May be just easier to do this:

# create header first

grep -w SP file1.csv > merged.csv

# grep out the header from each file
for i in {1..4};
do
grep -v SP file${i}.csv >> merged.csv;

done
[download]

A perlish hint after opening each file:

my @f1 = <F1>;
my @f2 = <F2>;
...

my @all;
<strike>
push @all, @f1(1..$#f1),@f2(1..$#f2),@f3(1..$#f3),@f4(1..$#f4);
</strike>


push @all, @f1[1..$#f1],@f2[1..$#f2],@f3[1..$#f3],@f4[1..$#f4];
# header
print "$f1[0]\n";

# lines
print join("\t", @all);
print "\n";
[download]

UPDATE: tiredness!