‘probably yes’, ‘probably not’, and ‘definitely not’. A period of two years was chosen to receive information about concrete fertility intentions. If respondents replied ‘probably not’ or ‘definitely not’ to have a first or another child within the next two years, then a question was asked whether they intend ever to have a first or another child. The answers could be differentiated in the same way as in the previous question. The investigation of tempo-related intentions to have a first or another child is only meaningful for individuals that actually want to have this child. Therefore, only respondents that want to have ever a first or another child are considered for this variable. For the variable about the quantum related fertility intentions, all respondents that probably or definitely want to have a child within the next two years are coded as respondents that also definitely ever want to have the respective child (for more details on the tempo and quantum of fertility intentions, see Philipov et al. 2005).
“可能”、“可能不会”和“肯定不会”。我们选择两年时间来收集有关具体生育意愿的信息。如果受访者回答说“可能不会”或“肯定不会”在未来两年内生育第一个孩子或另一个孩子，那么我们会询问他们是否打算生育第一个孩子或另一个孩子。答案的区分方式与上一个问题相同。对生育第一个孩子或另一个孩子的时机相关意愿的调查仅对真正想要生育这个孩子的个人有意义。因此，只有想要生育第一个孩子或另一个孩子的受访者才会被考虑为这个变量。对于与生育数量相关的生育意愿变量，所有可能或肯定希望在两年内生育孩子的受访者都被编码为肯定希望生育相应孩子的受访者（有关生育意愿的时机和数量的更多详细信息，请参阅Philipov等人，2005）。

Respondents’ social capital was measured by the amount of experienced and potential transfers of resources. Respondents were asked about the size of three different networks that provide them with three different kinds of resources during the last two years: ‘small help’, ‘substantial and important support’, and ‘borrowed money’. If the respondents reported that no network partners of a particular network provided them with the particular resource, they were asked to name the number of network partners from whom they would receive this resource if needed. For example, if a respondent replied that nobody provided ‘substantial and important support’, then the respondent was asked to name the number of network partners he or she could ask for this kind of support. For each of the three kinds of resources, the numbers of experienced and potential supportive network partners are summarised to three variables that provide information about the respondents’ amount of social capital according to ‘small help’, ‘substantial and important support’, and ‘borrowed money’. ${}^4$${}^4$^(4){ }^{4} To cover the respondents’ involvement in long-term exchange relationships and investments in social capital, they were also asked about the number of network partners to whom they provided ‘important and substantive support’. If no network partner was named, the respondents were asked about the number of network partners that would ask them to provide this resource. These two variables are, again, summarised to a new one giving information about the number of network partners that receive ‘important and substantive support’ from the respondents.
受访者的社会资本通过资源转移的量和潜力来衡量。受访者被问及在过去两年中，为他们提供三种不同资源的三种不同网络的大小：“小帮助”、“实质性和重要的支持”和“借款”。如果受访者报告说，特定网络中的任何网络合作伙伴都没有为他们提供特定的资源，则要求他们说出如果有需要，可以从哪些网络合作伙伴那里获得这种资源。例如，如果受访者回答说没有人提供“实质性和重要的支持”，则要求受访者说出可以要求这种支持的网络合作伙伴的数量。对于这三种资源中的每一种，根据“小帮助”、“实质性和重要的支持”和“借款”，将经历的和潜在的支持网络合作伙伴的数量总结为三个变量，以提供有关受访者社会资本数量的信息。为了涵盖受访者参与长期交换关系和社会资本投资的情况，还询问了他们向哪些网络合作伙伴提供了“重要和实质性的支持”。如果没有提到任何网络合作伙伴，则询问受访者有多少网络合作伙伴会要求他们提供这种资源。这两个变量再次总结为一个新变量，以提供有关从受访者那里获得“重要和实质性的支持”的网络合作伙伴数量的信息。

The subsequent analyses consider a very simple property of social networks, namely, size. However, this structural characteristic is a central dimension of social capital (Bourdieu 1983; Flap 2002). The analyses address the processes of interpersonal exchange as the basic mechanism of generating social capital. Structural properties of social networks, such as density, cohesiveness, openness, or structural similarity, are very much outcomes of these processes and are not considered subsequently.
随后的分析考虑了社交网络的一个非常简单属性，即规模。然而，这一结构特征是社会资本的核心维度（Bourdieu 1983；Flap 2002）。分析将人际交流过程作为社会资本产生的基本机制。社交网络的结构属性，如密度、凝聚力、开放性或结构相似性，很大程度上是这些过程的结果，因此不在后续考虑之列。

For a subgroup of network partners, the questionnaire collected personal characteristics and some attributes of the respondent’s relationships with them. For example, when a respondent reported about network partners from whom he received ‘small help’, he was asked to select up to five network partners that were most important within this context. Next, he was asked for the network partners’ gender, travelling distance between the respondent and them, their frequency of contact, and their role relationships, i.e. whether they belong to the respondent’s family or whether they are relatives, friends, acquaintances, colleagues etc. The latter variable can be used as an indicator for the kind of reciprocity that characterises the relationships. Members of the extended family and kin build systems of indirect reciprocal exchange. Relationships with the partner, friends, or colleagues rest on direct reciprocity. The same information was collected for up to five network partners that ‘lend money to the respondent’ and/or that received ‘important and substantive support’ from her. For the analyses, the network partners from these two networks as well as from the network that provided the respondent with ‘small help’ are pooled and their characteristics are aggregated. This is done to receive some independence between the composition of role relationships and the kind of resources received or given. Two new variables are created to indicate the number of indirect reciprocal relationships between the respondents and their network partners, measured by the number of extended family members and relatives, and to measure the number of direct reciprocal relationships, represented by the number of friends, colleagues, neighbours, and acquaintances. Husbands and partners are not considered, as the analyses concentrate on the exchange of resources with network partners outside the respondents’ core family. To cover a probably significant role the respondents’ direct parents may play in the relationships of indirect reciprocity, three additional variables are constructed. Two dummy variables that report whether the respondent named one or both direct parents and one variable about the number of indirect reciprocal relationships to all other relatives.
对于网络合作伙伴的子群体，问卷调查收集了受访者的个人特征以及他们与网络合作伙伴关系的某些属性。例如，当受访者报告说从网络合作伙伴那里获得了“小帮助”时，他会被要求选择最多五个在这种情况下最重要的网络合作伙伴。接下来，他会被要求提供网络合作伙伴的性别、受访者与他们的旅行距离、联系频率以及角色关系，即他们是否属于受访者的家人，或者他们是亲戚、朋友、熟人、同事等。后一个变量可以用作关系中互惠类型的指标。大家庭和亲属成员建立间接互惠交换系统。与合作伙伴、朋友或同事的关系建立在直接互惠的基础上。对于“借钱给受访者”和/或从受访者那里获得“重要和实质性支持”的最多五个网络合作伙伴，也收集了相同的信息。为了进行分析，来自这两个网络以及向受访者提供“小帮助”的网络中的网络合作伙伴被集中起来，其特征被汇总。这样做是为了在角色关系的构成与所获得或提供的资源类型之间获得一定的独立性。创建了两个新变量来指示受访者与他们的网络合作伙伴之间的间接互惠关系数量，以大家庭成员和亲戚的数量来衡量，并衡量直接互惠关系数量，以朋友、同事、邻居和熟人的数量来表示。 丈夫和伴侣不在考虑之列，因为分析的重点是受访者核心家庭以外的网络伙伴之间的资源交换。为了涵盖受访者直系父母在间接互惠关系中可能发挥的重要作用，我们构建了三个额外的变量。两个虚拟变量用于报告受访者是否指定了直系父母中的一方或双方，另一个变量用于报告与所有其他亲属的间接互惠关系数量。

Primarily for the purpose of control, the multivariate analyses consider the basic characteristics of the respondents, their husbands or partners, and their households. Husbands’, partners’, and households’ characteristics are considered because we assume that women form their fertility intentions not solely based on
出于控制的目的，多元分析考虑了受访者、其丈夫或伴侣以及家庭的基本特征。之所以考虑丈夫、伴侣和家庭的特征，是因为我们认为女性形成生育意愿时，不仅考虑
their personal characteristics. The respondents are portrayed by age, the completed or aspired level of education, the employment situation, their religiosity, and the number of siblings they have. Husbands or partners are attributed with their age, the level of education completed or aspired, and their employment situation. As the analyses take the respondents’ perspectives into account, they consider only the characteristics of husbands’ or partners’ that we assume most of the respondents to know accurately. However, information on these characteristics is taken from the husbands’ or partners’ interviews. The situation of the household is represented by its equivalence income, i.e. by the household members’ per capita income weighted by the age structure of the household. ${}^5$${}^5$^(5){ }^{5} Finally, one variable controls for the differences of fertility intentions between urban and rural areas.
他们的个人特征。受访者按年龄、已完成或期望的教育水平、就业状况、宗教信仰以及兄弟姐妹数量进行描述。丈夫或伴侣按年龄、已完成或期望的教育水平以及就业状况进行描述。由于分析考虑了受访者的观点，因此只考虑了我们认为大多数受访者都准确了解的丈夫或伴侣的特征。然而，这些特征的信息来自对丈夫或伴侣的访谈。家庭状况由等值收入表示，即按家庭年龄结构加权计算的家庭成员人均收入。 ${}^5$${}^5$^(5){ }^{5} 最后，一个变量控制了城乡地区生育意愿的差异。

The empirical results are presented in two steps. First, the distributions of the central variables in the analyses are reported: the respondents’ fertility intentions and the size and composition of their exchange networks. Next, estimates from ordered logistic regressions are discussed to explore the possible effects of the respondents’ social capital and their embeddedness in relationships of direct and indirect reciprocity on their fertility intentions.
实证结果分两步呈现。首先，报告分析中的中心变量的分布：受访者的生育意愿以及他们的交流网络的规模和组成。其次，讨论有序逻辑回归的估计值，以探讨受访者的社会资本以及他们在直接和间接互惠关系中的嵌入程度对他们的生育意愿可能产生的影响。

Of all women considered in the analyses, $35.9\mathrm{\%}$$35.9\mathrm{\%}$35.9%35.9 \% definitely intend to have a first or another child (see Table 2). This compares to $36.1\mathrm{\%}$$36.1\mathrm{\%}$36.1%36.1 \% who perceive their reproductive period to be completed and for certain do not intend to give birth to a child. However, as expected, these intentions depend significantly on the actual number of children belonging to the respondent. Most of the childless respondents ( $87.7\mathrm{\%}$$87.7\mathrm{\%}$87.7%87.7 \% ) intend to have at least one child for sure. However, this intention changes significantly in the case of a second child. Here, only $43.3\mathrm{\%}$$43.3\mathrm{\%}$43.3%43.3 \% are certain that they want to have a second child whereas $37.4\mathrm{\%}$$37.4\mathrm{\%}$37.4%37.4 \% intend not to have more than one child. Finally, among the respondents with two, three, or more children, only a marked minority that intends to have a further child can be identified.
在分析的所有女性中， $35.9\mathrm{\%}$$35.9\mathrm{\%}$35.9%35.9 \% 肯定打算生第一个或第二个孩子（见表 2）。相比之下， $36.1\mathrm{\%}$$36.1\mathrm{\%}$36.1%36.1 \% 认为她们的生育期已经结束，肯定不打算生孩子。然而，正如预期的那样，这些打算很大程度上取决于受访者的实际子女数量。大多数没有孩子的受访者（ $87.7\mathrm{\%}$$87.7\mathrm{\%}$87.7%87.7 \% ）肯定打算至少生一个孩子。然而，对于第二个孩子，这种打算发生了很大变化。在这里，只有 $43.3\mathrm{\%}$$43.3\mathrm{\%}$43.3%43.3 \% 肯定想要生第二个孩子，而 $37.4\mathrm{\%}$$37.4\mathrm{\%}$37.4%37.4 \% 打算只生一个孩子。最后，在有两个、三个或更多孩子的受访者中，只有极少数人打算再要一个孩子。

Table 2:  表2：
Intention ever to have a first or another child by the number of children belonging to the respondent (natural children, step, and fostered children)
根据受访者所生子女的数量（亲生子女、继子和领养子女），推测其是否打算生育第一个或更多孩子

Note: * Due to 81 cases with nonresponse or missing information, the total number of respondents in this table is smaller than the overall population ( $\mathrm{n}-2,016$$\mathrm{n}-2,016$n-2,016\mathrm{n}-2,016 ) considered in the analyses.
备注：* 由于有 81 个案例未回答或信息缺失，因此本表中的受访者总数小于分析中考虑的总人数（ $\mathrm{n}-2,016$$\mathrm{n}-2,016$n-2,016\mathrm{n}-2,016 ）。

The majority of the respondents who definitely or probably want to have a first or another child intends to have this child within the next two years (see Table 3). This holds especially for the timing of the first child. However, the intention to postpone the birth increases with parity.
绝大多数明确表示或可能想要生育第一个或第二个孩子的受访者都打算在未来两年内生育（见表3）。对于生育第一个孩子的时间安排尤其如此。然而，随着生育次数的增加，推迟生育的意愿也会增加。

The results in Tables 2 and 3 entail two limitations for the subsequent analyses. First, the distribution of the intentions of the childless women ever to have a first child is too skewed for a meaningful analysis ( $94.5\mathrm{\%}$$94.5\mathrm{\%}$94.5%94.5 \% definitely or probably
表2和表3中的结果对后续分析存在两个限制。首先，无子女女性生育第一个孩子的意愿分布过于偏斜，无法进行有意义的分析（ $94.5\mathrm{\%}$$94.5\mathrm{\%}$94.5%94.5 \% 肯定或可能

Table 3:  表3：
Intention to have a first or another child within the next two years by the number of children belonging to the respondent (natural children, step, and fostered children). All respondents that definitely or probably intend to have a first or another child
根据受访者所生子女的数量（亲生子女、继子和领养子女），在未来两年内生育第一个或第二个孩子的意愿。所有明确或可能打算生育第一个或第二个孩子的受访者

intend to have a first child). Second, the number of respondents with three or more children $(\mathrm{n}=40)$$(\mathrm{n}=40)$(n=40)(\mathrm{n}=40) as well as the number of respondents who probably or definitely intend to have a third $(\mathrm{n}=50)$$(\mathrm{n}=50)$(n=50)(\mathrm{n}=50) or a fourth or fifth one $(\mathrm{n}=3)$$(\mathrm{n}=3)$(n=3)(\mathrm{n}=3) within the next two years are too small. Therefore, the multivariate analyses will only consider the respondents’ intentions ever to have a second or a third child and their intentions to have a first or a second one within the next two years.
打算生第一个孩子）。其次，有三个或以上孩子的受访者数量 $(\mathrm{n}=40)$$(\mathrm{n}=40)$(n=40)(\mathrm{n}=40) 以及可能或肯定打算在未来两年内生第三个 $(\mathrm{n}=50)$$(\mathrm{n}=50)$(n=50)(\mathrm{n}=50) 或第四个或第五个孩子 $(\mathrm{n}=3)$$(\mathrm{n}=3)$(n=3)(\mathrm{n}=3) 的受访者数量太少。因此，多元分析将只考虑受访者是否打算生第二个或第三个孩子以及是否打算在未来两年内生第一个或第二个孩子。

The majority of the women reported to have access to the supportive resources that were addressed in the questionnaire. Of the respondents, only $6.0\mathrm{\%}$$6.0\mathrm{\%}$6.0%6.0 \% resp. $7.7\mathrm{\%}$$7.7\mathrm{\%}$7.7%7.7 \% replied not to know at least one network partner that provides ‘small help’ or ‘important and substantive support’ (see Table 4). $12.3\mathrm{\%}$$12.3\mathrm{\%}$12.3%12.3 \% do not know of network partners that lend money. $51.1\mathrm{\%}$$51.1\mathrm{\%}$51.1%51.1 \% of the respondents borrowed money from their network partners in the last two years. This money was repeatedly spent on goods for basic needs, such as food, clothing, or medicine ( $62.1\mathrm{\%}$$62.1\mathrm{\%}$62.1%62.1 \% ), but also on bills for heating and lighting ( $31.9\mathrm{\%}$$31.9\mathrm{\%}$31.9%31.9 \% ), or on unexpected payments of a smaller amount ( $25.6\mathrm{\%}$$25.6\mathrm{\%}$25.6%25.6 \% ), like repairs or small medical treatments (multiple answers were possible). The mean values of the size of the non-empty networks show that the respondents are able to reach on average 3.7 network partners when they need ‘small help’, followed by 2.7 persons that give ‘substantive and important support’, and 2.4 people that lend money to them. The results on the small networks agree with insights from other studies (see for example Bühler and Frątczak 2004, Pfenning 1995, or Bernard et al. 1990). Within the personal network, there is mostly only a small number of members able and willing to give substantive support. Of the respondents, $87.5\mathrm{\%}$$87.5\mathrm{\%}$87.5%87.5 \% gave or would give ‘important and substantive support’ to 3.0 network partners on average. A cross tabulation of the respondents who received and/or gave this kind of resource documents that $50.2\mathrm{\%}$$50.2\mathrm{\%}$50.2%50.2 \% of them were engaged in exchange processes, i.e. they received as well as provided ‘important and substantive support’ within the last two years.
大多数女性表示她们能够获得问卷中提到的支持资源。在受访者中，只有 $6.0\mathrm{\%}$$6.0\mathrm{\%}$6.0%6.0 \% 人表示不认识至少一个可以提供“小帮助”或“重要实质性支持”的网络伙伴（见表4）。 $12.3\mathrm{\%}$$12.3\mathrm{\%}$12.3%12.3 \% 人不知道有提供贷款的网络伙伴。 $51.1\mathrm{\%}$$51.1\mathrm{\%}$51.1%51.1 \% 的受访者在过去两年中向网络伙伴借钱。这些钱经常用于购买基本生活必需品，如食物、衣服或药品（ $62.1\mathrm{\%}$$62.1\mathrm{\%}$62.1%62.1 \% ），也用于支付取暖和照明费用（ $31.9\mathrm{\%}$$31.9\mathrm{\%}$31.9%31.9 \% ），或用于支付修理或小病治疗等小额意外费用（ $25.6\mathrm{\%}$$25.6\mathrm{\%}$25.6%25.6 \% ）（可多选）。非空网络规模的平均值表明，当受访者需要“小帮助”时，他们平均能够联系到3.7个网络伙伴，其次是2.7个提供“重要实质性支持”的人，以及2.4个借钱给他们的人。小网络的结果与其他研究结果一致（例如，参见Bühler和Frątczak 2004、Pfenning 1995或Bernard等人1990）。在个人网络中，通常只有少数成员能够并愿意提供实质性支持。在受访者中， $87.5\mathrm{\%}$$87.5\mathrm{\%}$87.5%87.5 \% 人平均向3.0个网络伙伴提供或愿意提供“重要实质性支持”。交叉分析显示，在获得和/或提供这种资源的人中， $50.2\mathrm{\%}$$50.2\mathrm{\%}$50.2%50.2 \% 人参与了交换过程，即他们在过去两年中既获得了“重要实质性支持”，也提供了这种支持。

Table 5 reports the networks’ compositions by the relationships of indirect and direct reciprocity. ${}^6$${}^6$^(6){ }^{6} Indirect reciprocal relationships form the majority. They make on average $55\mathrm{\%}$$55\mathrm{\%}$55%55 \% of the relationships between the respondents and their network partners. On the one hand, this is caused by the significance of the respondents’ direct parents ( $30\mathrm{\%}$$30\mathrm{\%}$30%30 \% ), but also by parents-in-law, members of the extended family, and relatives ( $25\mathrm{\%}$$25\mathrm{\%}$25%25 \% ). Direct parents are crucial providers of ‘substantive and important support’ and they are a valuable source of money, but they make also one fourth of the network partners that were supported by the respondents. Direct reciprocal relationships with friends, colleagues, neighbours, or acquaintances form a significant share of the respondents’ exchange activities within the last two years as well. This group of network partners is the most important source for borrowed money ( $37\mathrm{\%}$$37\mathrm{\%}$37%37 \% ) and they are also frequent beneficiaries of ‘important and substantive support’ given by the respondents. Furthermore, there are supportive relationships within the respondents’ core families. Husbands or partners make on average $27\mathrm{\%}$$27\mathrm{\%}$27%27 \% of the network partners that provided important
表5按间接和直接互惠关系列出了网络构成。 ${}^6$${}^6$^(6){ }^{6} 间接互惠关系占多数。它们平均占受访者与其网络伙伴之间关系的 $55\mathrm{\%}$$55\mathrm{\%}$55%55 \% 。一方面，这是由于受访者的直系亲属（ $30\mathrm{\%}$$30\mathrm{\%}$30%30 \% ）的重要性，另一方面，也是由于姻亲、大家庭成员和亲戚（ $25\mathrm{\%}$$25\mathrm{\%}$25%25 \% ）的重要性。直系亲属是“实质性和重要支持”的重要提供者，也是宝贵的资金来源，但他们也占受访者所支持的网络伙伴的四分之一。与朋友、同事、邻居或熟人的直接互惠关系也占受访者过去两年内交换活动的很大一部分。这一组网络伙伴是借款的最重要来源（ $37\mathrm{\%}$$37\mathrm{\%}$37%37 \% ），也是受访者给予的“重要和实质性支持”的常见受益者。此外，在受访者的核心家庭中也有支持关系。丈夫或伴侣平均占提供重要

support to the respondents and they form $22\mathrm{\%}$$22\mathrm{\%}$22%22 \% of the network partners that received this kind of support.
支持受访者，他们构成了接受过此类支持的 $22\mathrm{\%}$$22\mathrm{\%}$22%22 \% 网络合作伙伴。

Table 4:  表4：
Shares of empty networks and mean sizes of non-empty networks separated by the number of children belonging to the respondent (natural children, step, and fostered children)
空网络的份额和非空网络的平均规模，以受访者子女的数量（亲生子女、继子和养子）为分隔

\begin{tabular}{|c|c|c|c|c|c|c|c|c|c|c|}
\hline \multirow[t]{3}{*}{} & \multicolumn{5}{|c|}{Access to small help} & \multicolumn{5}{|c|}{Access to important and substantive support}
\hline \multirow[t]{3}{*}{} & \multicolumn{5}{|c|}{访问小帮助} & \multicolumn{5}{|c|}{访问重要且实质性的支持}

\hline & Total & \multicolumn{4}{|c|}{Number of children} & Total & \multicolumn{4}{|c|}{Number of children}
横线 & 总计 & 多栏{4}{|c|}{孩子数量} & 总计 & 多栏{4}{|c|}{孩子数量}

\hline & & 0 & 1 & 2 & $\ge 3$$\ge 3$>= 3\geq 3 & & 0 & 1 & 2 & $\ge 3$$\ge 3$>= 3\geq 3

\hline Empty net Share (in percent) & rks
空行净额 份额（百分比）&rks

6.1 & 7.0 & 5.2 & 7.3 & 9.8 & 7.7 & 6.6 & 7.5 & 8.4 & 9.5

\hline N & 2,002 & 259 & 1,096 & 606 & 41 & 2,001 & 259 & 1,093 & 607 & 42
横线 N & 2,002 & 259 & 1,096 & 606 & 41 & 2,001 & 259 & 1,093 & 607 & 42

\hline Non-empt Mean (Std.dev.) & etworks
\hline 非空均值（标准差）和etworks
3.7

$(2.53)$$(2.53)$(2.53)(2.53) & & [
$$\underline{\begin{array}{c}3.6\\ (2.24)\end{array}}$$$$\underline{\begin{array}{c}3.6\\ (2.24)\end{array}}$$[3.6],[(2.24)]\begin{gathered} 3.6 \\ (2.24) \\ \hline \end{gathered}

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

$$\underline{\begin{array}{c}2.6\\ (1.73)\end{array}}$$$$\underline{\begin{array}{c}2.6\\ (1.73)\end{array}}$$[2.6],[(1.73)]\begin{gathered} 2.6 \\ (1.73) \\ \hline \end{gathered}
]

\hline N & 1,879 & 241 & 1,039 & 562 & 37 & 1,847 & 242 & 1,011 & 556 & 38
hline N & 1,879 & 241 & 1,039 & 562 & 37 & 1,847 & 242 & 1,011 & 556 & 38

\hline & \multicolumn{5}{|c|}{Access to borrowed money} & \multicolumn{5}{|c|}{Important and substantive support given}
\hline & \multicolumn{5}{|c|}{获得借款} & \multicolumn{5}{|c|}{提供重要且实质性的支持}

\hline & Total & \multicolumn{4}{|c|}{Number of children} & Total & \multicolumn{4}{|c|}{Number of children}
横线 & 总计 & 多栏{4}{|c|}{孩子数量} & 总计 & 多栏{4}{|c|}{孩子数量}

\hline & & 0 & 1 & 2 & $\ge 3$$\ge 3$>= 3\geq 3 & & 0 & 1 & 2 & $\ge 3$$\ge 3$>= 3\geq 3

\hline Empty net Share (in percent) & rks ${}^{\text{n}}$${}^{\text{n }}$^("n "){ }^{\text {n }}
空行净额 份额（百分比）&rks ${}^{\text{n}}$${}^{\text{n }}$^("n "){ }^{\text {n }}
12.3 & 8.5 & 11.5 & 14.1 & 28.6 & 12.5 & 11.9 & 12.7 & 11.2 & 31.7

\hline N & 2,006 & 259 & 1,097 & 608 & 42 & 1,965 & 252 & 1,071 & 601 & 41
横线 N & 2,006 & 259 & 1,097 & 608 & 42 & 1,965 & 252 & 1,071 & 601 & 41

\hline Non-empty Mean (Std.dev.) & etwork 2.4 (1.57) & $$\begin{array}{c}2.5\\ (1.92)\end{array}$$$$\begin{array}{c}2.5\\ (1.92)\end{array}$${:[2.5],[(1.92)]:}\begin{gathered} 2.5 \\ (1.92) \end{gathered} & [
非空平均值（标准差）和误差 2.4 (1.57) & & [

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

] & [  ]和[

$$\underline{\begin{array}{c}2.8\\ (1.69)\end{array}}$$$$\underline{\begin{array}{c}2.8\\ (1.69)\end{array}}$$[2.8],[(1.69)]\begin{gathered} 2.8 \\ (1.69) \\ \hline \end{gathered}
]

\hline N & 1,760 & 237 & 971 & 522 & 30 & 1,719 & 222 & 935 & 534 & 28
hline N & 1,760 & 237 & 971 & 522 & 30 & 1,719 & 222 & 935 & 534 & 28

\hline   水平线
\end{tabular}

Table 5:  表5：
Networks’ compositions by relationships of direct and indirect reciprocity
网络的构成取决于直接和间接互惠关系