd – Yr )

Additional cost of the BRS to the DoD compared to T20R for an individual SM:
           MCR * BP* n + 2.5 * BP12

Savings to the DoD compared to T20R:
           ( .5 – RR ) * BPr * 12 * ( Y– Y)

Cost of BRS:
           ( MCR * BP* n + 2.5 * BP12 ) * SMBRS( .5 – RR ) * BPr * 12 * AYURD

The general solution to find the optimal solution is:
RD(RR, n) = 6BPr + 12nRRBPr – 6nBPr -(3nRRBPry(1 – nMRR))/(25*(1 – MRR))

The particular solution:
RD(RR,n)= 13896nn + 129504nRR – 64752.2n – RR 27792nn

Importantly, we should set realistic constraints on RR and n:
           .2 < RR < .5   ;   0 < n < 50

The constraints are chosen because the markets could fail and SMs should still have a safety net of 20% of their final base pay.  A 50-year cap is chosen because of the assumption that retirees will want to eventually utilize their TSP accounts (at this age, perhaps setting up a trust).

Rather than try to save the DoD money in this model, I am routing all the money from the savings from the reduced pension payouts back into the TSP matching contributions for all service members. Because there are now 5 SMs to provide matching TSP contributions for instead of just 1 SM to provide a pension for, the savings from a reduced payout will be divided by 5, and this will be the amount of money the DoD can pay out to each SM over their careers.

Normally, rank determines compensation, but to simplify this model, everyone will receive the same amount of matching contributions over their careers. Thus, this model divides the amount of TSP matching contribution money available to each SM by 20 and provides this as the maximum amount of matching contributions available to each SM every year.

See Appendix B for an Excel sheet that uses Solver to optimize for the pension and matching contributions rate.  What we find is that the function is non-convex in the domain defined by our constraints, so there are no relative maxima.  There is only an optimum RR for any given n (and vice versa).

Further Discussion: If one is willing to wait over 24 years after retiring to touch their TSP accounts, it seems that my current model is better than the T20R. However, what I have failed to account for is how much money extra money a SM retired under the T20R could invest in private investment accounts.  This amount is modeled by the expression:

(.5 – M) * BP* (1 – MRR*n) / (1 – MRR)

For example, an O-5 SM retiring under T20R and investing the difference between the two pensions could expect to have more money regardless of RR (as long as it is the same as the TSP) up until year 36.  The BRS only surpasses 36 years after military retirement due to the 20 years of earlier TSP contributions that would finally catch up and make the difference.

While it takes a long time for lifetime compensation for my modeled BRS SMs to surpass the T20R SMs, the benefit of this really seen elsewhere.  Now, every SM, regardless of how long they stay in the DoD, has a retirement benefit. Not only does this help millions of individuals financially, but it also will considerably promote talent management.

Appendix A

appendix2BRS

Appendix B

appendix3BRS

Link to Google Sheets File

Sources

[1] https://budget.house.gov/hbc-publication/364048/

[2] https://www.rand.org/pubs/research_reports/RR1373.html

[3] https://warontherocks.com/2015/03/military-retirement-too-sweet-a-deal

Risk-Tolerance is Not a Thing

Traditional finance literature says that individuals may choose their risk-tolerance in order to adequately model their financial preferences. A person who chooses a high-risk financial option is willing to deal with greater variability; a person who chooses a low-risk option prefers outcomes with low-variability.

Underlying these notions is the idea that risk-tolerance is a stand-alone, unique, irreplaceable concept.

I like to compress ideas to their simplest form, and I think we can model individual financial preferences without the notion of risk-tolerance:

Risk-tolerance can be wholy framed as prefering a utility function with a particular curve of marginal diminishing utility.  There is no need for extra calculus beyond standard von Neumann-Morgenstern (vNM) rationality, which involves acting so as to maximize expected utility (considering one’s utility function). Financial modeling is as simple as assigning an amount of utility per dollar over the range of the real numbers, and acting so as to maximize one’s expected utility.

If one appropriately updates their utility function, they don’t have to deviate from this simple application of vNM.  If my preference is to definitely make $9,000 over a 1% chance of $1 million, you could just say I give greater utility to the first $9k, less utility per dollar for higher amounts of money, and I still should just act so as to maximize expected utility as per vNM.

I am not saying the use of the term “risk-tolerance” has to go, but let’s not think it’s this special thing when it is not.

Is Life Insurance Rational?

BLUF: Life insurance might be rational for many effective altruism-minded people, but the devil is in the details- that is, one’s probable risk of death and the payout/cost of a plan.

US military servicemembers are afforded the option of paying $29 a month for $400,000 in life insurance. The norm is that the overwhelming majority of servicemembers pay for this. I decided that if I happen to die while in the military, I wanted the $400k to go to the most effective charities. The separate $100k death gratuity would go to my family members, but I would divide up the $400k payout among four think-tanks/charities I care about:
the Machine Intelligence Research Institute,
80,000 Hours,
GiveDirectly
the Future of Humanity Institute at Oxford.

While this may not be the optimal way of distributing funds and there is an argument that in a sufficiently-sized donation market, it’s optimal for people to just donate to the one charity they think is the best, I still hate the idea of putting all my eggs in one basket. Additionally, if I die, I expect my donations to go on the local news or at least be on my obituary, and I think it would be better for people to see four organizations than one. Hence, I decided to focus on the four cause areas of AI-alignment (MIRI), meta- effective altruism (80,000 Hours), direct impact on global poverty (GiveDirectly), and general existential risk research (FHI).

Are my odds of death good enough to warrant this $29 a month that cannot go to these charities? I assumed my odds of death were good enough when I was an enlisted deep sea diver (laundry list of diving hazards for the uninitiated). I feel they are still decent enough as a West Point Triathlete who frequently bikes on terrible New York roads 8+ months of the year to warrant the $29 a month. However, I recently realized that EA is probably big enough now that I don’t need to make sure that these organizations get a minimum amount of money from me, but that the community is big enough now that I should just maximize the expected utility of my donations. In other words, I might actually be wasting a portion of $29 a month depending on my actual risk of death. Thus, I figured it was time to just shut up and multiply:

It turns out the calculus really doesn’t have to be that hard. $29 a month * 12 months a year * 7.3 more years in the Army= $2540. $400k/2540= 157. Essentially, if I have greater than a 1 in 157 chance in dying over my expected time remaining in the Army, it makes sense that I buy the life insurance. I think this is definitely within a natural log order-of-magnitude of the true probability, which is between a 1 in 55 chance and a 1 in 403 chance (ln157=5 ; e^4= 55 ; e^6= 403). Generic actuarial data supports this (average odds of dying while a 25 year old male (my average age over the next 7 years) is 0.001451. This yields a 1 in 94 chance of dying, which is riskier than 1 in 157. Additionally, while I may not be doing the risky driving which fuels deaths for people my age, I think the road cycling and risks while on deployment (assuming we don’t go to war with a near-peer) will keep my number at least in the ballpark of the national average. Hence, it seems rational for me and others in the military to buy life insurance if we want to maximize expected dollars donated to the causes we care about.

Why is SGLI so cheap if the expected utility works out for most people? This could signal my math is faulty.  Contributing factors include:
-SGLI is run by the Department of Veteran Affairs and they are not making a profit.
-they probably make a lot of interest from investing the money.

Does my math check out? Thanks for reading!

Possible Moral Trade Implementation

I’ve been thinking about Toby Ord’s Moral Trade paper, and think a new Repledge website is a desirable thing, legal questions aside. Here’s the idea (edited with my own takes) for those unfamiliar:

Create a website where people can donate to a cause, but where if someone else donates to the opposite cause, both peoples’ money is instead diverted to a 3rd charity that both parties also support (e.g. GiveWell). To discourage GiveWell supporters from waiting and donating to whatever interest group is necessary to double their donations, the running balance is kept private. After a set time (say, once a week, Saturday at 1800), the tied money goes to GiveWell and the surplus money goes to the interest group it was intended for.

People interested in supporting interest groups should be interested in funding this way if:
1) they believe their opponent’s interest group could advance their interest better with a dollar than their own
2) they would rather give $2 to GiveWell than $.5001 to their own interest group
3) some reconciliation of #1 and #2.

Trust problems can be resolved with smart open-source software and 3rd party (not GiveWell) auditing.

Given only the option of donating X dollars through the site or outside it, I think a rational agent should donate according to the following procedure so as to maximize utility:

uA= a utility/$ ———utility per dollar of one’s interest group
uB= -b utility/$——–negative utility per dollar of the opposition’s interest group
uG= g utility/$ ———utility per dollar of the neutral 3rd group (GiveWell)

If abs(uB)>uA:
Donate through site so as to get 2uG
If uA>2uG:
Donate directly to A
uD=uA-uB
If uD>2uG:
Donate directly to A
Else:
Donate so as to get 2uG

If this is a good idea in theory, the next obstacle to tackle is the question of legality.  I imagine that people should be able to consent to their money being used in this way, but laws, especially campaign finance laws, are not always intuitive.

The next question is whether the expected donations to GiveWell would be worth the effort to tackle this project. The effort, of course, could widely vary; we could hire a team of software engineers to build a secure system where humans are effectively out of the loop and this could be verified by 3rd party investigators. Or we could make two Venmo-like accounts (one for each side of a partisan issue that a poll shows people are interested in funding on both sides), and literally just live stream and post a video weekly of the site’s owner subtracting the difference between pairs of accounts, donating the money to the winning site (with the camera still rolling), and donating the matched money to GiveWell.

There is a very good chance that we will not find prospective donators on opposite sides of an issue that both buy into the calculus and trust the site enough, but it’s possible.  The cost is low enough however that this simpler system could be implemented within hours by one trusted third party should a community find itself sharply divided on an issue and be willing or already spending money on organizations with opposing missions.

Thanks for reading! I would love your feedback 🙂