Impact of legalization of same-sex marriage on the new HIV infection rate in 172 countries around the world based on difference-in-differences (DID) method

doi:10.21203/rs.3.rs-2105087/v1

Download PDF

Research Article

Impact of legalization of same-sex marriage on the new HIV infection rate in 172 countries around the world based on difference-in-differences (DID) method

https://doi.org/10.21203/rs.3.rs-2105087/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Acquired immunodeficiency syndrome is a major global health challenge, which is exceedingly harmful to human health. This study aimed to explore the impact of legalization of same-sex marriage on the new HIV infection rate in countries around the world.

Methods: The study is conducted on 172 countries whose observed data were collected from 1996 to 2019 in an open-source website. To evaluate the effect of legalization of same-sex marriage, difference of differences (DID) method was applied. The treatment effect was reported in static DID model, and the time-varied treatment effect and its 95% CI were reported in dynamic DID model.

Results: In the static DID model, we found that compared with countries without legalization of homosexuality, countries with legalization of same-sex marriage had an increase of 0.129‰ (95%CI: [-0.234,0.491]) in the new HIV infection rate, but the result was not statistically significant. In the dynamic DID model from 1996 to 2010, we found that the fifth year after the implementation of same-sex marriage legalization policy significantly increased the new HIV infection rate by 0.7‰. But in other years the effect are not significant after the policy enforcement. The implementation of same-sex marriage legalization policy has no significant effect on the new HIV infection rate during 1996-2019, but increase the rate in early year from 1996-2010, although the latter result seems incredible because of the unbalance of the sample in 1996-2010.

Conclusion: Based on current study, the policy on legalization of same-sex marriage has no significant effect on new HIV infection rate. It seems that legalizing same-sex marriage do not intervene HIV transmission, but more evidence is needed and the mechanism remains to be studied.

Acquired immunodeficiency syndrome

same-sex marriage

legalization

new HIV infection rate

Acquired immunodeficiency syndrome (AIDS) is a major global health challenge, caused by HIV infection and transmitted through sex, blood and mother to child, which is exceedingly harmful to human health[1]. The prevalence of AIDS has brought a huge burden of disease to mankind. In 2019, the global HIV/AIDS burden was 36.9 million cases, and the prevalence rate was 476 cases per 100,000[2].

The causes of homosexuality are complex, including genetic factors and environmental factors[3, 4]. AIDS was first recognized in 1981 and it was found that some homosexuals suffered from rare infectious diseases and malignant tumors, and the epidemic of the disease was related to the gay community[5]. At present, in many countries in the world, men who have sex with men (MSM) bear a high burden of AIDS transmission[6, 7].

Same-sex marriage has been legally recognized in many countries in the past two decades and the legalization of same-sex marriage has an important impact on the psychology and behavior of homosexual group[8]. Our study included countries that have formal legislation to protect the rights of same-sex couples, and countries that recognize civil unions of the same sex although they have no formal legislation. The legalization of same-sex marriage involves many social, cultural, legal and other factors, which has always been a controversial issue in countries around the world. At present, countries where same-sex marriage is legal include the Netherlands, Spain, Canada, France, Brazil, the United Kingdom, the United States, Germany and so on. Countries that have not yet legalized same-sex marriage include China, South Korea, Russia, Poland, Egypt, Tanzania, etc. According to the impact of the legalization of same-sex marriage on the behavior of homosexual group, this policy may affect the incidence of safe or dangerous same-sex behaviors to affect the new infection rate of AIDS.

The mechanism of this study is shown in Fig. 1. AIDS spreads through many ways. Based on the direct impact of the same-sex marriage legalization policy on same-sex sexual behavior, this study explores its indirect impact on the new HIV infection rate in various countries. At present, most of the research on AIDS is concentrated in the fields of biomedicine and public health, and less in the fields of humanities and social sciences. Based on the legalization policies of same-sex marriage in various countries, this study explored its impact on the new HIV infection rate, in order to provide scientific reference for the prevention and control strategies of AIDS.

Data source

The panel data of new AIDS infection rate and control variables in each country in this study were from the open source data website "our world in data" ( https://ourworldindata.org/ ). The data of 172 countries from 1996 to 2019 on the number of new HIV infections, GDP per capita, population, Average total years of schooling for adult population, Human Development Index (HDI), Share of deaths due to unsafe sanitation, Life expectancy at birth, Urban population, Human rights protection, Prevalence of drug use disorders were extracted. The new infection rate was calculated according to the number of new infections and the total population in that year, and the unbalanced panel data was constructed, To ensure the analyzability of the data, we only used the data of countries with observation time of more than ten years.

Statistical analysis

This study regarded the implementation of the same-sex marriage legalization policies in various countries as a quasi-natural experiment, and explores the impact of the same-sex marriage legalization policies in various countries on the new HIV infection rate based on the difference in difference method (DID).

DID method is often used in quasi-experiment to evaluate treatment effect such as enforcement of a policy. In this paper, we divided the countries into two groups: countries with the policy enforcement were in treatment group (${G}_{i}=1$), while others were in control group(${G}_{i}=0$). Let t represent the treatment stage: t = 1 before treatment and t = 2 otherwise, and D = 1 if the unity was treated by the policy and 0 otherwise. Therefore, ${Y}_{it}\left(D\right)$ was explained as outcome of country i at certain stage t and whether it is treated at the time. To evaluate the treatment effect in treatment group (ATT) is to estimate $\text{E}\left\{{Y}_{i2}\left(1\right)-{Y}_{i2}\left(0\right)|{G}_{i}=1\right\}$. Under parallel trend assumption and after following transformation we can get:

The last equation holds because of parallel trend assumption: If no treatment to both group, all countries will have the same trend of their outcomes. To estimate ATT, we can use linear model which is equivalent to former equation. A simple model is shown as :

$${Y}_{it}={\alpha }_{0}+{\alpha }_{1}{G}_{i}+{\alpha }_{2}{T}_{t}+{\alpha }_{3}{D}_{it}+{\beta }{X}_{it}+{\epsilon }_{it}$$

Where ${G}_{i}$ and ${T}_{t}$ are all dummy variables, ${D}_{it}= {G}_{i}\times {T}_{t}$, ${X}_{it}$ are covariate variables. In the equation, we can see ${\alpha }_{3}$ is the treatment effect and we want to estimate it in later models.

Our study divides countries into treatment group and control group according to the legalization of same-sex marriage. The treatment group is countries that implement policies related to the legalization of same-sex marriage, including countries that have formal legislation to protect the rights of same-sex couples (the Netherlands, Canada, etc.), countries that recognize same sex civil union (Chile, Italy, etc.) and countries where the approval of same-sex marriage by the court is not unconstitutional and civil unions of the same sex have been recognized in some regions (Japan). After screening, 35 countries including the Netherlands, Canada, Argentina, Uruguay, France, Brazil, the United Kingdom, the United States, Germany, South Africa and so on are selected as the treatment group of policy impact, and 137 countries including China, South Korea, Russia, Poland, Egypt and Tanzania are selected as the control group. The explained variable is the new HIV infection rate of each country (region). Control variables include GDP per capita, Average Total Years of Schooling for Adult Population, HDI, Share of deaths due to unsafe sanitation, Life expectancy at birth, Urban population, Human rights protection, Prevalence of drug use disorders were extracted. Table 1 reports the characteristics of the core explanatory variables and outcome variables, Table 2 reports the new HIV infection rate in countries of the treatment group, and Table 3 reports the characteristics of the control variables.

Static DID model

We used a static DID model to identify the impact of legalization of same-sex marriage on the new HIV infection rate. The basic assumption of DID model is that if the countries that legalized same-sex marriage did not implement the policy, the new HIV infection rate of it should have the same time trend as the countries that did not legalize same-sex marriage (i.e. the parallel trend assumption, which will be verified in event study). A potential threat to this hypothesis is that countries do not adopt policies randomly. To address this issue, we estimate the effects by controlling fixed effects across countries. The model is shown as Eq. (1) :

$${AIDS}_{it}=\alpha +\beta {D}_{it}+{{X}_{it}}^{{\prime }}\theta +{\mu }_{i}+{\gamma }_{t}+{\epsilon }_{it}, \left(1\right)$$

where the subscript i represents $\text{i}\text{t}\text{h}$ country and t represents the year. ${AIDS}_{it}$ indicates the new rate of HIV infection in country i in year t. ${D}_{it}$ indicates whether country i has implemented the policy of legalizing same-sex marriage in year t, which is a dummy variable and $\text{D}=1$ indicates that the policy has been implemented in the country.${\mu }_{i}$ represents fixed effects across countries, which are used to capture all country characteristics that are not changing over time and have the potential to influence the new HIV infection rate. ${\gamma }_{t}$ represents year fixed effects, which control shocks affecting all countries in a given year. ${X}_{it}$ represents other control variables. ${\epsilon }_{it}$ represents a random error term, and here we adopt the cluster standard error in the country dimension.

Dynamic DID model

The representation of dynamic DID model is as following equation

$${AIDS}_{it}=\alpha +\sum _{k=5-}^{-2}{\beta }_{k}{D}_{ik}+\sum _{j=0}^{5+}{\beta }_{j}{D}_{ij}+{X}_{it}^{{\prime }}\theta +{\mu }_{i}+{\gamma }_{t}+{\epsilon }_{it} \left(2\right)$$

Where ${AIDS}_{it}$ represents the number of newly infected persons in $t$th year of the $i$th country;${D}_{ik}$ and ${D}_{ij}$ represent that for each country in the treatment group, D = 1 when the difference between current year t and treatment year ${t}_{0}$ is equal to j or k. When the relative treatment year is less than or equal to -5 or greater than or equal to 5, ${D}_{-5}$ and ${D}_{5}$=1,otherwise 0; The values of ${D}_{ik}$ and ${D}_{ij}$ are constantly 0 for the countries in the control group.${X}_{it}$ is the control variable mentioned above. ${\mu }_{i}$ and ${\gamma }_{t}$ are fixed effect of country and year, respectively, and ${\epsilon }_{it}$ is the error term. We take period ${D}_{1}$, namely the period before the policy occurred, as the baseline period to study the treatment effect of each period after the legalization of same-sex marriage on the new incidence of HIV infection.

Table 1

Core explanatory variables and outcome variables
Core explanatory variables and outcome variables	Total observed value	Mean	Standard deviation
Legalization of same-sex marriage	3654	0.2137	0.4100
New HIV infection rate (‰) (policy not implemented)	3364	0.9450	2.6365
New HIV infection rate (‰) (policy has been implemented)	290	0.7991	2.6993
Data source: “Our World in Data”(https://ourworldindata.org/)

Table 2

Treatment group: new HIV infection rate before and after policy treatment (‰)
Country	Time of legalization of same-sex marriage	Before policy implementation		After policy implementation
Country	Time of legalization of same-sex marriage	Mean	Standard deviation	Mean	Standard deviation
Netherlands	2000	0.0510	0.0017	0.0410	0.0077
Spain	2005	0.0329	0.0060	0.0619	0.0276
Canada	2005	0.0719	0.0028	0.0750	0.0142
Portugal	2010	0.1356	0.0613	0.1323	0.0324
Argentina	2010	0.1922	0.0123	0.2331	0.0185
Uruguay	2013	0.1566	0.0260	0.2272	0.0226
New Zealand	2013	0.0178	0.0055	0.0305	0.0020
France	2000	0.0535	0.0029	0.0408	0.0049
Brazil	2013	0.2640	0.0225	0.3139	0.0065
United Kingdom	2014	0.0845	0.0104	0.0816	0.0007
United States	2015	0.1589	0.0137	0.1992	0.0067
Mexico	2009	0.0945	0.0122	0.1244	0.0076
Germany	2000	0.0332	0.0048	0.0256	0.0020
Australia	2017	0.0530	0.0078	0.0585	0.0010
South Africa	2000	14.4938	0.8845	10.3493	2.3629
Italy	2016	0.0452	0.0073	0.0618	0.0034
Croatia	2014	0.0102	0.0024	0.0107	0.0009
Chile	2015	0.1272	0.0260	0.1963	0.0162
Cyprus	2015	0.0078	0.0015	0.0103	0.0001
Czech Republic	2006	0.0063	0.0010	0.0093	0.0005
Estonia	2014	0.1156	0.0575	0.1380	0.0058
Greece	2015	0.0147	0.0057	0.0148	0.0003
Luxembourg	2014	0.0226	0.0071	0.0318	0.0030
Denmark	2012	0.0424	0.0029	0.0434	0.0025
Iceland	2010	0.0180	0.0053	0.0202	0.0025
Austria	2010	0.0695	0.0136	0.0771	0.0061
Norway	2008	0.0358	0.0025	0.0381	0.0038
Belgium	2003	0.0959	0.0076	0.0949	0.0133
Finland	2014	0.0090	0.0016	0.0106	0.0004
Columbia	2016	0.1204	0.0290	0.1727	0.0032
Slovenia	2017	0.0032	0.0009	0.0037	-
Mali	2014	0.0351	0.0112	0.0547	0.0018
Ireland	2015	0.0149	0.0025	0.0248	0.0015
Sweden	2009	0.0119	0.0025	0.0220	0.0024
Data source: “Our World in Data”(https://ourworldindata.org/)

Table 3

Descriptive statistics of control variables
Control variable	Observed value	Mean	Standard deviation	Min	Max
GDP per capita (1000 dollars)	3,654	17.56	18.99	0.51	114.89
Average Total Years of Schooling for Adult Population (year)	3,654	7.75	3.17	0.90	14.20
HDI	3,654	0.67	0.16	0.24	0.95
Share of deaths due to unsafe sanitation (%)	3,654	1.4	2.1	0.0	11.1
Life expectancy at birth (year)	3,654	69.10	9.51	35.38	84.36
Urban population (%)	3,654	55.2	22.6	7.4	100.0
Human rights protection	3,654	0.65	1.49	-2.56	5.34
Prevalence of drug use disorders (%)	3,654	0.7	0.4	0.2	3.7
Data source: “Our World in Data”(https://ourworldindata.org/)

Static DID model

Figure 2 reports the trends in new HIV infection rate over time of the control group (we take 2010 as the base year) and the treatment group (the year same-sex marriage was legalized). It was observed that the new HIV infection rate in the treatment group is almost parallel to that in the control group before base year. After the treatment year, the average new HIV infection rate in the treatment group tends to decrease, while the rate in the control countries increases after 2010. So it is speculated that this policy may reduce the rate of new HIV infections.

Table 4 reports the results of the static DID regression model. It can be seen that compared with the countries without legalization of same-sex marriage, the countries with legalization of same-sex marriage have an increase of 0.142‰ of the new HIV infection rate, but the result is not statistically significant. The possible reason is that although the legalization of same-sex marriage can allow homosexual population to obtain more rights under the legal protection, the government can actively guide the adoption of correct protective measures to reduce the incidence of high-risk homosexual behaviors, so as to decrease the incidence of new HIV infection. But in fact, the legalization of same-sex marriage could not change the cognition on the prevention of HIV infection caused by homosexual behaviors among the homosexual community to a certain extent, nor did it affect people's choice of sexual attitude, so the new incidence of HIV infection did not change significantly. Therefore, in this study, the legalization of same-sex marriage has no significant impact on the new incidence of HIV infection.

Table 4

Effect of same-sex marriage legalization on HIV infection rate: static DID model
Variables	Coefficient and standard error
D	0.129 (0.184)
GDP per capita	0.016 (0.013)
Average Total Years of Schooling for Adult Population	-0.267** (0.132)
HDI	15.590*** (4.885)
Share of deaths due to unsafe sanitation	-0.029 (0.161)
Life expectancy at birth	-0.269*** (0.062)
Urban population	-0.028 (0.034)
Human rights protection	0.184** (0.092)
Prevalence of drug use disorders (%)	1.975* (1.070)
Intercept	10.946*** (4.293)
N	3654
R²	0.9106
Note: 1. The standard error in this study is cluster standard error in the country dimension 2. Significance criterion: * p < 0.05, p < 0.01, * p < 0.001

Dynamic DID model

In order to accurately test whether the aforementioned DID method meets the Parallel trend assumption and specifically explore the impact of same-sex marriage legalization policies on the new HIV infection rate at different time points, we used event-study to evaluate the dynamic effects of the policies.

Parallel trend hypothesis verification

The dynamic DID model was used for full-sample regression, and the results are shown in column (1) of Table 5. The treatment effect and confidence interval are shown in Fig. 3. In the first five periods before policy enforcement, except for the first period, the coefficients obtained are not significant. Therefore, this study roughly meets the hypothesis of parallel trend, which indicates the validity of assumptions in DID study.

Evaluation of policy effects

Full sample analysis

The event study method was still used to analyze all countries during the period from 1996 to 2019. The regression results and figures of dynamic treatment effects are shown in column (1) of Table 5 and Fig. 3. It can be seen that, except for the periods of legalization of same-sex marriage and the fifth period after legalization, the coefficients of the other three periods were all positive, but there was no statistically significant significance. Therefore, we can roughly conclude that the implementation of same-sex marriage legalization has no significant impact on the new incidence of HIV infection. At the same time, we can see that compared with the treatment period closer to the base period, the longer the time from the base period, the larger the standard error is estimated, which may be related to the long observation period. We will talk more about it later.

Study of treatment effect events in the period from 1996 to 2010

We limited the time interval in 15 years from 1996 to 2010 to evaluate the treatment effect of same-sex marriage legalization on the new incidence of HIV infection during this period. In this period, the countries that legalize same-sex marriage were in the treatment group, while the countries that legalize same-sex marriage after 2010 were in the control group. The results of analysis using dynamic DID model are shown in column (2) of Table 5. The dynamic effect is shown in Fig. 4. As can be seen from the figure, in the first five periods of treatment, the regression coefficients are not significant, indicating that the parallel trend hypothesis is satisfied in this case. In the period of policy enforcement and the following four periods, the regression coefficients are not statistically significant, while in the fifth period, the estimated regression coefficient is 0.707, which is statistically significant. This result indicates that the legalization of same-sex marriage from 1996 to 2010 increased the new HIV infection rate by 0.7‰ five years after the policy enforcement. However, this result is not necessarily credible because of the small number of countries that implemented the policy during this period, which will be talked further in discussion.

Study of treatment effect events in the period from 2005 to 2019

We limit the time interval in the period from 2005 to 2019 to evaluate the effect of same-sex marriage legalization on the incidence of new HIV infection rate during this period. In this analysis, the countries that have already legalized same-sex marriage in the period from 1996 to 2004 are excluded from the sample. The results are shown in column (3) of Table 5, and the dynamic treatment effect is shown in Fig. 5. The result shows that in the first four years before the policy enforcement period, the regression coefficients are not statistically significant, indicating that the hypothesis of parallel trend in this period is also satisfied, and DID research is applicable. In the current period of policy enforcement and five periods after policy implementation, the regression coefficient is close to 0 and there is no statistical significance. This suggests that legalizing same-sex marriage did not have a significant effect on new HIV infection rate between 2005 and 2019.

Table 5

Estimation of the effect of same-sex marriage legalization on new HIV/AIDS infection rate: analysis based on dynamic DID model
Explanatory variables	(1)	(2)	(3)
Explanatory variables	Full- sample analysis	Analysis in time interval 1996–2010	Analysis in time interval 2005–2019
eventpre5	-0.220 (0.179)	-0.130 (0.237)	-0.015 (0.054)
eventpre4	-0.009 (0.066)	-0.042 (0.134)	0.007 (0.027)
eventpre3	0.002 (0.043)	-0.047 (0.082)	0.017 (0.027)
eventpre2	0.008 (0.022)	-0.024 (0.038)	-0.003 (0.022)
event0	-0.006 (0.022)	0.055 (0.044)	-0.005 (0.014)
eventpost1	0.007 (0.039)	0.008 (0.126)	-0.005 (0.030)
eventpost2	0.025 (0.055)	-0.062 (0.202)	-0.001 (0.036)
eventpost3	0.025 (0.081)	-0.036 (0.298)	-0.022 (0.043)
eventpost4	0.007 (0.119)	0.411 (0.354)	-0.016 (0.070)
eventpost5	-0.004 (0.247)	0.707* (0.324)	-0.014 (0.095)
GDP per capita	0.014 (0.013)	0.027 (0.016)	0.000 (0.005)
Average total years of schooling for adult population	-0.274* (0.134)	-0.164 (0.107)	-0.238* (0.101)
HDI	16.014** (5.030)	14.284* (6.098)	5.611** (1.956)
Share of deaths due to unsafe sanitation	-0.034 (0.161)	0.332* (0.158)	-0.291*** (0.073)
Life expectancy at birth	-0.270*** (0.062)	0.015 (0.074)	-0.341*** (0.062)
Urban population	-0.027 (0.034)	-0.028 (0.037)	-0.001 (0.016)
Human rights protection	0.180 (0.091)	0.153 (0.108)	-0.027 (0.028)
Prevalence of drug use disorders	1.902 (1.074)	2.465* (1.071)	-0.466 (0.372)
Intercept	10.840* (4.303)	-9.194 (5.390)	23.457*** (3.668)
N	3654	2224	2162
R²	0.911	0.96	0.98
Note: 1. Standard errors in parentheses are cluster standard errors calculated by individual fixed effects of countries 2. The regression takes the period before the implementation of the policy as the benchmark group and assumes that its treatment effect is 0 3. Significance criterion: * p < 0.05, p < 0.01, * p < 0.001

The legalization of same-sex marriage has been highly discussed in the 21st century, especially after the Netherlands became the first country to legalize same-sex marriage in 2000, and more countries corresponding policies later. However, few research has been conducted on the public health effects and sociological implications of this policy. For example, in 2006, Jeffrey D. Klausner et al. investigated and analyzed the community population of gay men in Chicago, Los Angeles and other places in the United States, and concluded that compared with the situation of non-regular sexual partners, the prevalence of sexually transmitted diseases is significantly lower in homosexual couples with stable marriage and family relationships[9]. But the study was conducted in parts of the United States and did not analyze the effects of policies legalizing same-sex marriage. Diederik Boertien's survey and analysis in England and Wales showed that after the implementation of the same-sex marriage legalization policy, the subjective well-being of same-sex couples increased significantly[10], reflecting the positive effect of the legalization of same-sex marriage. However, this study did not study the related transmitted diseases.

This study based on 1996–2019 years of global countries do static DID and event study, analysis to legalize same-sex marriage to the influence of new HIV infections, although in this paper, based on the existing methods, the data and the results showed that the policy for the new infections of AIDS and had no significant effect, but still had some limitations, this study analysis is as follows:

The time span of this study is relatively large. Even if our event study results suggest that the parallel trend hypothesis can be met, we are still not sure whether the parallel trend of development can be maintained between the treatment group and the control group with different treatment times in the 24-year time span from 1996 to 2019. In particular, it is impossible to estimate the impact that a series of international events will have on the policy. In addition, the time interval in this study is from the five years before policy enforcement to five years later, 11 years in total. However, the length of observation is 24 years, which will cause bias in estimation, especially at the end point (the fifth year before and after policy implementation). That’s why we conduct DID in different shorter time intervals.

From 1996 to 2019, 35 countries have legalized same-sex marriage, while 137 are countries in the control group. In comparison, fewer countries implemented policies will lead to greater bias in the estimation of policy effects. Only 10 countries implemented the policy during the period 1996–2010, which makes our estimated effect unstable. Even if the policy has a significant positive effect in the fifth period after implementation in our study, it is not necessarily credible. In contrast, the results of the study between 2005 and 2019 are more credible due to the short time span and the appropriate proportion of countries in the treatment group (about 21%), which leads to the conclusion of this paper.

In this paper, we only explored the impact of the legalization of same-sex marriage on the new HIV infection rate, but we did not explore the possible mechanism in detail. In terms of life experience, same-sex marriage may reduce the new HIV/AIDS infection rates by giving homosexual population legal protection and leading them to avoid risk sexual behaviors. However, it may also encourage unsafe sex behaviors and then add new infections. In this article we only explore the comprehensive effect due to limitation of length of this article, and specific mechanism remains to be explored.

At present, the global HIV/AIDS infection rate is still high, and AIDS prevention and control faces serious challenges. While homosexual behavior is an important way of HIV infection, it is of great significance to study the impact of same-sex marriage legalization on AIDS. Under the quasi-experiment, we inspect the world legalize same-sex marriage to the influence of new HIV infections, the conclusion is as follows: in combination with existing evidence, we cannot presume that the implementation of same-sex marriage has a significant effect on new HIV infection rate, and this is only a kind of comprehensive effect, which still needs more evidence for further analysis of the mechanism.

Ethics approval and consent to participate

Not applicable

Consent for publication

Not applicable

Availability of data and materials

The original dataset presented in this study is included in the supplementary material, where we provide the raw data downloaded from the website mentioned before. To complete missing data, we also use data from Human Development Report (you can get access to it at https://hdr.undp.org/reports-and-publications) in2018 and 2019.Further inquiries can be directed to the corresponding authors.

Competing interests

The authors declare there are no competing interests.

Funding

Not applicable

Author Contributions

Jinzhu Jia supervised this research. Kuan Jiang and Mintao Su contributed to the study conception, implementation of the study, design, interpretation and critical was involved in drafting of the manuscript. Kuan Jiang and Mintao Su are dedicated to data collection, data analysis and results interpretation. Zhongsong Zhang made major contributions to the descriptive statistics of the data and the drafting and revision of the manuscript. All authors have read and approved the manuscript and ensured that this is the case.

Acknowledgements

Here we acknowledge Jinzhu Jia for his revising our article. Meanwhile, Xuan Wang, Shiyuan Li and Xinru Huang also taught us methodology in class.

Sharp PM, Hahn BH. Origins of HIV and the AIDS pandemic. Cold Spring Harbor perspectives in medicine 2011, 1(1).
Govender RD, Hashim MJ, Khan MA, Mustafa H, Khan G. Global Epidemiology of HIV/AIDS: A Resurgence in North America and Europe. J Epidemiol Glob Health. 2021;11(3):296–301.
Tu D, Xu R, Zhao G, Wang B, Feng T. [Research progress on molecular genetics of male homosexuality]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi. 2016;33(4):569–72.
Acosta FX. Etiology and treatment of homosexuality: a review. Arch Sex Behav. 1975;4(1):9–29.
Greene WC. A history of AIDS: looking back to see ahead. Eur J Immunol. 2007;37(Suppl 1):94–102.
Szmulik-Misiurek K, Niedźwiedzka-Stadnik M, Rosińska M. HIV and AIDS in Poland in 2018. Przegl Epidemiol. 2020;74(2):223–38.
Smith AD, Tapsoba P, Peshu N, Sanders EJ, Jaffe HW. Men who have sex with men and HIV/AIDS in sub-Saharan Africa. Lancet. 2009;374(9687):416–22.
Drabble LA, Wootton AR, Veldhuis CB, Riggle EDB, Rostosky SS, Lannutti PJ, Balsam KF, Hughes TL. Perceived psychosocial impacts of legalized same-sex marriage: A scoping review of sexual minority adults' experiences. PLoS ONE. 2021;16(5):e0249125.
Klausner JD, Pollack LM, Wong W, Katz MH. Same-sex domestic partnerships and lower-risk behaviors for STDs, including HIV infection. Journal of homosexuality 2006, 51(4).
Boertien D, Vignoli D. Legalizing Same-Sex Marriage Matters for the Subjective Well-being of Individuals in Same-Sex Unions. Demography. 2019;56(6):2109–21.

No competing interests reported.

rawdata.zip

Download PDF

Version 1

posted

You are reading this latest preprint version

Impact of legalization of same-sex marriage on the new HIV infection rate in 172 countries around the world based on difference-in-differences (DID) method

Status:

Version 1

Abstract

Figures

Background

Methods

Data source

Statistical analysis

Static DID model

Dynamic DID model

Result

Static DID model

Dynamic DID model

Parallel trend hypothesis verification

Evaluation of policy effects

Full sample analysis

Study of treatment effect events in the period from 1996 to 2010

Study of treatment effect events in the period from 2005 to 2019

Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1