Spectrophotometers measure the amount of light at a very specific wavelength transmitted through the sample. A turbidimeter only needs to know how much light is scattered (wavelength is relatively unimportant). So to make a turbidimeter, you need a light source (a simple red LED would work), a few lenses (focus light), and a detector (there are some very simple/cheap light detectors out there). Put it all in a box with the light source/sample/detector forming a right angle, put a readout on the front, and you have it.
Precise turbidimeters are still pretty expensive - LaMotte sells one for $720. However, if you sacrifice some of the precision (only use one wavelength, less concern over stray light or stability), you could cut the cost greatly.
For a spectrophotometer you need a light source that provides the specific wavelength(s) you need and then you need to filter/isolate to only allow one wavelength to hit the detector at a time - that is often the expensive part. If your spectrophotometer only looks at one wavelength (not adjustable), you just put in the right light source and filter. If you want to be able to look at a wide range of wavelenghts, you need a white light source and variable wavelength monochromator - much more expensive.
For example, Fisher Scientific sells LaMotte single-parameter colorimeters for $300-$450 each (depends on the chemical to be analyzed), but sells a LaMotte 4-wavelength colorimeter (evidently enough for analysis of at least 16 different chemicals) for $704.24. A true variable-wavelength spectrometer (the Spectronic Genesys 20 - from 325 to 1100nm) goes for $1725.
72g bowfront planted, CO2, 4x - T5HO, Eheim 2213 and 2217, 2 angels, pristella tetras, blue tetras, betta, albino bristlenose pleco, albino cories. Sword, vals, hygros, ludwigias, java moss and fern, anubias
2g Mac-quarium. Clown gravel, fluorescent plastic plants, and 2 guppies.