The majority of Americans do not appear to have done much retirement planning, according to the Financial Industry Regulatory Authority’s (FINRA’s) report, “Financial Capability in the United States 2016.”

Despite the study’s finding of overall improvement in Americans’ ability to make ends meet, the percentages of those who have planned for retirement or have a retirement account are little changed since 2009. In 2009 37% had tried to figure out their retirement savings needs, compared to 39% in 2015. Fifty-seven percent of Americans had employer-sponsored or individual retirement accounts in 2009, while 58% had them in 2015.

Respondents with lower income levels are much less likely to be prepared for retirement than those with higher incomes. Only 19% of those with incomes less than $25,000 have tried to plan for retirement, compared to 60% of those with $75,000 or more in income. Similarly, the likelihood to have a retirement account increases dramatically with income, such that only a small minority of respondents with less than $25,000 in income have a retirement account (18%), while a strong majority of respondents with $75,000 or more income have one (87%).

Respondents who feel they are working towards long-term financial goals are much more likely than those who do not to have calculated retirement savings needs (53% vs. 17%) and to have a retirement account (67% vs. 42%). Among respondents who reported planning for time horizons of more than 10 years, more than half (57%) have tried to calculate their retirement needs, and nearly three-quarters (73%) have a retirement account.

The study finds that more than half of Americans (56%) are worried about running out of money in retirement. Women are somewhat more likely than men to be worried about running out of money in retirement (59% vs. 53%, respectively). Respondents ages 35 to 54 are the most likely to be worried (65%), followed by those 18 to 34 (57%). While those in the highest income group ($75,000 or more) are less likely than those with lower incomes to be worried about retirement, more than half of them are worried.

Among non-retired respondents, those who have tried to calculate retirement savings needs are slightly more likely than those who have not to be worried about having enough money in retirement (64% vs. 59%). Non-retired respondents with retirement accounts are just as worried as those without retirement accounts (62% vs. 60%).

Respondents who receive retirement income (pension plans, Social Security, or withdrawals from retirement accounts) are the most likely to have no difficulty making ends meet. This may be due in part to the higher likelihood of having multiple income sources among these respondents, the report says. For example, among those receiving pension payments, 95% have at least one other additional source of income (out of the seven listed in the survey). In contrast, among those receiving salaries or wages, only 55% have more than one source of income.

Nearly two-thirds (65%) of respondents receiving income from a traditional pension plan report no difficulty making ends meet. Fifty-seven percent of those receiving Social Security retirement benefits report no difficulty, as do 55% of those receiving withdrawals from retirement accounts (e.g., 401(k), IRA, Keogh).

An important determinant of how people choose to invest their savings and retirement wealth is their attitude towards financial risk. Only about one in five Americans (21%) say they are willing to take financial risks. However, risk tolerance has increased considerably relative to 2009, when 12% of respondents reported being willing to take risks. Men are much more likely than women to say they are willing to take risks in financial investments (28% vs. 14%, respectively).

To evaluate financial knowledge, respondents were exposed to a series of questions covering fundamental concepts of economics and finance that may be encountered in everyday life, such as calculations involving interest rates and inflation, principles relating to risk and diversification, the relationship between bond prices and interest rates, and the impact that a shorter term can have on total interest payments over the life of a mortgage. The survey reveals relatively low levels of financial literacy among Americans as measured by these standard questions. While the correct response to some individual questions reaches 75%, only 14% of respondents were able to answer all five questions correctly, and 37% were able to answer at least four questions correctly.

Slightly less than one-third of respondents (31%) report having been offered financial education at a school, college, or workplace, and 21% say they participated. On the surface, exposure to financial education appears to be associated with better performance on the financial literacy quiz questions. Respondents who stated that they participated in financial education score higher than those who were offered but did not participate, who in turn score higher than those who were not offered financial education.

However, the report says it is important to note that these findings do not imply a causal relationship between financial education and financial literacy, and may be entirely attributable to differences in education, employment, and other demographic factors. It is also possible that those who are more interested in financial literacy might be more likely to seek out financial education.