Improvement of nonprofits through an active survey to donors : The case of four major Spanish non-governmental organisations ( NGOs )

We comment on a survey done to the donors of four major Spanish non-governmental organisations (NGOs). The main objective of the survey is to try to improve their revenue and functionality. We sent it to 38,794 donors and received 4,334 valid responses. The survey may have been capable in i tself of increasing the number of donors that recommend the NGO and to measure it in the same survey. We may have developed the first survey that is also an awareness tool from which effect can be measured in the same survey. We call this type of data collection "active surveys". Also several pieces of advice to NGOs on fundraising policies are derived from the survey, like that NGOs should: concentrate their efforts on the population of over 40 years old; address men and women with the same intensity; concentrate on middle, high and very high socioeconomics neighborhoods etc.


INTRODUCTION
A way of improving NGOs is through making surveys that are useful for them, for example because it may help to increase their resources.In Spain this is a greater need because only 19.4% of adult population made gifts in 2012, while in Germany is 32%, in France 53%, in t he United Kingdom 56% and in Austria 61% (AEFR, 2013).United States' population which donates money to NGOs reached on average the 70%, but in some regions it rose up to 80%, such as in metro Atlanta (Van Slyke and Brooks, 2005).In Mexico, most studies show that it is around 50% (García-Colín and Sordo, 2016).
Table 1 reflects articles that have used surveys relat ed to nonprofits and, especially, to donors.It is the result of searching in scientific bibliographic databases for the terms nonprofit, donors and s urveys, most of t he times combining two of them.Also there are included some others articles that we did find in our researc h.It was a very wide search, although not totally exhaustive.We have included all the surveys that we have found, but many must be missing.We could have done a much more extensive searc h, but we did not bec ause we consider that the sample was large enough to analyze the diversity of surveys related wit h nonprofit and because the article became too long.
In general, the response rates obtained are quite good: 70.5% are equal or above 30 and 42.6% are equal or above 50 and 19.7% are equal or above 70%.The majority of authors have concluded that mail surveys *Corresponding author.E-mail: fernando.molini@uam.es.Tel: 0034 914976637.
Authors agree that this article remain permanently open access under the terms of the Creativ e Commons Attribution License 4.0 International License obtain better response rates than online survey (Ardalan et al., 2007;Shih and Fan, 2009;Denniston et al., 2010;Fan and Yan, 2010;Lin and Van Ryzin, 2012).
Table 1 does not corroborate that conclusion, showing that the best results in this comparison of different studies are obtained first by the face to face method, secondly online and thirdly by phone.In any case, the advantage of mail surveys for obtaining better response rate, found by the majority of the studies, is compensated and overtaken when email surveys are capable of reaching a very much larger number of individuals than mail surveys.A very large number of answers may compensate lower response rates.
Table 1 show s that our survey is the one related to donors and nonprofits w ith the largest amount of valid responses (4,334) of surveys done by the researcher, although not in the case of surveys made by the government, associations or external companies.
In our survey, w e w ere aw are of the problem of social desirable bias ( Hall, 2001;Lee and Woodliffe, 2010) in w hich donors over or under r eport depend on social desirability.We try to minimize this problem by emphasizing that it w ould be completely anony mous, done by an independent institution (a University) and by email, which makes answ ers less personal.We established that computers IP number w ould not be collected.We have tried to develop a survey maybe capable in itself of contributing to the increase of the number of donors that recommend the NGO and to measure it in the same survey.
One of the study objectives w as not only to analyze how nonprofit organizations function, but also to improve their management.Specifically, w e w anted to increase the number of donors that recommend the NGO.This active approach has been adopted before by other researches, including O' Donoghue et al. (2006) w ho, besides increasing know ledge, also pursued the strengthening of nonpr ofits.It is the idea of action research that considers that scholars should try to change a problem and not only to understand it or to make predictions about it.We have tried to develop both an academic and an applied perspective, in w hich w e wanted to r aise aw areness in donors.We have called it active surveys.It is characterized by its attempt of influencing population as much as possible, w hile collecting ideas w hich are scarce and valuable.Active surveys should therefore use the greatest demographic sample possible, the entire population being the ideal target.

RESULTS AND DISCUSSION
To increase the number of donors that recommends the NGO We asked donors: How many times have you recommended t his NGO to family, friends or acquaintances?The category with most answers was "above five", with 35%, but the second category was 0, with 22% of the ans wers.It is surprising that there are more men that never recommended the NGO (25.07%) than women (18.07%) and it is statistically significant: = 28.557,p < 0.001.Table 2 expresses the relationship between the number of recommendations and the question To whic h degree are you willing to recommend this NGO in the future more than up until now?It is seen that 910 donors who have never recommended their NGO answered the question To which degree are you willing to recommend this NGO in the future more than up until now?Of these, 35 consider the chanc es that they will recommend it more in the future are very high, 152 high and 329 average.This reflects a change in attitude that may be, in part, an effect of the survey.If we c onsider the res ults for "very high" and "high," we can estimate that the survey may have helped 187 donors to say that they will recommend the NGO more, although they never did so previously.
21% of the donors who have never recommended their NGO before, say that they are willing to recommend it more in the future to a very high or high degree.If we add half of the donors that have answered average, they would be 352 donors.
In that assumption, 39% of the donors that have never recommended their NGO and have answered this question declare that they are planning to do it more in the future.Also, the survey may have partly influenc ed the attitude of 622 donors with one or more recommendations, who say they will recommend their NGO more to a very high degree and the 1, 443 that say they would do it to a high degree.E ven if only 10% of t he donors that say they will recommend the NGO more really do it, it would be a significant achievement.
The results show that it is possible to develop a survey capable in itself of increasing the number of donors that recommend the NGO and t o measure in the same survey that there will be some increase.There is a significant positive relations hip bet ween the number of recommendations and the willingness to recommend: Kendall's Tau-b = 0.485, p<0.05;Kendall's Tau-c = 0.453, p<0.05;Gamma= 0.632, p<0.05.They all prove that there is a monotonic increasing relationship bet ween the number of previous recommendations and donors' willingness to recommend in the future.

There are more donors among the over-60 population
Table 3 reflects the age of the donors.The proportion of donors and the proportion of Spanish population is statistically different: (5, N = 4,307) = 2. 048.17, p<0.05.The greater group is over 60 years old (31.3% of the donors).That proportion is higher than the number of over-60 year-olds in Spain (22.9%, INE, 2016a).
This data is partly consistent with many studies that have found t hat generosity increases with age (Bekkers, 2006;Bekkers, 2007;Belfield and Beney, 2000;García-Colín and S ordo, 2016) and with AEFR (2015), which states that the average age for donors in Spain is 55 years old.However, it is partly contradictory because, in  relation to the Spanish population, the donors between 51 to 60 years old and between 41 to 50 years old have a higher difference (11.7 and 10.7 respectively), than donors over 60 years old (8.4).This may be explained because t he lat er includes pensioners, who s ee their incomes reduced as their salaries are replaced with retirement pensions.Our data shows that NGOs should pay a special attention to the segment of the population over 40 years old in their fundraising campaigns.Over 60 year-old contribut ors say, by a greater degree than the average of the whole survey, that they will contribut e more to the NGO if they did not have to help relatives or friends in need (46.45% against 18.49%).That reflects a common tendency in S pain: many pensioners are economically supporting their children and grandchildren.This is a consequence of the high unemployment rate, which was 20.9% in the fourth quarter of 2015 (INE, 2016b).In that aspect older people seem to have a greater level of solidarity with their own families.However, donors over 60 will increas e their contribution to the NGO less if their income, or their family income, increased (26.50% against 71.84% ) and some are thinking of reducing their aid in a great er degree than the average (17.54%against 2.90%).In both cases it may be explained by the family problems due to the Spanish economic crisis.In the "under 30 year-olds" category, there are nearly no donors.This is expected as many do not have a job yet, or at least not a stable one.

Age (year-olds) Number Percentage (%) donors Percentage (%) Sp Percentage (%) donors -Percentage (%) Sp
Other studies have shown that donors behave differently according to the age.For example, personal circumstances like having a friend that has died from cancer makes them give more to cancer charities than younger donors (Bennett, 2003).For Rooney et al. (2007), husband's age is irrelevant, but increases in the age of the wife is positively associated to the likelihood of being a donor.

Despite expectations, there seem to be more male than female donors
The majority of the donors that answered our survey are men (59.49% ), and it is statistically significant: 2P(Y > 2,135 | n = 4,271, p = 0.5) < 0.001.This is not what would be expected by the predominant literature, that stated that women are more likely to donate than men (Andreoni et al., 2003;Mesch et al., 2006Mesch et al., , 2011;;Piper and Schnepf, 2008;Rooney et al., 2005Rooney et al., , 2007)).
Also AEFR (2015) contradicts the data of this study research.In their study, they found t hat 58% of Spanish donors are women.And Cnann et al. ( 2011) found that females are 53.1% of the donor base.Nevertheless, some authors found that men donate more than women (Brown-Krus e and Hummels, 1993).Einolf (2011) suggests that females are more motivated to help others, but that "men's advantage in resources and social capital balance out women's advantage in motivations, so men's and women's contributions to charitable giving are similar".Belfield and Beney (2000) conclude that women have a higher probability of giving, but that they are likely to give less.Other authors corroborate that the behavior of men and women is sometimes different, for example "temporarily creat ed social norms influence giving by men significantly more than by women" (Croson et al., 2010).
According to Andreoni and Vesterlund (2001) "when the price of giving is low, men appear more altruistic, and when the price is high, women are more generous."They consider that this may explain why the literature has been divided by inconsistent findings.The lower number of women in our survey (40.51%) may not be explained by having a lower level of solidarity, but maybe because in Spain men are in a much greater degree in charge of the finances of the household.Bryant et al. (2003) suggest that men are asked more than women to donate money because nonprofits know that they are the ones frequently in control of the monetary resources.It can be concluded that NGOs should address equally men and women because there are strong arguments for both of them.

Reasons that would make you increase your degree of economic collaboration with that NGO
Another question was, What would mak e you increase your degree of economic collaboration with that NGO?The low score of this last data corroborates the findings of Wiepking and Bekkers (2010) that state that there is a great consensus in the family about donations.It is of great interest with the fact that there are very few donors that are thinking of reducing their economic contribution, although S pain is imm ersed in an economic crisis.The result of the study is lower than t he one obtained by the AEFR (2013), which found that 6% of t he donors say that they will probably stop donating and 4% that say that they will stop for sure.
Table 4 shows the relationship between the question What is your degree of satisfaction with this NGO? and the donors that, in anot her question answered "I am thinking of reducing t he economic contribution".There is an association between both variables: (4, N=114) = 40.123,p<0.05.
The degree of satisfaction with the NGOs is extremely high, as expected, because if not donors would have terminat ed their donations.There are many more donors with a very high or high degree of satisfaction with t he NGO (87.29%), than with a very low or low satisfaction degree (1.9%).
The degree of satisfaction has less influence on the willingness to contribute, bec ause with a very low satisfaction there is only the 0.65% of the donors, while they are the 7. 02% of those that are thinking of reducing their economic contribution to the NGO.W ith low satisfaction there is only 1.25% of the donors, while they are 16.67% of those that are thinking of reducing their economic contribution to the NGO.
Our survey shows that the NGO can do relatively little to increase donors' contributions, because it depends more on external factors, like the economic growth and the rate of unemployment.The main possible measure to be implemented by NGOs is to let donors know that the NGO is spending their money more efficiently (17.5%).Donors' main financial concern is the expenses dedicat ed to their mission, in relation to others such as fundraising costs and other administrative costs (Hyndman, 1991).Higher efficiency in terms of lower total operating costs is positively associated with private donations (Marcuello and Salas, 2000;Xie et al., 2016).For Trussel and Parsons' (2008, p. 264) "efficiency, stability, reputation and the amount of available information" are the four relevant factors which exert an influence on donors ' or potential donors' decisions.For Khumawala and Gordon (1997) donors are more interested in nonfinancial information such as mission and accomplishments than in financial statements.Zainon et al. (2011) found that in Malaysia for institutional donors the most important criteria were basic background, financial and future information.
Those donors who think that they will increase their contribution if they were sure that the money is spent more efficiently consider to a greater degree than the average of t he whole survey, that the information that they receive from the NGO is scarce (12.02% against 4.98%).
Another factor in which NGOs could exert influence is to dedicate the money of the donors to the project that they consider of major interest (12.39% ).Although this percentage is not very big, it should be taken into serious consideration as it is one of the few measures that NGOs can take in order to increase its donors' gifts.
The geographical area of the NGOs' projects is an issue of less import ance, whic h matters to only 6.85% of donors.Cont ribut ors may have their geographical preferences, but they do not care much about them, among other reasons because many may consider that the NGO has a better knowledge of where the great est necessities are.In general NGOs have small need to condition t he geography of t heir expenses to their donor opinions.
It is also of very high interest that only 4.21% of t he donors will not increase their contribution under any circumstances and that the 71.84% declare t hat they would do it if their income increased.With the mitigation of the economical crisis, it will be more realistic to ask current donors for more money.
Some donors (18.49%), besides supporting the NGOs, are also helping relatives and friends in need.When t he economic crisis comes to an end, these donors will be more predisposed to greater contributions to nonprofit organizations.NGOs may concent rate on new donors, especially the ones that gain employment, were promoted, got married, etc.Although it is complicated for NGOs to get such specific information, it may be wort hwhile to try, for example asking for the collaboration of employment agencies or large companies.
For NGOs it is very useful to know that the 8.74% of the donors are willing to increase their economic contribution with the rate of inflation, even if it is a low percentage.When recruiting new donors the NGOs could ask them if t hey wanted their cont ribution to be increas ed automatically by the rate of inflation every year.E ven if few agree, it may be worthwhil e because it usually implies more revenues without the nee d of asking for them regularly.
Women may increase their contribution more than men if they stopped having to help relatives or friends in need (20.93% against 16.64%) and if their income, or that o f their families, increased (74.21% against 70.40% ).The differenc es are statistically significant: = 15.221,p < 0.001 and = 11.218,p < 0.001, res pectively.Men may increase their contribution more than women if tax deductions were higher (17.28% against 12.52% ) and if the money were dedicated to t he project that they consider of most interest (13.61% against 10.72%).The differenc es are statistically significant: = 11.484,p = 0.001 and = 4.627, p = 0.031, respectively.

Satisfaction with the information that donors receive
We asked donors how they feel about the information they receive from the NGOs.The responses were: very excessive (0.46%); excessive (5.59%); adequate (88.31%); scarce (4.98% ) or very scarce (0.65%).The great majority of donors are satisfied, but even so the NGOs Moliní et al. 7 should work on obt aining better results.There is a 11.68% of donors that disagree with the amount of information received.NGOs should know which donors want a greater or lesser feedback and act in consequence.Some NGOs offer their donors an easy option to receive less information, but many others do not.

Preferred way to be contacted for an increase in donor contributions
Another question was, If every t wo years this NGO gets in contact with you to ask for an increase in your contribution, what would you prefer?They could mark one option.The results were: 58.78% preferred to be contacted by email; 29.02% preferred not to be contacted; 7.19% to be contacted by post mail; 4% to be contacted by telephone; and 1.01% to be contacted by a social network in a private way.That shows the need for NGOs to differentiate donors from the very beginning.When they become donors and every time they ask donors for increases in their contribution, the NGO should ask the way that they would prefer to be asked in the future for new increases, if any.They should not bother the 29% that preferred not to be contacted, which is an important percentage.Some may contact the NGO by themselves any way, but the NGO should not call by telephone (as they often do) the donors that preferred to be contacted by email.
The majority of donors preferred to be contacted by email, and that is a problem because the NGO may not even know if they open the messages.In any case, when a NGO sends several personal emails to their donors asking for an increase in their contribution but receive no answer, it may be more ethical to respect donors preferences and to not call them if t hey have not expressed that they wanted to be contacted t hat way.For Sargeant (2001) it is essential that donors are given t he choice of when and how frequently they are solicited, and the format of those communications.
Men preferred to be contacted by email more than women (61.23% against 55.26%) and women preferred in a greater manner not to be contacted for an increase in the contribution (31.18% against 27.5%).Both differenc es are statistically significant: = 14.937, p < 0.001 and = 6.674, p = 0.01.

Number of NGOs to which donors donate
We also wanted to know if there was strong competition between NGOs in getting and keeping donors.We consider that the competition would be low if a great percentage of donors contributed to only one NGO.We asked, To how many NGOs are you a donor on a regular Int.NGOJ.basis?Only 15.28% donated to one NGO, 22.69% to two, 19.14% to three and 42.90% to four or more.This last percentage is surprisingly high and implies that NGOs should be extremely careful in keeping their donors.It may also show that donors prefer to diversify their donations, among other reasons because in case of fraud in one or more NGO, there are more chances that at least some of their money reac hes its destination.Andreoni et al. (2003) s aid that women donate to a larger number of charities than men, but give a lower amount to each of them.On the contrary, our survey did not find any significant difference in the number of NGOs to which men and women donate.Rooney et al. (2007) in the field of education found that "income has a positive effect on the likelihood of being a donor, but surprisingly, wealth (excluding home) does not matter".On the contrary, some authors have concluded that there is no statistical significant relatio n bet ween income and donations (B elfield and Beney, 2000).That may be because it is a complex relationship.Schervish and Havens (1998) state that at all levels of income and wealt h there are high and low givers, and that popular comments about the stingy wealthy and generous poor are clearly exaggerated.

Economic level of donors
For A uten and Rudney (1990) the generosity of t he wealt hy is actually due to the great generosity of a small portion of them, rather than a wides pread attitude.In relation to their income, low income donors donate more than middle class donors.According to Van Slyke and Brooks (2005), "In the Unit ed States, people with the lowest and highest incomes donate the highest percentages of their taxable inc omes-generally, between 3 and 8%-whereas those in the middle donate less than 2%."For Schervish and Havens (1998), the percentage donated by the income groups below $100,000 is around 2% or a little less.But in relation to total contribution, people with high incomes provide t he greater proportion of the donations to charities.In another study, Schervish and Havens (2001, p. 22) stated that "families with a net worth of $1 million or more made up 46% of the total contributions to charitable organizations in 1994".The same authors found that just 3.5 perc ent of households give about 35% of all charit able dollars and that the top 20% gave between 52 and 64% (Schervish and Havens, 1998).
In relation to inc ome we asked, What type of socioeconomic level is the neighborhood where you live?As far as we know, it is the first time the question of income has been asked in such a way.We did it because it is indelicate to ask donors directly about income as they often do not feel comfortable answering it.Donors respond about their income relatively little, for example in the case of Van Slyke and Brooks (2005) only 53%.Our question obt ained the following responses: very low, 0.64%; low, 7.80%; average, 73.57%; high, 16.91%; and very high, 1.09%.There are donors even in the low and very low neighborhoods, although in a very small proportion.This reflects that there are people with cert ain levels of solidarity despite their poor socioeconomic conditions.But as can be expected, there are more donors in rich neighborhoods than in poor ones.
NGOs should focus their recruitment efforts on average, high and very high socioeconomic level neighborhoods, because: there are more donors; it can be assumed that in general they will donat e more; and AEFR (2015) have found t hat the dropping -out probability of donors with low level of income is above the high incomes ones.In some aspects, donors that live in different types of neighborhoods have different behaviors, as Table 5 shows.
The main difference is that people who live in neighborhoods of very high or high socioeconomic level would increase their cont ribution more than the people that live in very low or low socioeconomic level neighborhoods if the tax deductions were higher (difference of 7.14) and if they were sure that the money is spent more efficiently (difference of 6.75).On the contrary, they would increase their economic contribution less than the people who live in very low and low socioeconomic level neighborhoods if their income, or that of their family increased (difference of -7.11), and if they stopped having to help relatives or friends in need (difference of -2.93).
Difference between both types of neighborhoods is that the donors who live in low and very low socioeconomic level neighborhoods collaborate to a greater degree with only one NGO (around 22% against 15% on average), and to a lesser degree with more than three NGOs (37% against 43% on average).That is what can be expected.In any case, it is surprising that there is such a high percentage of donors in low and very low socioeconomic level neighborhoods which contribute to more than three NGOs, although it is lower than in rich neighborhoods.

Conclusions
This research shows that it may be possible to develop a survey capable in itself of increasing the number of donors that recommend the NGO, and it may be the first time that it has been applied.It can be seen that 910 donors, who admitted to never having recommended their NGO, answered the question: To which degree are you willing to recommend this NGO in the future more than up until now?Of these, 35 considered that the chances of recommending it more in the future are very high, 152 high and 329 average.There seems to be a change in attitude that may be partly an effect of t he survey.If this were true, the survey has also become a tool for awareness: it not only measures reality, but also influences society and measures how it is influencing it.That should be of great interest to the NGOs and a major contribution to practice.Also, among the surveys that we have found related to donors and nonprofits, ours is the one with the largest amount of valid responses (4,334) of surveys done by the researcher, not in case of using surveys from institutions.
The study data shows that in relation to the Spanish population there are more donors as the age increases, but only until a certain level.It seems that NGOs should concentrate their fundraising efforts on the population of over 40 years old.
The advice that we provide to NGOs in these conclusions are contributions to practice.Against what would be expected by the predominant literature, the majority of the donors that answer our survey are men (59.49%), and it has statistical significance.From our analysis it can be concluded that NGOs should address men and women with the same intensity in their fundraising campaigns.
With a great difference, the main reason that would mak e donors increase their degree of economic collaboration wit h the NGO is "If my income, or the income of my family, increased" (71.84% of the answers).NGOs cannot act upon the main reasons that would raise donors' contributions, because they depend of factors beyond its control, like the evolution of the economic crisis.There is not an answer in relation to which the NGOs can act until the third place answer: "If I was surer that money is spent more efficiently" (17.05% of the answers).
A great majority of the donors (88.31% ) consider that the information they receive from the NGOs is adequate.A great proportion of donors (42.90%) donate to four or more NGOs.The preferred way by donors to be contacted for an increase in their contribution is by email (58.78%).

Int. NGOJ.
NGOs should try to do as much as possible what donors prefer because it is more ethical.The NGO may have more s uccess in asking for economic increases by telephone rather than by email, but it should not do it if it is not the choice of the donor.
There are donors in all types of socioeconomic neighborhoods, even in the ones with very low and low levels, although in that case in a very small proportion.This reflects that there are people with certain levels of solidarity even if their socioeconomic conditions are difficult.In any case, there are more donors in middle class and rich neighborhoods than in poor ones.For t he NGOs it would be more efficient to concentrate their fundraising campaigns in middle and high neighborhoods because of the number of donors, and in very high neighborhoods because of the amount of the donations.
We conclude that the management of NGOs should promote active surveys, as they can help to increase their revenue and to improve their activity.

Table 1 .
Surveys related w ith nonprofit.
*Survey not made by the author (made by the government, associations or external companies); ** because possible overlap in associational membership.

Table 2 .
Number of recommendations (Willing to recommend).
Source: Ow n survey.

Table 3 .
Age of the donors and % of Spanish population (Sp).

Table 4 .
Degree of satisfaction w ith the NGO (*I am thinking of reducing the economic contribution).

Table 5 .
Socioeconomic level of the neighborhoods *Reasons for increasing your donation.

for increasing your donation High and very high socioeconomic level neighborhood % (number of responses) Low and very low socioeconomic level neighborhood % (number of responses) Difference in % between high and very high, and low and very low
Source: Ow n elaboration.