It is not, necessarily
The initial capital cost of a plant is not the same thing as its lifecycle costs. The best way to analyze the cost impact of an energy source is to look at the total expected costs over the lifetime of the plant, and compare to expected energy generation. This gives you a simple measure that you can use to compare systems.
There are three main variables in the lifetime cost of a power source: capital costs (upfront money to build the plant), operations and maintenance (paying to keep the plant operational) and fuel. Obviously, not all plants have fuel costs.
The other thing to keep in mind is that the numbers calculated by various groups use various methodologies, and these methodologies are surely tweaked to represent the user's ideological interests. Therefore, it is smart to get the opinion of both ends of the spectrum in our polarized world; the truth probably lies somewhere in the middle.
Data Sources
Here is a graphic from the Institute for Energy Research, which is generally aligned with free market energy solutions. However, the underlying data comes from the EIA's 2013 report.
The color key has an error where Fixed O&M and Capital cost are swapped.
The next set of information comes from Energy Innovation Policy and Technology, LLC; with the underlying data from Lazard's 2014 cost of energy analysis. There was no graphic I could embed, so you'll have to settle for a link.
Summary
We can compare the two sources to see the range of estimates for the total lifetime energy expense of various power sources in dollars per MWh:
EIA Lazard's
Conventional Coal 100.1 110
Gas Combined Cycle 67.1 73
Nuclear 108.4 113
Onshore Wind 86.6 60
Geothermal 89.6 116
Solar Thermal 261.5 124
Solar PV 144.3 80
Here is the same comparison for initial capital investment:
EIA Lazard's
Conventional Coal 65.7 77
Gas Combined Cycle 15.8 39
Nuclear 83.4 92
Onshore Wind 70.3 48
Geothermal 76.2 81
Solar Thermal 214.2 107
Solar PV 130.4 72
Discussion
There are three relevant things to point out.
Firstly, both data sets are obviously biased for/against renewable energy. While fossil fuel and nuclear costs are similar, there are widely different interpretations of the costs of wind and solar. I'm not particularly interested in attempting to interpret differing methodologies, so lets just say the truth is probably somewhere in the middle.
Second, the linked example is Solar Thermal. This type of plant is the most expensive for both lifetime return on the dollar and initial capital investment, for both datasets.
Third, given that solar thermal is more expensive, it appears that the quoted statement could be true about solar thermal compared to fossil fuel for initial capital investment according to the Lazard's source, or it could be true about total lifetime levelized cost according to the EIA. So either interpretation could be supported with the data.
Hope that helps.