Against Loneliness: Cause Exploration

Cross-posted from Cold Button Issues. Thanks to Dominic Roser for comments and suggestions.

Americans are born into smaller families than before, with fewer brothers, sisters, and cousins. They’re less likely to marry and less likely to have children. Membership in churches and other religious organizations is in decline. Fraternal organizations are emptying out. Many Americans will live increasingly lonely lives and die alone.

The costs of loneliness, to the extent they can be measured, are immense. Based on estimates of the psychological benefits of marriage, it’s plausible that the decline in marriage rates since 2000 produce a loss equivalent to $1 trillion annually in the United States.

Support for stronger, earlier marriages and social organizations could help stem the tide of loneliness. The most straightforward way to work on this would be to back political advocacy to modify the federal tax and welfare system to be more favorable to marriage. Some other interventions I view as promising include seeking regulatory privileges for fraternal organizations, and funding research to evaluate the effect of premarital counseling, marital counseling, and pornography consumption on marital satisfaction.

How bad is loneliness?

Social isolation is associated with all kinds of ills, including poor health outcomes. The National Institute on Aging, part of the NIH, summarizes the findings, stating “[r]esearch has linked social isolation and loneliness to higher risks for a variety of physical and mental conditions: high blood pressure, heart disease, obesity, a weakened immune system, anxiety, depression, cognitive decline, Alzheimer’s disease, and even death.” Luanaigh and Lawlor (2008) offer their own review on the effects of loneliness in older people and find that loneliness contributes to depression, mental decline, impaired  sleep, blood pressure, and immune response. Luo et al (2012) found that feelings of loneliness were associated with higher levels of mortality, even when controls for obvious confounders were included. 

Obviously, a serious downside of much of this research is that it relies on observational studies. There aren’t large scale studies where people are purposely isolated from their social network that allow us to better estimate the direct effects of loneliness on health. But it is possible to have experiments or interventions that try to reduce loneliness that show positive effects of increased social interaction. Pitkala et al (2009) use a randomized controlled trial to show that seniors at adult day care centers who were enrolled in groups that fostered socialization and shared activities  were 7 percentage points less likely to die over the course of the next two years and reported better subjective health. The costs of this program were 881 €/person and  the cost of the program per participant was more than offset by a decline in healthcare consumption costs. This implies that elderly people who use adult day care centers could  gain another year of life for about $12,000. While this may compare favorably to many health interventions in the developed world, it does not compare favorably to interventions favored by GiveWell in developing economies. 

Americans have weaker social ties than in the past.. Americans now have fewer friends - one in ten Americans report having no close friends outside their family.

On average, Americans spend only half an hour socializing per day, compared to three hours spent watching television.

Special attention should be paid to the decline of marriage. Marriage appears to be hugely beneficial. Married people are consistently shown to be happier than unmarried people, across countries (Diener et al 2000; Stack and Eshelman1998). Is this a selection effect- are happier people more likely to marry or people who are poised to become happy just more likely to marry right before their happiness would have soared anyways? There isn’t conclusive proof, but marriage probably does cause happiness to increase. Tao (2019) shows that in Taiwan, the soon-to-be married don’t have a happiness advantage but the married do.In Germany, differences in happiness do predate marriage but selection effects do not explain the entire happiness gap between married and single people (Stutzer and Frey 2004).

Blanchflower and Oswald (2004) use widowhood as a natural experiment, and show that in terms of happiness, marriage is worth $100K of additional income annually. In 2000, the Census reports that 54% of those over 15 years of age in America were married. That fell to 48% in 2020. With approximately 266 million Americans over the age of 15, there are 16 million more Americans unmarried than there would have been if marriage rates stayed constant since 2000. If we accept Blanchflower and Oswald’s estimate, this decline in marriage would require $1.6 trillion annually in additional to offset the fall in happiness. Of course likely such a naive estimate overstates the value of marriage if one makes the reasonable assumption that those who would benefit the most from marriage are most likely to get married, and those who benefit the least from marriage are least likely to get married. Still, even if we assumed that “missing marriages” due to the decline in the marriage rate are only worth $10,000 annually, the United States is still losing $160 billion of value annually. 

Who else cares about this?

There are people who care about this and there are no organizations that are openly pro-loneliness. The straightforwardly named Cost of Loneliness Project “seeks to provide a unifying voice to recognize and combat loneliness across all demographics. It is our current healthcare imperative.” However, the group appears to not have been active since 2018. In the United Kingdom, loneliness is now part of the ministerial portfolio of the Minister for Sport and and Civil Society.

There are also organizations that exist to reduce social isolation for specific groups, frequently the elderly. Pets for the Elderly provides animal companions to elderly Americans so they can have animal companionship. Many collegiate sororities like Sigma Kappa promote volunteering to visit with the elderly, and so on.

While religious organizations exist for reasons beyond promoting social ties and alleviating loneliness, that is one of their functions. They nurture and provide relationships for their members and many also engage in evangelism to outsiders, offering social ties and relationship  benefits to those who join. That said membership in religious organizations is in precipitous decline. 

Few people seem to be directly working on reducing American loneliness in general, at scale or via policy interventions. 

Possible Interventions

There are many possible anti-loneliness interventions. I primarily consider ways to strengthen marriages and romantic relationships as they are the central relationship in many people’s lives. I also discuss ways to strengthen friendships via fraternal organizations.

Marriage Incentives and Disincentives

Governments can encourage or discourage marriage, intentionally or inadvertently. Rewriting tax laws and eligibility requirements for certain welfare programs could make marriage more attractive. Philanthropists could support political advocacy to rework the US tax code and welfare system to be more friendly to married couples, thereby encouraging more people to get married or stay married.

The so-called marriage penalty refers to cases where couples are financially disadvantaged by the tax code or other government policies for being married. A marriage bonus is where a couple is rewarded for being married. One well-known cause of a marriage penalty is that single parents are able to claim a larger standard deduction by filing as “head of household” yet if they marry the other parent of their children, they are no longer able to claim such a deduction. Marriage penalties are more likely to occur in couples where each member makes a similar income. Couples can also be discouraged from marrying when their joint incomes would make them ineligible for government subsidies, or for smaller government subsidies.

Ilin, Kotlikoff, and Pitts (2022) estimate the effect of the “marriage tax” on marriage rates. They define the marriage tax to  “include these income tax penalties and transfer claw-backs associated with marriage but also expands it to include the impact of these marriage penalties on remaining lifetime net taxes and, thus, remaining lifetime spending.” They conclude that for women with children a “ one percentage point increase in the marriage tax rate decreases the probability of marrying in 2018 by 3.69 percentage points” with smaller effects on other groups. Revising tax and welfare codes to better support married couples could encourage marriage.

Senator Mitt Romney recently proposed a bill that would remove certain marriage penalties. Encouraging marriage is well within the US Overton window and there are already politicians such as Senator Romney that support the goal.

Overall, I am the most confident about this proposal. There is reasonable evidence that marriage can be encouraged or discouraged via government policy. There are high-profile American political figures who would like to use government policy to increase marriage and the apparent benefits of marriage are quite large. If a political advocacy campaign could expect to undo 1% of the decline in marriage rates since 1990 and the estimated annual value of marriage is discounted to $10,000 annually, that advocacy campaign would generate value worth $1.6 billion each year. If the effects of the campaign lasted for a decade, it would generate $16 billion in value.

Reduce Wedding Costs

Generally, if something is cheaper, more of it is consumed. If the cost of weddings fell in the United States, would there be more weddings in the United States?

Several jurisdictions have tried to boost marriage rates (and fertility rates) by discouraging expensive weddings. Korea has run ads discouraging conspicuous consumption related to weddings. A Chinese city has offered to reimburse some of the the cost of weddings.

In the United States, there is a negative correlation  between wedding costs (as measured by the Knot) and marriage rates, both with and without adjusting for the state median income (Alaska and the Dakotas are excluded due to  lack of statewide data from the Knot). Hawaii and Nevada are excluded for extremely high marriage rates presumably due to destination weddings.) 

Although there is a negative  correlation it is not that strong (-0.16 when controlling for state income) and it’s very likely the causation could run the other way. In states where marriage is prized, couples may marry young where they have less income, keeping marriage rates high and wedding costs down.

Some researchers have argued that marriage costs, including wedding costs, can deter marriage. Ragab and Saad (2022) show that negative economic shocks to men in the West Bank reduced marriage rates, in part due to strict cultural expectations about proper weddings. Dhillon (2008) argued that this is a problem across the Middle East in general. Smock, Manning and Porter (2005) interviewed unmarried couples who lived together in the United States and found that a fifth of the couples said that they were putting off marriage until they could afford a wedding.

I would recommend philanthropists interested in the problem of high wedding costs to pay for additional research on this topic, to both get better data on wedding spending in the United States and to examine its relationship with marriage rates.

Are there ways to lower the costs of weddings? There are no precedents of clearly successful initiatives to lower the cost of weddings across an entire nation. That said, it might be possible to try to shift cultural norms in favor of affordable weddings or to directly or indirectly provide affordable weddings.

Cultural interventions such as paying Youtubers, bloggers, TikTok stars, and other influencers to promote budget weddings is another possibility. Current bridal magazines offer a wide variety of ways to spend more money. There are some popular sites that encourage nontraditional and often more affordable weddings, but often encourage couples to plan expressive and individualistic weddings that require substantial time investment. Ideally, we could encourage weddings that are both financially affordable and standardized enough to limit stress and unpaid labor on behalf of the couple, their families, and their friends.

American wedding planning sites generally state that most of the cost of a wedding comes from the reception venue and catering. Buying desirable wedding venues and offering very cheap prices to hold weddings could drive down prices. Forming a wedding planning company that offers standardized and cheap wedding packages could also make weddings more accessible.

This could be done by nonprofits, by B corporations, or perhaps even by a standard for-profit firm that a philanthropist could invest in. In Korea, a steel company is trying to promote affordable weddings to its employees by offering the use of low-cost wedding venues. 

Premarital and Marital Counseling

Some couples seek premarital counseling before their wedding. Some couples seek marriage or relationship counseling during their marriage. Does such counseling work? Probably to some extent, but the evidence is not particularly strong. German researchers tried to survey clients undergoing counseling to see if their relationships improved- they did or said they did but most of their subjects dropped out of the project (Hahlweg and Klann 1997). A RCT in Iran showed that couples struggling with fertility appeared to be helped  by couples counseling (Vizheh et al 2013). The American Association for Marriage and Family Therapy reports that 98% of marriage and family therapy clients  think their counselors are good or excellent. In a meta-review, Hawkins et al (2008) summarize the evidence on the  efficacy of providing marriage and relationship education by saying counseling ”produces modest but reliable effects comparable with those of other psychoeducational interventions of interest to policy makers. However, the question of efficacy for more diverse and disadvantaged remains an important area for research that will inform practitioners and policy makers.” A recent Canadian review of the effectiveness of couples therapy found results “were mixed.”

For premarital counseling, there is some evidence that it’s helpful but hardly conclusive. Couples who go through premarital counseling score higher on a test of marriage preparedness that predicts marital satisfaction (McGeorge and Carlson 2006). In a very small-scale survey, Kepler (2015) finds a positive but significant relationship between premarital counseling and marital satisfaction. 

A philanthropist could fund research to better estimate the effects of premarital counseling on marriage duration, relationship satisfaction, number of children, frequency of sexual intercourse, symptoms of anxiety and depression, and so on. A RCT could directly provide premarital counseling to a treatment group. A philanthropist could also seek to partner with the Roman Catholic Church in the United States. In 2020, there were 100K Catholic weddings out of 1.7 million total weddings. Couples marrying within the Catholic Church are expected to participate in church-based preparation programs that include premarital counseling, often known as pre-Cana. Perhaps it would be possible to offer a grant to the United States Conference of Catholic Bishops to evaluate the different pre-Cana programs offered across dioceses to see what practices are associated with good marriage outcomes. 

Marital counseling could be supported and assessed through similar means.

Reduce Consumption of Pornography?

Sixteen US states have declared that pornography consumption is a public health crisis. There are multiple subreddits where people try to deal with the harmsof porn, often discussing its impact on relationships. NPR reports that pornography use causes divorce.

What evidence is there that porn consumption is harmful? Fight The New Drug, one of the major anti-porn groups, offers a selection of peer-reviewed studies that claim pornography is harmful. There are many studies that show a correlation between all kinds of undesirable behavior and outcomes and porn consumption. Wright, Tokunaga, and Kraus (2016) offer a metareview which shows that consumption of pornography is associated with sexual aggression. Skoda and Pedersen (2019) find that porn consumption is linked to lower self-esteem in men. Its consumption is linked to lower marital satisfaction, especially among religious couples (Perry, 2016).

Probably the best evidence for the claim that pornography is bad for relationships comes from Perry and Schleifer (2018) who use a longitudinal study of American couples and shows that couples where one member begins using pornography roughly doubles the risk of divorce, while couples where porn use ceases see a decline in their divorce risk. 

What is missing is evidence that pornography consumption causes or increases these negative outcomes behavior. Ley (2018) criticizes states who called porn consumption a public health crisis by arguing that the links between porn consumption and negative outcomes fail to consider reverse causality and spurious correlation.

I don’t think the current evidence that pornography is harmful for relationships is conclusive or even especially convincing, but it is a plausible claim. (Of course there are other powerfully made objections to porn- its production may be exploitative and abusive, it may treat people as means rather than ends, it may immorally incite lust and violate religious teachings, etc…) But there is a politically and culturally influential anti-porn movement. If it could be convincingly shown that consumption of pornography harmed relationships that would find a receptive audience and many potential allies.

For this reason, I think sponsoring further research on the potential harms of pornography consumption is worthwhile. If it could be more firmly established that pornography consumption has clear harms, it would strengthen the current anti-porn movement.

Current research on pornography heavily depends on observational data. Experimental data would be very valuable. For instance, you could pay a treatment group to install “Covenant Eyes” or some other anti-porn software on their personal devices and see if relationship quality improved following this intervention. Alternatively, a treatment group could be offered gift certificates to popular porn websites and changes in relationship quality could be observed, although this would only work if pornography consumption were limited by access to monetized sites and might have trouble getting IRB approval.

Fraternal Organizations

Fraternal organizations, clubs, service societies and so forth have a long history in the United States. Examples include Freemasons, The Elks, the Moose, and Lions Club International. Some exist primarily as social clubs, others place greater emphasis on philanthropic service, and others use rituals to convey metaphysical teachings. Whatever the specific emphasis of a fraternal organization, these groups are collapsing and have been for decades. Putnam (1995) writes: 

“Fraternal organizations have also witnessed a substantial drop in membership during the 1980s and 1990s. Membership is down significantly in such groups as the Lions (off 12 percent since 1983), the Elks (off 18 percent since 1979), the Shriners (off 27 percent since 1979), the Jaycees (off 44 percent since 1979), and the Masons (down 39 percent since 1959). In sum, after expanding steadily throughout most of this century, many major civic organizations have experienced a sudden, substantial, and nearly simultaneous decline in membership over the last decade or two.”

This trend has continued. Is it possible to revive these fraternal organizations and if so how?

One way to support fraternal organizations at scale would be to liberalize regulation of alcohol and tobacco consumption on fraternity property and at fraternity events for members. In the United Kingdom, “working men’s clubs” suffered a decline in membership and attendance after smoking was prohibited in their venues. Marlow (2010) suggests that smoking bans are also harmful to American fraternal organizations (note that Marlow acknowledges previous research funding from Phillip Morris). 

Access to alcohol is also very important to fraternal organizations. Some US states made the ability of fraternal organizations to sell alcohol at club-owned properties contingent on allowing women to be members, an action upheld by the US Supreme Court, thus prompting some fraternal organizations to go co-ed. Perhaps states could spark a revival in fraternal organization membership by waiving taxes on alcoholic beverages sold to members at club events and venues.

Another way to increase membership would be to hire a management consulting firm to provide turn-around advice for declining fraternal organizations or offer grants to organizations that have a plausible pilot program for rapid growth. A philanthropist who thinks revitalized fraternal organizations would be valuable, but thinks spending money on this cause would be fruitless, could offer a prize for fraternal organizations that meet certain benchmarks in terms of growth or membership engagement.

Conclusion

. The most dramatic reductions in loneliness, in my opinion, would be driven by increases in marriage rates, fertility rates, and membership in religious organizations. The most neglected and novel intervention I have identified is support for fraternal organizations.

A robust lobbying effort aimed at removing federal “marriage penalties” and creating marriage incentives would be cost-effective in expected value terms even if the odds of success are quite low.  The value and cost-effectiveness of other proposed interventions- research on the efficacy of premarital and marital counseling, research on the effects of porn consumption on marriage quality, support for fraternal organizations- is hard to quantify but are areas that are currently extremely neglected.

That said there are many other ways that seem promising, but fall outside the scope of this essay. Raising the birth rate would provide more people with children and grandchildren, nieces and nephews, brothers and sisters. Better urban planning could mean shorter commutes and more time for relationships. Cheaper housing might mean families and friends could live closer together. There’s also more research that could be done on the effects of social media.

Loneliness is pretty common and pretty bad. But it’s possible to change that, at least a little.