The research, published by the York-based Joseph Rowntree Foundation, also finds that changes that might be expected to lead to lower poverty, such as reducing the pay gap between men and women or re-balancing regional growth, are by themselves unlikely to do so.
It said key changes forecast in the labour market over the next decade including those in the type of jobs available and expected general improvements in the qualifications and skills of the workforce, will tend to widen inequalities rather than reduce them, continuing recent trends.
It found the main reason for the projected rise in poverty rates is that earnings are forecast to grow in real terms over the decade, while benefits typically will not.
As earnings are a more important income source for middle and top earners than those at the bottom of pay scales, this causes inequality and relative poverty to rise.
Projected changes to employment, such as the continuing growth of both high-skilled and low-skilled jobs at the expense of those in the middle, will also cause relative poverty to rise slightly, said the report, which calculated there will be an extra 1.5 million jobs by 2020.
Chris Goulden, poverty programme manager at the foundation, said: “We need to think again about what would be most effective in reducing poverty.
“Under current tax and benefit policies, and with the labour market we expect to have by 2020, both relative and absolute poverty rates for families will rise to around one in four.”
Prof Rob Wilson, from the Warwick Institute for Employment Research, who led the work, said: “Previous research on the importance of skills for improving individual pay and employment prospects suggested that investment in skills could help alleviate poverty and reduce inequality.
“Our results suggest that things are much more complicated than that.”
Robert Joyce, a report author from the Institute for Fiscal Studies, said: “This highlights the need for governments to be clear about how labour market aims fit with objectives relating to poverty and inequality and to not assume that the two are complementary.
“Policy-makers may well believe that certain labour market changes are desirable even if they do not reduce measures of poverty or inequality.
“But it is vital that such trade-offs are acknowledged – if overlooked, then governments will underestimate how much they need to do via other means to achieve goals such as the large reductions in child poverty currently being sought by 2020.”
Researchers tried to model outcomes to see whether certain improvements to the labour market such as reducing the gender pay gap or improving qualifications would reduce poverty, but none was found to have a significant impact.
They said low earners are not concentrated in low-income households, but are spread across the whole of the household income distribution which meant that increasing pay for all low-earners is not well-targeted on low-income households and can even make inequality worse.
Although an anticipated 1.5m new jobs were built into the forecast, the study projects that many individuals in relative poverty in 2020 will still be in households where no-one works, and so are not affected by changes in the labour market structure, suggesting more jobs are needed. Half of all individuals earning the bottom fifth incomes live in households with no workers.